YOU’VE JUST PURCHASED

MORE THAN
A TEXTBOOK
ACTIVATE THE COMPLETE LEARNING EXPERIENCE
THAT COMES WITH YOUR BOOK BY REGISTERING AT

http://evolve.elsevier.com/Canada/LoBiondo/Research
Student Resources

• Audio Glossary provides pronunciations for selected key

terms

• Research Article Library provides practice in reviewing

and critiquing

• Writing Proposals for Funding provides a brief outline to

help you write research proposals

REGISTER TODAY!
NURSING RESEARCH
IN CANADA
Methods, Critical Appraisal,
and Utilization
This page intentionally left blank
 NURSING RESEARCH
IN CANADA
Methods, Critical Appraisal,
and Utilization

GERI LoBIONDO-WOOD, RN, PhD, FAAN JUDITH HABER, PhD, APRN, BC, FAAN
Director of Nursing Research and Evidence-Based The Ursula Springer Leadership Professor in Nursing
Practice, Planning and Development Associate Dean for Graduate Programs
The University of Texas New York University
MD Anderson Cancer Center College of Nursing
Houston, Texas New York, New York
Adjunct Associate Professor
University of Texas Health Sciences Center
School of Nursing
Nursing Systems and Technology
Houston, Texas

Canadian Editors
CHERYLYN CAMERON, RN, PhD MINA D. SINGH, RN, PhD
Associate Vice President Coordinator, Program Evaluation Unit
University Partnership Centre, Research and Scholarship York Institute for Health Research
Georgian College Toronto, Ontario
Barrie, Ontario
Associate Professor
School of Nursing
York University
Toronto, Ontario

Third Canadian Edition

Copyright © 2013 Elsevier Canada, a division of Reed Elsevier Canada, Ltd.

Adapted from Nursing Research: Methods and Critical Appraisal for Evidence-Based Practice, 7th edition, by
Geri LoBiondo-Wood and Judith Haber. Copyright © 2010. Previous editions copyrighted 2006, 2002, 1998,
1994, 1990, 1986.

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval
system, without permission in writing from the publisher. Reproducing passages from this book without such
written permission is an infringement of copyright law.

Requests for permission to make copies of any part of the work should be mailed to: College Licensing
Officer, access©, 1 Yonge Street, Suite 1900, Toronto, ON, M5E 1E5. Fax: (416) 868-1621. All other
inquiries should be directed to the publisher.

Every reasonable effort has been made to acquire permission for copyright material used in this text and to
acknowledge all such indebtedness accurately. Any errors and omissions called to the publisher’s attention
will be corrected in future printings.

NOTICES
Neither the Publisher nor the Authors assume any responsibility for any loss or injury and/or damage to
persons or property arising out of or related to any use of the material contained in this book. It is the
responsibility of the treating practitioner, relying on independent expertise and knowledge of the patient,
to determine the best treatment and method of application for the patient.
—The Publisher

Library and Archives Canada Cataloguing in Publication

LoBiondo-Wood, Geri
Nursing research in Canada: methods, critical appraisal, and utilization/
Geri LoBiondo-Wood, Judith Haber.—3rd Canadian ed./
Canadian editors, Cherylyn Cameron, Mina D. Singh
Includes biographical references.
ISBN 978-1-926648-54-5
1. Nursing–Research–Canada–Textbooks. I. Haber, Judith II. Cameron, Cherylyn, 1958– III. Singh,
Mina D. IV. Title.
RT81.5.N8724 2012 610.73072’071 C2011-907013-8

Vice President, Publishing: Ann Millar

Managing Editor: Roberta A. Spinosa-Millman
Developmental Editor: Jerri Hurlbutt
Publishing Services Manager: Jeffrey Patterson
Project Manager: Mary G. Stueck
Copy Editor: Anne Ostroff
Proofreader: Claudia Forgas
Cover, Interior Design: Teresa McBryan
Typesetting and Assembly: Toppan Best-set Premedia Limited

Elsevier Canada
905 King Street West, 4th Floor, Toronto, ON, Canada M6K 3G9
Phone: 1-866-896-3331
Fax: 1-866-359-9534

2 3 4 5 16 15 14 13

Working together to grow

libraries in developing countries
www.elsevier.com | www.bookaid.org | www.sabre.org
Contents
PART ONE PART TWO
Research Overview, 1 Qualitative Research, 143
RESEARCH VIGNETTE: How Nurses Use and Respond to RESEARCH VIGNETTE: Creating Qualitatively Derived
Research, 2 Knowledge for a Practice Discipline, 144
Barbara Paterson Sally Thorne

1 The Role of Research in Nursing, 4 7 Introduction to Qualitative Research, 147

Cherylyn Cameron Julie Barroso, Cherylyn Cameron

2 Theoretical Framework, 23 8 Qualitative Approaches to Research, 167

Joan Samuels-Dennis, Cherylyn Cameron Julie Barroso, Cherylyn Cameron

3 Critical Reading Strategies: Overview of the

Research Process, 48 PART THREE
Geri LoBiondo-Wood, Judith Haber, Cherylyn Cameron
Quantitative Research, 195
4 Developing Research Questions,
RESEARCH VIGNETTE: Tackling the Prevention of Falls
Hypotheses, and Clinical Questions, 65
Among Older Adults, 196
Judith Haber, Cherylyn Cameron
Nancy C. Edwards

5 Finding and Appraising the Literature, 90

Stephanie Fulton, Barbara Krainovich-Miller, Cherylyn Cameron
9 Introduction to Quantitative Research, 198
Geri LoBiondo-Wood, Mina D. Singh

6 Legal and Ethical Issues, 112

Judith Haber, Mina D. Singh
10 Experimental and Quasiexperimental
Designs, 216
Susan Sullivan-Bolyai, Carol Bova, Mina D. Singh

11 Nonexperimental Designs, 234

Geri LoBiondo-Wood, Judith Haber, Mina D. Singh

v
vi CONTENTS

PART FOUR PART SIX

Processes Related to Application of Research:
Research, 253 Evidence-Informed Practice, 475
RESEARCH VIGNETTE: Developing and Implementing a RESEARCH VIGNETTE: Nursing Best Practices, 476
Framework for Interprofessional Education in Health Barbara Davies
Sciences Education: Can Nursing Contribute? 254
Julie Gaudet 20 Developing an Evidence-Informed
Practice, 479
12 Sampling, 258 Marita Titler, Susan Adams, Cherylyn Cameron
Judith Haber, Mina D. Singh

13 Data-Collection Methods, 287

Susan Sullivan-Bolyai, Carol Bova, Mina D. Singh
Appendixes
14 Rigour in Research, 306 A Understanding Nursing on an Acute Stroke
Geri LoBiondo-Wood, Judith Haber, Mina D. Singh Unit: Perceptions of Space, Time and
Interprofessional Practice, 514
15 Qualitative Data Analysis, 331 Cydnee C. Seneviratne, Charles M. Mather, Karen L. Then
Cherylyn Cameron
B Decisional Involvement of Senior Nurse
16 Quantitative Data Analysis, 346 Leaders in Canadian Acute Care
Susan Sullivan-Bolyai, Carol Bova, Mina D. Singh Hospitals, 526
Carol A. Wong, Heather Laschinger, Greta G. Cummings,
17 Presenting the Findings, 383 Leslie Vincent, Patty O’Connor
Geri LoBiondo-Wood, Mina D. Singh
C The Effect of Music on Parental
Participation During Pediatric Laceration
PART FIVE Repair, 542
Critiquing Research, 395 Gregory Sobieraj, Maala Bhatt, Sylvie LeMay, Janet Rennick,
Celeste Johnston

RESEARCH VIGNETTE: Evolution in a Program of Research

D Patient Real-Time and 12-Month
Focusing on Family Violence, Caregiving, and Women’s
Retrospective Perceptions of Difficult
Health, 396
Communications in the Cancer Diagnostic
Judith Wuest
Period, 552
Sally Thorne, Elizabeth-Anne Armstrong, Susan R. Harris,
18 Critiquing Qualitative Research, 399 T. Gregory Hislop, Charmaine Kim-Sing, Valerie Oglov,
Helen J. Streubert, Cherylyn Cameron John L. Oliffe, Kelli I. Stajduhar

19 Critiquing Quantitative Research, 434

Nancy E. Kline, Mina D. Singh GLOSSARY, 568

INDEX, 581
Author Biographies
eri Lo iondo ood RN, PhD, FAAN, is the Dr. LoBiondo-Wood has been active locally
Director of Nursing Research and Evidence- and nationally in many professional organiza-
Based Practice, Planning and Development at the tions, including the Southern Nursing Research
MD Anderson Cancer Center, Houston, Texas, Society, the Midwest Nursing Research Society,
and an Adjunct Associate Professor at the niver- and the North American Transplant Coordinators
sity of Texas Health Sciences Center at Houston Organization.
( THSC-Houston), School of Nursing. She She has received local and national awards for
received her Diploma in nursing at St. Mary’s teaching and contributions to nursing. In 1997,
Hospital School of Nursing in Rochester, New she received the Distinguished Alumnus Award
York her Bachelor’s and Master’s degrees from from New York niversity, Division of Nursing
the niversity of Rochester and her PhD in Alumni Association. In 2001 she was inducted as
Nursing Theory and Research from New York a Fellow of the American Academy of Nursing,
niversity. At MD Anderson Cancer Center, Dr. and in 2007 to The niversity of Texas Academy
LoBiondo-Wood developed and implemented the of Health Science Education.
Evidence-Based Resource nit Nurse (EB-R N)
Program, which is a hospital-wide program that udith aber PhD, APRN, BC, FAAN, is the
involves all levels of nurses in the application of rsula Springer Leadership Professor in Nursing
research evidence to practice. She also has imple- and Associate Dean for Graduate Programs in the
mented a mentorship program for nurses wishing College of Nursing at New York niversity. She
to conduct research. Dr. LoBiondo-Wood also received her undergraduate nursing education at
teaches research and evidence-based practice Adelphi niversity in New York, and she holds a
principles to undergraduate, graduate, doctoral, Master’s degree in Adult Psychiatric–Mental
and doctor of nursing practice students at THSC- Health Nursing and a PhD in Nursing Theory and
Houston, School of Nursing. She has extensive Research from New York niversity. Dr. Haber
experience guiding nurses and other health care is internationally recognized as a clinician and
professionals in the development and utilization educator in psychiatric–mental health nursing.
of research in clinical practice. Dr. LoBiondo- She has extensive clinical experience in psychi-
Wood is currently a member of the Editorial atric nursing, having been an advanced practice
Board of Progress in Transplantation and a psychiatric nurse in private practice over 30
reviewer for Nursing Research and Nephrology years, specializing in treatment of families coping
Nursing ournal Her research and publications with the psychosocial sequelae of acute and
focus on chronic illness and the impact of solid chronic catastrophic illness. Dr. Haber is cur-
organ transplantation on pediatric or adult recipi- rently on the Editorial Board of the ournal of
ents and their families throughout the transplant the American Psychiatric Nurses Association
process. At MD Anderson Cancer Center, her APNA Her areas of research involvement
research focuses on symptom clusters in adult include tool development, particularly in the area
patients with cancer. of family functioning. She is internationally
vii
viii AUTHOR BIOGRAPHIES

known for developing the Haber Level of Dif- in Education (OISE)/ niversity of Toronto. Her
ferentiation of Self Scale. Another program of dissertation, The Lived Experience of Transfer
research addresses physical and psychosocial Students from a Baccalaureate Nursing Program,
adjustment to illness, focusing specifically on won the Best Dissertation award from the Council
women with breast cancer and their partners. On for Study of Community Colleges in the nited
the basis of this research, she and Dr. Carol States. Dr. Cameron currently teaches courses as
Hoskins have written and produced an award- an adjunct professor for Central Michigan ni-
winning series of evidence-based psycho- versity in the Master of Arts degree in education
educational videotapes, ourney to Recovery: or program with a community college specializa-
omen ith Breast Cancer and Their Partners, tion. At Georgian College, she works with uni-
which has been tested in a randomized clinical versity partners to provide degree-level studies
trial funded by the National Cancer Institute. for college students and is responsible for research
Dr. Haber has been active locally and nation- and scholarship at the college. Her research inter-
ally in many professional organizations, including ests include college-university relations, student
the American Nurses Association, the American experiences, transfer from college to university,
Psychiatric Nurses Association (APNA), and the and the implementation of best practice
American Academy of Nursing. She has received guidelines.
numerous local, state, and national awards for
public policy, clinical practice, and research, ina Singh RN, PhD, is an Associate Profes-
including the APNA Psychiatric Nurse of the sor at York niversity School of Nursing and
Year Award in 1998 and 2005, the APNA Out- Coordinator of the Program Evaluation nit
standing Research Award in 2005, and the 2007 at the York Institute for Health Research. She
New York niversity College of Nursing Distin- received her Bachelor of Science in Nursing
guished Alumnus Award. In 1993, she was degree from the niversity of Toronto and started
inducted as a Fellow of the American Academy her clinical nursing neurosurgery program but
of Nursing. spent most of her career in mental health and
public health nursing. Her interests in program
CANADIAN EDITORS evaluation encouraged her to pursue her educa-
Cherylyn Ca eron RN, PhD, is the Associate tion in quantitative research. She earned her doc-
Vice President of the niversity Partnership torate in measurement and evaluation from the
Centre, Georgian College. She received her Bach- niversity of Toronto. Her current interests, in
elor of Science in Nursing degree from the ni- her role as a specialist in quantitative methods,
versity of Alberta and a Master of Arts degree in relate to program evaluation in health care and
education from Central Michigan niversity. She nursing education. She also conducts research in
received her doctorate in theory and policy studies chronic disease prevention and cultural diversity
in education from the Ontario Institute for Studies in health care.
Contributors
Susan Adams, RN, PhD Stephanie Fulton, MSIS
Associate Director Assistant Library Director
Research Translation and Dissemination Core Research Medical Library
Gerontological Nursing Interventions Research Center The University of Texas
College of Nursing MD Anderson Cancer Center
University of Iowa; Houston, Texas
Director
National Nursing Practice Network Julie Gaudet, RN, MN (Admin.)
Iowa City, Iowa Professor
School of Nursing
Julie Barroso, PhD, ANP, APRN, BC, FAAN Centre for Health Sciences
Associate Professor and Specialty Director George Brown College
Adult Nurse Practitioner Program; Toronto, Ontario
Research Development Coordinator
Office of Research Affairs Nancy E. Kline, PhD, RN, CPNP, FAAN
Duke University School of Nursing Director, Center for Evidence-Based Practice and Research
Durham, North Carolina Department of Nursing
Memorial Sloan-Kettering Cancer Center
Carol Bova, PhD, RN, ANP New York, New York
Associate Professor of Nursing and Medicine
Graduate School of Nursing Barbara Krainovich-Miller, EdD, APRN, BC,
University of Massachusetts, Worcester ANEF, FAAN
Worcester, Massachusetts Clinical Professor
College of Nursing
Barbara Davies, RN, PhD New York University
Professor New York, New York
School of Nursing
Faculty of Health Sciences Barbara Paterson, RN, MEd, PhD
University of Ottawa; Professor and Dean
Co-Director, Nursing Best Practice Research Unit Thompson Rivers University
University of Ottawa and RNAO School of Nursing
Ottawa, Ontario Kamloops, British Columbia

Nancy C. Edwards, RN, BScN, MSc, PhD Joan Samuels-Dennis, RN, PhD
Full Professor Assistant Professor, Faculty of Health
Department of Epidemiology and Community Medicine School of Nursing
School of Nursing York University
University of Ottawa Toronto, Ontario
Ottawa, Ontario

ix
x CONTRIBUTORS

Helen J. Streubert, EdD, RN, CNE, ANEF Marita Titler, RN, PhD, FAAN
Vice President of Academic Affairs Professor of Nursing, Rhetaugh Dumas Endowed Chair
Our Lady of the Lake University Associate Dean of Practice and Clinical Scholarship
San Antonio, Texas Development
University of Michigan School of Nursing
Susan Sullivan-Bolyai, DNSc, CNS, RN Ann Arbor, Michigan
Associate Professor
Graduate School of Nursing and Department of Pediatrics Judith Wuest, RN, BScN, MN, PhD
University of Massachusetts, Worcester Professor Emerita, Faculty of Nursing
Worcester, Massachusetts University of New Brunswick
Fredericton, New Brunswick
Sally Thorne, RN, PhD, FCAHS
Professor
School of Nursing
University of British Columbia
Vancouver, British Columbia
Reviewers
Davina Banner, RN, PhD Kathy Quee, RN, MSN
Assistant Professor Clinical Placement Advisor
School of Nursing Chair, Research Ethics Board
University of Northern British Columbia British Columbia Institute of Technology
Prince George, British Columbia Burnaby, British Columbia

Barbara Brady-Fryer, BSc, MN, PhD, RN Louise Racine, RN, MScN, PhD
Instructor Associate Professor
Baccalaureate in Nursing Program College of Nursing
Grant MacEwan University University of Saskatchewan
Edmonton, Alberta Saskatoon, Saskatchewan

Shelley Cobbett, MN, RN, Gnt, EdD Jasna K. Schwind, RN, PhD
Adjunct Assistant Professor Associate Professor
School of Nursing—Yarmouth Campus Daphne Cockwell School of Nursing
Dalhousie University Ryerson University
Halifax, Nova Scotia Toronto, Ontario

Cheryl Forchuk, RN, MScN, PhD Kathy F. Spurr, BSc, RRT, MHI, FCSRT
Associate Director, Nursing Research Assistant Professor
Arthur Labatt Family School of Nursing School of Health Sciences
The University of Western Ontario Dalhousie University
London, Ontario Halifax, Nova Scotia

Angela Gillis, RN, PhD Dr. Darlene Steven, RN, PhD

Professor Full Professor
School of Nursing School of Nursing
St. Francis Xavier University Lakehead University
Antigonish, Nova Scotia Thunder Bay, Ontario

Patricia Grainger, RN, BN, MN Lela V. Zimmer, BScN, PhD

Nurse Educator Assistant Professor
Centre for Nursing Studies School of Nursing
St. John’s, Newfoundland University of Northern British Columbia
Prince George, British Columbia
Cindy Murray, RN, MN, PhD
Assistant Professor
School of Nursing
Memorial University of Newfoundland
St. John’s, Newfoundland

xi
Acknowledgements
This major undertaking was accomplished with Anne Ostroff, Copy Editor, and Mary Stueck,
the help of many people, some of whom made Senior Project Manager, whose attention to
direct contributions to the new edition and some detail ensured that we had a very readable
of whom contributed indirectly. We acknowledge text for our students
with deep appreciation and our warmest thanks Teresa McBryan, Design Manager, who
the following people who made this third redesigned this edition so that it is a plea-
Canadian edition possible: sure to look at and read
Nursing educators across Canada who pro- Our vignette contributors, whose willing-
vided valuable and insightful comments ness to share their wisdom and evidence of
that helped to direct the revisions featured their innovative research made a unique
in this edition and contributed to improving contribution to this edition
the content All of the reviewers, who provided thought-
Our students, particularly the nursing ful feedback not only on the first and second
students of the Georgian College, Seneca editions but also on the third Canadian
College, and York niversity Second-Entry edition manuscript
programs, who inspired us with their feed- Our families, who supported us and picked
back and willingness to use the informa- up the loose ends while we wrote and
tion in this text by becoming research revised:
assistants
Ann Millar, Publisher, Elsevier Canada, To my husband John Cameron, who has
who had faith in us and provided much- always been there loving, supporting, and
needed guidance and discussion encouraging me. Thanks for keeping our
Roberta A. Spinosa-Millman, Managing family and household together so that I
Editor, who got us started with encourage- could stay focused on this third edition.
ment, a sense of humour, and great insight Cherylyn Cameron
Jerri Hurlbutt, Developmental Editor, who
encouraged us with positive feedback, made To my husband Neranjan, my daughter
sense of the process, and extended dead- Sandhya, and my parents Ram and Betty
lines graciously Laljie, for their support and encouragement.
Mina D. Singh

xii
Preface
The foundation of the third Canadian edition of not only in the undergraduate nursing research
Nursing Research in Canada: Methods, Critical course but also throughout the curriculum. The
Appraisal, and tili ation continues to be the research role of baccalaureate graduates calls for
belief that nursing research is integral to all levels evidence-informed practice competencies central
of nursing education and practice. Since the first to this are critical appraisal skills: that is, nurses
edition of this textbook, we have seen the depth should be competent research consumers.
and breadth of nursing research grow: More Preparing students for this role involves devel-
nurses are conducting research and using research oping their critical thinking and reading skills,
evidence to shape clinical practice, education, thereby enhancing their understanding of the
administration, and health policy. research process, their appreciation of the role of
The Canadian Nurses Association promotes the critiquer, and their ability to actually appraise
the notion that nurses must provide care that is research critically. An undergraduate course in
based on the best available scientific evidence. nursing research should develop this basic level
This is an exciting challenge to meet. Nurses are of competence, which is an essential requirement
using the best available evidence, combined with if students are to engage in evidence-informed
their clinical judgement and patient preferences, clinical decision making and practice. This is in
to in uence the nature and direction of health care contrast to a graduate-level research course, in
delivery and to document outcomes related to the which the emphasis is on carrying out research,
quality and cost effectiveness of patient care. As as well as understanding and appraising it.
nurses continue to develop a unique body of The primary audience for this textbook remains
nursing knowledge through research, decisions undergraduate students who are learning the steps
about clinical nursing practice will be increas- of the research process, as well as how to develop
ingly evidence informed. clinical questions, critically appraise published
As editors, we believe that all nurses not only research literature, and use research findings to
need to understand the research process but also inform evidence-informed clinical practice. This
need to know how to critically read, evaluate, and book is also a valuable resource for students at
apply research findings in practice. We realize the master’s and doctoral levels who want a
that understanding research, as a component of concise review of the basic steps of the research
evidence-informed practice, is a challenge for process, the critical appraisal process, and the
every student, but we believe that the challenge principles and tools for evidence-informed
can be accomplished in a stimulating, lively, and practice.
learner-friendly manner. This text is also a key resource for doctoral
Consistent with this perspective is a commit- students who are preparing to be experts at leading
ment to advancing implementation of the evi- evidence-informed initiatives in clinical settings.
dence-informed practice paradigm. nderstanding Furthermore, it is an important resource for prac-
and applying nursing research must be an integral tising nurses who strive to use research evidence
dimension of baccalaureate education, evident as the basis for clinical decision making and
xiii
xiv PREFACE

development of evidence-informed policies, pro- By emphasizing the role of evidence-

tocols, and standards, rather than rely on tradi- informed practice as the basis for informing
tion, authority, or trial and error. It is an important clinical decision making and nursing inter-
resource for nurses who collaborate with nurse- ventions that support nursing practice,
scientists in the conduct of clinical research and demonstrating quality and cost-effective
evidence-informed practice. outcomes of nursing care delivery.
Building on the success of the second edition, By presenting numerous examples of
we maintain our commitment to introduce evi- recently published research studies that
dence-informed practice and research principles illustrate and highlight each research
to baccalaureate students, thereby providing a concept in a manner that brings abstract
cutting-edge research consumer foundation for ideas to life for students new to the research
their clinical practice. and critical appraisal process. These exam-
Nursing Research in Canada: Methods, Criti- ples are a critical link for reinforcement of
cal Appraisal, and tili ation prepares nursing evidence-informed concepts and the related
students and practising nurses to become knowl- research and critiquing process.
edgeable nursing research consumers in the fol- By showcasing, in esearch ignettes the
lowing ways: work of renowned nurse researchers whose
By addressing the evidence-informed prac- careers exemplify the links among research,
tice role of the nurse, thereby embedding education, and practice.
evidence-informed competence in the clini- By providing numerous pedagogical chapter
cal practice of every baccalaureate graduate. features, including Learning Outco es
By demystifying research, which is some- ey Ter s ey Points new Critical
times viewed as a complex process. Thin ing Challenges el ul ints Evi
By using an evidence-informed approach to dence In or ed Practice Ti s new Prac
teaching the fundamentals of the research tical A lications, and revised Critical
process. Thin ing ecision Paths as well as
By teaching the critical appraisal process in numerous tables, boxes, and figures. At the
a user-friendly but logical and systematic end of each chapter that presents a step of
progression. the research process, we feature a revised
By promoting a lively spirit of inquiry that section titled A raising the Evidence
develops critical thinking and critical which reviews how each step of the research
reading skills, facilitating mastery of the process should be evaluated from a con-
critical appraisal process. sumer’s perspective. This section is accom-
By developing information literacy, search- panied by an updated Criti uing Criteria
ing, and evidence-informed practice com- box.
petencies that prepare students and nurses By providing a Study uide that promotes
to effectively locate and evaluate the best active learning and assimilation of nursing
available research evidence. research content.
By elevating the critical appraisal process By offering an Evolve site presenting free
and research appreciation to a position of Evolve esources or Instructors that
importance comparable to that of producing includes a Test Bank, an Instructor’s
research. Before students become research Manual, PowerPoint slides, critiquing exer-
producers, they must become knowledge- cises, an Image Collection, and new critical
able research consumers. appraisal activities. There are also Evolve
 PREFACE xv

resources for both the student and faculty and Appraising the Literature, showcases cut-
that include a research article library, an ting-edge information literacy content, providing
audio glossary, and instructions on how to students and nurses with the tools necessary to
write proposals for funding. effectively search, retrieve, manage, and evaluate
The third edition of Nursing Research in research studies and their findings. This chapter
Canada: Methods, Critical Appraisal, and tili- also develops research consumer competencies
ation is organized into six parts. Each part is that prepare students and nurses to critically read,
preceded by an introductory section and opens understand, and appraise a study’s literature
with an exciting Research Vignette by a review and framework. The final chapter in this
renowned nurse researcher. section, Chapter 6, Legal and Ethical Issues,
Part One esearch Overvie contains six provides an overview of the increased emphasis
chapters. Chapter 1, The Role of Research in on the legal and ethical issues facing researchers
Nursing, provides an excellent overview of in Canada.
research and evidence-informed practice pro- Part T o ualitative esearch contains
cesses that shape clinical practice. This chapter two interrelated qualitative research chapters.
introduces the role research plays in practice and Chapter 7, Introduction to ualitative Research,
education, the roles of nurses in research activi- provides a framework for understanding qualita-
ties, a historical perspective, and future directions tive research designs and literature, as well as the
in nursing research. The style and content of this significant contribution of qualitative research to
chapter are designed to make subsequent chapters evidence-informed practice. Chapter 8, ualita-
more user-friendly. Chapter 2, Theoretical tive Approaches to Research, presents, illus-
Framework, focuses specifically on how theo- trates, and, in examples from the literature,
retical frameworks guide and inform knowledge showcases major qualitative methods. This
generation through the research process. Chapter chapter highlights the questions most appropri-
3, Critical Reading Strategies: Overview of the ately answered through the use of qualitative
Research Process, addresses students directly methods.
and highlights critical thinking and critical Part Three uantitative esearch contains
reading concepts and strategies, thereby facilitat- Chapters 9 ( Introduction to uantitative
ing students’ understanding of the research Research ), 10 ( Experimental and uasiexperi-
process and its relationship to the critical appraisal mental Designs ), and 11 ( Nonexperimental
process. This chapter introduces a model evi- Designs ). These chapters delineate the essential
dence hierarchy that is used throughout the text. steps of the quantitative research process, with
The next two chapters address foundational published, current clinical research studies used
components of the research process. Chapter 4, to illustrate each step. Links between the steps
Developing Research uestions, Hypotheses, and their relationship to the total research process
and Clinical uestions, focuses on how research are examined.
questions, hypotheses, and evidence-informed Part our Processes elated to esearch
practice questions are derived, operationalized, describes the specific steps of the research process
and critically appraised. Numerous clinical exam- for qualitative and quantitative studies. The chap-
ples illustrating different types of research ters make the case for linking an evidence-
questions and hypotheses maximize student informed approach with essential steps of the
understanding. Students are also taught how to research process by teaching students how to
develop clinical questions that are used to guide critically appraise the strengths and weaknesses
evidence-informed inquiry. Chapter 5, Finding of each step of the research process. Students
xvi PREFACE

learn how to select participants (Chapter 12, to reading and understanding qualitative and
Sampling ), gather data (Chapter 13, Data- quantitative research literature and evaluating its
Collection Methods ), analyze the results strengths and weaknesses. Extensive Internet
(Chapter 15, ualitative Data Analysis, and resources are provided on the accompanying
Chapter 16, uantitative Data Analysis ), and Evolve site that can be used to develop evidence-
present their results (Chapter 17, Presenting the informed knowledge and skills.
Findings ). Chapter 14, Rigour in Research, The Evolve Web site that accompanies the
gives students the tools for assessing the quality third Canadian edition provides interactive learn-
and trustworthiness of a study. ing activities that promote the development of
Part ive Criti uing esearch makes the critical thinking, critical reading, and information
case for linking an evidence-informed approach literacy skills designed to develop the competen-
with essential steps of the research process by cies necessary to produce informed consumers
teaching students how to critically appraise the of nursing research. Instructor resources are
strengths and weaknesses of each step of the available at a passcode-protected Web site that
research process. Each chapter critiques two gives faculty access to all instructor materials
examples of actual published research. Chapter online, including the Instructor’s Manual, Image
18, Critiquing ualitative Research, focuses on Collection, PowerPoint Slides, a Test Bank that
qualitative research, whereas Chapter 19, Cri- allows faculty to create examinations through
tiquing uantitative Research, is based on the the use of the ExamView test generator program,
quantitative research process. and more.
Part Si A lication o esearch Evidence The development and refinement of an evi-
In or ed Practice contains the final chapter in dence-informed foundation for clinical nursing
the book. Chapter 20, Developing an Evidence- practice is an essential priority for the future of
Informed Practice, provides a dynamic review professional nursing practice. The third Canadian
of evidence-informed models. These models can edition of Nursing Research in Canada: Methods,
be applied—step by step, at the organizational or Critical Appraisal, and tili ation will help stu-
individual patient level—as frameworks for imple- dents develop a basic level of competence in
menting and evaluating the outcomes of evidence- understanding the steps of the research process
informed health care. that will enable them to critically analyze research
Stimulating critical thinking is a core value of studies, evaluate their merit, and judiciously
this text. Innovative chapter features such as apply evidence in clinical practice. To the extent
Critical Thinking Decision Paths, Evidence- that this goal is accomplished, the next generation
Informed Practice Tips, Helpful Hints, Practical of nursing professionals will include a cadre of
Applications, and Critical Thinking Challenges clinicians who inform their practice by using
enhance critical thinking and promote the devel- theory and research evidence, combined with
opment of evidence-informed decision-making their clinical judgement, and specific to the health
skills. To be consistent with previous editions, care needs of patients and their families in health
we now promote critical thinking by including and illness.
sections called Appraising the Evidence,
which describe the critical appraisal process Cherylyn Cameron
ccameron@georgianc.on.ca
related to the focus of the chapter. In addition,
Critiquing Criteria are included in this section to Mina Singh
stimulate a systematic and evaluative approach minsingh@yorku.ca
To the Student
We invite you to join us on an exciting nursing this text are designed to help you develop your
research adventure that begins as you turn the first skills in critical thinking, critical reading, infor-
page of the third Canadian edition of Nursing mation literacy, and evidence-informed clinical
Research in Canada: Methods, Critical Appraisal, decision making, while providing a user-friendly
and tili ation The adventure is one of discov- approach to learning that expands your compe-
ery You will discover that the nursing research tence to deal with these new and challenging
literature sparkles with pride, dedication, and experiences. The companion Study uide, with
excitement about the research dimension of pro- its chapter-by-chapter activities, will serve as a
fessional nursing practice. Whether you are a self-paced learning tool to reinforce the content
student or a practising nurse whose goal is to use of the text. The accompanying Evolve Web site
research evidence as the foundation of your prac- offers summative review material to help you
tice, you will discover that nursing research and reinforce the concepts discussed throughout the
a commitment to evidence-informed practice book.
positions our profession at the forefront of change. Remember that evidence-informed practice
You will discover that evidence-informed prac- skills are used in every clinical setting and can be
tice is integral to meeting the challenge of provid- applied to every patient population or clinical
ing quality health care in partnership with patients practice issue. Whether your clinical practice
and their families and significant others, as well involves primary care or specialty care and pro-
as with the communities in which they live. vides inpatient or outpatient treatment in a hospi-
Finally, you will discover the richness in the tal, clinic, or home, you will be challenged to
who, what, where, when, why, and apply your evidence-informed practice skills and
how of nursing research and evidence-informed use nursing research as the foundation for your
practice, and you will develop a foundation of evidence-informed practice. The third Canadian
knowledge and skills that will equip you for clini- edition of Nursing Research in Canada: Methods
cal practice today and into the future. Critical Appraisal, and tili ation will guide you
We think you will enjoy reading this text. Your through this exciting adventure, where you will
nursing research course will be short but filled discover your ability to play a vital role in con-
with new and challenging learning experiences tributing to the building of an evidence-informed
that will develop your evidence-informed prac- professional nursing practice.
tice skills. The third Canadian edition of Nursing
Research in Canada: Methods, Critical Appraisal, Cherylyn Cameron
ccameron@georgianc.on.ca
and tili ation re ects cutting-edge trends for
developing evidence-informed nursing practice. Mina D. Singh
The six-part organization and special features in minsingh@yorku.ca

xvii
This page intentionally left blank
1

PA RT ON E
PART ONE Part Title

Research Overview

1 The Role of Research in

Nursing
2 Theoretical Framework
3 Critical Reading Strategies:
Overview of the Research
Process
4 Developing Research
Questions, Hypotheses,
and Clinical Questions
5 Finding and Appraising the
Literature
6 Legal and Ethical Issues
RESEARCH VIGNETTE facilities and, to a lesser extent, in
surgical units. Therefore, the
How Nurses Use and Respond existing research did not address
the nurses’ particular concerns
to Research pertaining to the nutritional status
of patients who use injection
drugs, late-night admissions of
Barbara Paterson, RN, MEd, PhD Practising nurses’ positive atti- homeless patients who might not
Professor and Dean tudes toward research are not have eaten before admission, and
Thompson Rivers University
unique. During the 3 years I worked patients with neurological deficits
School of Nursing
Kamloops, British Columbia on the unit, I observed as nurses who experienced difficulty han-
searching for research evidence to dling cutlery and cups.
answer their practice questions, Nevertheless, the nurses were
A few years ago, I was fortunate chatting with each other about able to apply some findings of the
to work as a nurse scholar on an research articles, and quoting research conducted in long-term
acute care medical unit in a large research findings to patients and care facilities to their unit. For
urban hospital. This experience other health care professionals. Put example, the nurses learned that if
gave me an opportunity to observe simply, these nurses used research patients are fed by nursing staff
first-hand how nurses use and to do the job of nursing more who are in a hurry, those patients’
respond to research. During that effectively. For example, they con- nutritional status was often poor.
time, one of the staff nurses on the sulted research articles to find Many of the unit’s patients re-
unit made a comment that I believe solutions to several practice issues quired assistance in feeding, and
captures many Canadian nurses’ on the unit, such as how to help an the findings of research conducted
attitudes toward research. She anxious patient cope with invasive in long-term care facilities caused
said, I used to think that nursing procedures how to help nurses the nurses to re ect on the staff
research was bunk. Just something deal with their feelings about who fed the patients and how long
we had to learn in school that I patients who use illegal injection the feeding took. This consider-
could put away with my textbooks drugs and leave the hospital in ation, in turn, caused the nurses to
when I graduated. I thought that compromised medical conditions, change the feeding directions they
we learned research in school against medical advice and how gave to support staff (e.g., unit as-
because the professors were best to structure nursing assign- sistants) and to reinforce the fact
hoping to convince some of us to ments so that patients receive the that it is the nurse’s responsibility
go on to graduate school. I really care they need. to make sure that the patient has
didn’t see it as having anything to Sometimes, when nurses dis- eaten enough.
do with the real world of bedside covered that little research existed The nurses also initiated
nursing. She added that because about the issues that arose, they research on the unit. For example,
of our regular research discus- were forced to imagine how other they were continually frustrated
sions on the unit and our continual research might apply. For example, by the behaviour of many patients
efforts to use research findings in when the nurses became con- who use injection drugs. The nurses
the care of patients, her attitude cerned that some patients were reported that these patients were
about research had changed. She not eating enough and not eating often hostile, manipulative, and
said, When I saw research as rel- nutritionally sound foods, they noncompliant. Hospital statistics
evant to what I did and not just consulted the research about the demonstrated that this particular
some pie-in-the-sky idea from nutritional status of hospitalized patient population accounted for
school, I began to use research as patients. As it turned out, the most of the nurses’ security calls.
a tool—something that helped me existing research had been con- Several nurses said they dreaded
do my job better. ducted largely in long-term care working with this type of patient.

2
Initially, they suggested strategies investigating whether nurses’ research than are others. In
that were inadequate or unrealistic medication errors could be dec- 3 years, however, the nurses I
(e.g., building a seclusion room in reased by using personal digital worked with began to see research
the unit for patients who use injec- assistants (PDAs), such as Palm in a new light: as something they
tion drugs). ltimately, the nurses Pilot devices. The idea for the needed to care for patients effec-
decided to seek advice from the research project arose when the tively. When I told the nurses
hospital’s chemical dependency nurses discovered that many of mentioned in this vignette that I
experts. After consulting with the the medication errors occurred was writing this piece, one nurse
nurses, the chemical dependency because nurses did not recognize said, They will read it and think
team developed a program of that a new order had been pro- that our stories do not represent
workshops and conducted weekly vided, misread the order, or forgot the real world. They will think it
rounds over the course of a year to that a medication had not been can’t happen in most hospitals or
help nurses better understand and administered. The nurses submit- for most nurses. They will think
care for patients who use injection ted a proposal for the research that research took place in our
drugs. project and received funding for it case only because you were here
Because other units had similar from the niversity of British with us and you understood re-
issues with the same patient popu- Columbia. The first component of search. If that nurse is correct and
lation, the nurses decided to report this research consisted of observ- that is what you are thinking, I
the results of the program to others ing the nurses’ administering invite you to consider the follow-
in a tangible way, as research find- medication and writing documen- ing: To inspire nurses to use re-
ings. The nurses formally evalu- tation. On the basis of the observa- search as a method of helping
ated the program that had been tions, the nurses then developed them do their jobs more effective-
created by the chemical dependen- a program for PDAs that told ly, what if all we need is someone
cy experts. For example, the nurses nurses which medications should who believes that research can do
administered questionnaires and be given and which had not yet just that What if that someone
participated in group interviews. been administered. This program is you Are you prepared to be
The research revealed that most included the ability to link to other the nurse who says to colleagues,
nurses no longer dreaded caring information about patients and Let’s see what research has been
for patients who used illegal injec- their medications. In the end, the done about that. I bet someone has
tion drugs. The nurses better un- research confirmed that the use of studied it and that the results will
derstood the behaviour of these PDAs did reduce medication be helpful to us.
patients and what could be done to errors. Subsequently, the research After reading this section and
help them. After a presentation of findings were submitted to a pub- this book, I hope that you will use
the research findings, the nurses lishing company that may incor- research as the foundation of your
celebrated their accomplishments porate them into new PDA nursing practices. I have seen how
of the program with a cake and software for nurses. much research in uences practis-
coffee party on the unit. Commitment to research does ing nurses, and I’m here to tell you
Another example of nurses’ not occur overnight, and some that it works. ■
initiating research involved nurses are more excited by

3
C H A PTER 1

The Role of Research in Nursing

Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• State the significance of research to the practice of nursing.
• Identify the role of the consumer of nursing research.
• Discuss the differences in trends within nursing research in Canada.
• Describe how research, education, and practice are related to each other.
• Evaluate the nurse’s role in the research process as it relates to the nurse’s level of
education.
• Identify future trends in nursing research.
• Formulate the priorities for nursing research in the twenty-first century.

KEY TERMS
consumer evidence-informed practice phenomena
data generalizability research
evidence-based practice

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

4
 CHAPTER 1 The Role of Research in Nursing 5

WE INVITE YOU TO JOIN US on an exciting nursing of congestive heart failure but not with
research adventure that begins as you read the another
first page of this chapter. The adventure is one of What is the effect of using computers to
discovery You will discover that the nursing educate children about self-management of
research literature sparkles with pride in, dedica- asthma
tion to, and excitement about this dimension of What is the experience of men who undergo
professional nursing practice. As you progress prostate surgery and therapy
through your educational program, you are taught What is the quality of studies on therapeutic
how to ensure quality and safety in practice touch
through acquiring knowledge of the various sci- What nursing-delivered smoking cessation
ences and health care principles. Another compo- interventions are most effective
nent critical for clinical knowledge is research This book will help you begin your adventure
knowledge as it applies to practising from an into evidence-informed nursing practice by giving
evidence-informed approach. you the tools to use research as a foundation for
Whether you are a student or a practising nurse evidence-informed practice.
whose goal is to use research as the foundation
of your professional practice, you will discover SIGNIFICANCE OF RESEARCH AND
that nursing research and evidence-informed EVIDENCE-INFORMED PRACTICE IN
practice position the nursing profession at the THE FIELD OF NURSING
cutting edge of change and improvement in The health care environment is changing at an
patients’ outcomes. You will also discover that increasingly rapid pace. Hinshaw (2000), in por-
nursing research is integral to achieving the goal traying twenty-first–century nursing, stated that
of providing quality outcomes in partnership with an unprecedented explosion of nursing knowl-
patients, their families and significant others, and edge guided practice and advanced the health and
the communities in which they live. Finally, you well-being of individual clients, families, and
will discover the who, what, where, when, why, communities (p. 117). Carper (1978) described
and how of nursing research and develop a foun- four fundamental patterns of knowing in nursing:
dation of knowledge, evidence-informed prac- (1) empirical knowledge ( empirics ), the science
tice, and competencies that will equip you for of nursing (2) aesthetics, the art of nursing (3)
twenty-first century nursing practice. the component of personal knowing and (4)
Your nursing research adventure will be filled ethics, the component of moral knowledge of
with new and challenging learning experiences nursing. Empirical knowledge, which is based on
that develop your evidence-informed practice research findings, represents one source of
skills. Your critical thinking, critical reading, and knowledge within a larger body of knowledge
clinical decision-making skills will expand as you (Tarlier, 2005, p. 131). The challenges associated
develop clinical questions, search the research with nursing’s rapid pace of growth can best be
literature, evaluate the research evidence found in met by integrating evidence-informed knowledge
the literature, and make clinical decisions about into nursing practice. Nursing research provides
applying the best available evidence to your specialized scientific knowledge that empowers
practice. For example, you will be encouraged nurses to anticipate and meet these constantly
to ask important clinical questions, such as the shifting challenges and maintain the profession’s
following: societal relevance.
What makes an intervention effective with In learning about nursing research, it is impor-
one group of patients who have a diagnosis tant to differentiate among the terms research,
6 PART ONE Research Overview

evidence-based practice, and evidence-informed made, which results in practice that is evidence
practice. esearch is the systematic, rigourous, based. According to the Canadian Nurses Asso-
logical investigation that aims to answer ques- ciation (CNA), Evidence-based nursing refers
tions about nursing phenomena. Pheno ena can to the incorporation of evidence from research,
be defined as occurrences, circumstances, or facts clinical expertise, client preferences and other
that are perceptible by the senses. Phenomena, available resources to make decisions about
such as the expression of pain or loss, are the clients (CNA, 2002). Through research utiliza-
circumstances of interest to nurses. tion efforts, knowledge obtained from research is
There are two types of research: quantitative transformed into clinical practice, which results
and qualitative. You will be introduced to these in nursing practice that is evidence based.
two types of research in more depth in Chapter 2. Evidence in or ed ractice extends beyond
The methods used by nurse researchers are the the early definitions of evidence-based practice
same methods used in other disciplines the dif- described previously. Building upon the founda-
ference is that nurses study questions relevant to tion of evidence-based practice, evidence-
nursing practice. Nurse researchers also conduct informed practice also involves acknowledging
research collaboratively with researchers from and considering the myriad factors beyond such
other disciplines. Through the conducting of evidence as local indigenous knowledge, cultural
research, they produce knowledge that is reliable and religious norms, and clinical judgement. With
and useful for clinical practice. The methods and evidence-informed practice, the methods for
findings of studies provide evidence that is evalu- gathering evidence (use of published research
ated, and their applicability to practice is used to studies) are the same as the processes used for
inform clinical decisions. evidence-based practice however, the evidence
In the past 20 years, many health care disci- also incorporates expert opinion, clinical exper-
plines have adopted the tenets of evidence- tise, patient preference, and other resources
informed practice to provide the best health (CNA, 2010b). It is important to remember that
care possible for their patients. The roots of evidence-informed practice focuses on a more
evidence-informed practice stem from Dr. Archie inclusive and interactive process:
Cochrane’s investigation of the efficacy of health
care, particularly in the work of the medical pro- Evidence-informed decision-making is a continuous
interactive process involving the explicit, conscientious
fession. His work resulted in the establishment of and judicious consideration of the best available
the Cochrane Collaboration, which provides sys- evidence to provide care. It is essential to optimize
temic reviews of health care interventions. In outcomes for individual clients, promote healthy
1996, Sackett and colleagues defined evidence- communities and populations, improve clinical
based medicines as the conscientious, explicit, practice, achieve cost-effective nursing care and ensure
accountability and transparency in decision-making
and judicious use of current best evidence in
within the health-care system.
making decisions about the care of individual (CNA, 2010b, p. 1)
patients (p. 71). Since then, most health profes-
sions have adopted the tenets of evidence based For example, to understand the importance of
ractice evidence-informed practice, consider the work of
Much of the evidence used as a basis for prac- Dr. Judith Ritchie, who won the Canadian Health
tice is from research that has been completed, Services Research Foundation’s 2010 Excellence
written about in papers, and then published. Pub- through Evidence Award for her work on the suc-
lished research studies are assessed so that deci- cessful implementation of best practice guide-
sions about application to clinical practice can be lines to reduce falls, manage pain, and protect
 CHAPTER 1 The Role of Research in Nursing 7

skin integrity among patients. It has been esti- literature critically and use it to inform your clini-
mated that as a result of the implementation of cal decisions.
best practices, the incidence of pressure ulcers Throughout this text, the steps of the research
was reduced from 21 to 10.6 in 5 years. Not and evidence-informed practice processes are
only does this result in better outcomes for described. The steps are systematic and orderly
patients, but the potential cost savings are esti- and relate to the development of evidence-
mated at 2.9 million for every 1,000 people informed practice. nderstanding the step-by-
(CNA, 2009). step process that researchers use will help you
When you first read about the research and the develop the assessment skills necessary to judge
evidence-informed practice processes, you will the soundness of research studies. Chapter 20 will
notice that both processes may seem similar. Each describe how you can implement evidence into
begins with a question. The difference is that in practice to improve patient outcomes.
a research study, the question is tested with a Throughout the chapters, research terms perti-
design appropriate for the question and with spe- nent to each step are identified and illustrated
cific methods (sample, instruments, procedures, with many examples from the research literature.
and data analysis). In the evidence-informed Four published research studies are found in the
practice process, a question is used to search the appendixes and used as examples to illustrate sig-
literature for studies already completed that you nificant points in each chapter. Judging not only
will critically appraise in order to answer your the study’s strength and quality but also a study’s
clinical question. applicability to practice is key. Before you can
It is proposed that all nurses share a commit- judge a study, it is important to understand the
ment to the advancement of nursing science by differences between and among studies. Many
conducting research and using research evidence different study designs exist, which you will see
in practice. Scientific investigation promotes as you read through this text and the appendixes.
accountability, which is one of the hallmarks of There are standards not only for critiquing the
the nursing profession and a fundamental compe- soundness of each step of a study but also for
tency for all registered nurses (CNA, 2010a). judging the strength and quality of evidence pro-
What does this mean for you as a nurse There is vided by a study and determining its applicability
a consensus that the research role of the bacca- to practice.
laureate and master’s graduate calls for the skills This chapter provides an overview of the role
of critical appraisal that is, you must be a knowl- that research plays in practice and education, the
edgeable consumer of research, whereby you can roles of nurses in research activities, a historical
appraise research evidence and use existing stan- perspective, and future directions in nursing
dards to determine the merit and readiness of research.
research for use in clinical practice: Nurses
support, use and engage in research and other RESEARCH: THE ELEMENT THAT LINKS
activities that promote safe, competent, compas- THEORY, EDUCATION, AND PRACTICE
sionate and ethical care, and they use guidelines Research links theory, education, and practice.
for ethical research that are in keeping with Theoretical formulations supported by research
nursing values (CNA, 2008, p. 9). Therefore, to findings may become the foundations of theory-
use research (evidence-informed practice), you informed practice in nursing. Your educational
may not necessarily be able to conduct research, setting, whether a nursing program or the health
but you can understand and appraise the steps of care organization where you are employed, pro-
the research process in order to read the research vides an environment in which you, as a student
8 PART ONE Research Overview

or an employee, can learn about the research from each study discussed thus far have clearly
process. In the setting of a nursing program or a demonstrated implications for society and prac-
health care organization, you can also explore tice. In an era of continuing concern about health
different theories and begin to evaluate them in care costs, empirically supported programs that
light of research findings. See the Practical are cost effective without compromising quality
Application box for an example of applying are essential. Many researchers directly evaluate
theory to nursing practice. the cost effectiveness of treatment models for
example, Forchuk and colleagues (2005) found
that a transition discharge model of care for
Practical Application patients experiencing chronic mental illness
helped the patients achieve discharge from a psy-
Consider the research program that focused
on building knowledge about improving the chiatric hospital early, by an average of 116 days,
health and health care of women who experience which resulted in considerable cost savings.
intimate partner abuse. This program extended for Several studies have demonstrated that the use of
more than a decade and focused on the experiences
of women who left abusive partners, the health of research-based interventions is more likely to
the women after leaving the relationships, and the result in better outcomes than traditional or ritual-
degree of health care improvement for women after based nursing care (McGinty Anderson, 2008
they left their abusive partners (Ford-Gilboe, Wuest,
Varcoe, & Merrit-Gray, 2006). The knowledge gained Williams, 2004).
from this body of interrelated research led to the At this point in your study of nursing
development of comprehensive interventions to research, you may be wondering how education
support the health and quality of life of women who
in nursing research links theory and practice.
leave abusive partners.
The findings of that 2006 study have had The answer is twofold. First, learning about
enormous implications both for nurses and for other nursing research will provide you with an
health care professionals, who make up the appreciation and understanding of the research
interdisciplinary health care team that works with
women who experience intimate partner violence. process so that you can more easily become a
Meaningful intervention is crucial for these women participant in research activities. Second, learn-
because health problems persist after they leave their ing the value of nursing research helps you to
partners. Many of these women report continued
abuse and harassment from their ex-partners and become an intelligent consumer of research. A
experience financial hardships. Although health care consu er of research actively uses and applies
workers have an interest in improving health care for research. To be a knowledgeable consumer, you
these women, most health care workers do not have
must have knowledge about the relevant subject
the knowledge required to address their patients’
needs in a meaningful and effective way. As the matter, the ability to discriminate and to evalu-
researchers noted, “the vast majority of health ate information logically, and the ability to
practitioners [are] unprepared to recognize and apply the knowledge gained. You need not
respond to [intimate partner violence] in ways that
are sensitive to the complexity of women’s actually conduct research to be able to appreci-
experiences and respectful of women’s safety and ate and use research findings in practice. Rather,
choices” (Ford-Gilboe et al., 2006, p. 148). to be an intelligent consumer, you must under-
stand the research process and develop the criti-
cal evaluation skills needed to judge the merit
The example in the Practical Application and relevance of evidence before applying it to
box is an attempt to answer a question that you practice. The success of evidence-informed
may have asked before taking this course: How practice depends on your ability, as a consumer
will the theory and research content of this of research, to understand the research process
course relate to my nursing practice The data and to evaluate the evidence.
 CHAPTER 1 The Role of Research in Nursing 9

research about health risks and risk/harm reduction to

ROLES OF THE NURSE IN THE support evidence-informed practice (e.g., second-hand
RESEARCH PROCESS smoke, Pap smears). [Competency HW-20]
Every nurse practising in the twenty-first century On a provincial level, each province in Canada
has a role to play in the research process. The has its own standards for entry into nursing prac-
Canadian Association of Schools of Nurses tice, and many of these standards have specific
(CASN) (n.d.) states that it strongly supports related research competencies. For example, the
nursing research and the training of nurse College of Nurses of Ontario (2009) outlined the
researchers across Canada and works to encour- following competencies that incorporate research:
age and improve the quality of nursing research, Engages in nursing or health research by reading and
as well as to promote research findings and critiquing research reports and identifying research
support Canada’s current and upcoming nursing opportunities. [Competency 35]
research leaders. More specifically, CASN’s Supports involvement in nursing or health research
objectives include the education of nurse scholars/ through collaboration with others in conducting
research, participating in research, and implementing
researchers securing of sufficient funding for
research findings into practice. [Competency 36]
nursing scholarship/research enhancement of
scholarship/research cooperation dissemination Nurses must also be intelligent consumers of
of scholarship/research and cooperation with research that is, they must understand all steps
other nursing and non-nursing organizations, pro- of the research process and their interrelation-
vincial, national, and international. ships. The registered nurse interprets, evaluates,
The Canadian Registered Nurse Examination and determines the credibility of research find-
(CRNE) outlines competencies related to research ings. The nurse discriminates between interesting
on which examination questions for licensing to findings for which further investigation is required
practise are based. The following sections from and those that are sufficiently supported by evi-
the Registered Nurse examination competencies dence before applying findings to practice. The
explicitly highlight the importance of research for nurse should then use these competencies to
practice (CNA, 2010a): advance the nursing or interdisciplinary evidence-
informed practice projects (e.g., developing clini-
Professional Practice
. . . Registered nurses are expected to use knowledge
cal standards, tracking quality improvement data,
and research to build an evidence-informed or coordinating implementation of a pilot project
practice . . .. The registered nurse . . . uses evidence to test the efficacy of a new wound care protocol)
and critical inquiry to challenge, change, enhance or of the workplace committees to which he or she
support nursing practice (e.g., questioning accepted belongs.
practice, participating in research). [Competency PP-15]
Registered nurses are also responsible for gen-
Nurse–Client Partnership erating clinical questions to identify nursing
. . . The registered nurse . . . supports the informed issues that necessitate investigation and for par-
choice of the client in making decisions about care (e.g., ticipating in the implementation of scientific
right to refuse, right to request care, right to choose,
right to participate in research). [Competency NCP-9]
studies. Clinicians often generate research ideas
or questions from hunches, gut-level feelings,
Nursing Practice: Health and Wellness intuition, or observations of patients or nursing
. . . Nursing practice is influenced by continuing
care. These ideas often become the seeds of
competency, determinants of health, life phases,
demographics, health trends, economic and research investigations.
political factors, evidence-informed knowledge and For example, Paterson (2002) related the story
research . . . . The registered nurse . . . incorporates of Vini and Suzanne, two nurses working in an
10 PART ONE Research Overview

intensive care unit. They observed that the blood the findings of a research report that was cri-
pressure of a certain patient dropped each time tiqued, found to have merit, and believed to have
he received acetaminophen for a fever. He then the potential for application to practice. In a more
needed to receive additional treatment for the cor- formal way, it may involve joining a health care
responding low blood pressure. This situation agency’s research committee or its quality assur-
raised several questions for Vini and Suzanne, ance or quality improvement committee, in which
and they wondered whether treating the fever was research articles, integrative reviews of the litera-
truly beneficial and what the effects of leaving the ture, and clinical practice guidelines are evalu-
fever untreated would be. They spoke to other ated for evidence-informed clinical decision
health care professionals, conducted a literature making.
review, and concluded that treatment with anti- Nurses who have graduate degrees must
pyretics, such as acetylsalicylic acid, should not also be sophisticated consumers of research
be applied immediately and that, in some patients, and are specially prepared to conduct research
leaving the fever untreated had clear benefits. as co-investigators or primary investigators. At
Vini and Suzanne subsequently disseminated the master’s level, nurses are prepared to be
their research findings to their colleagues and active members of research teams. Nurses with
other health care professionals at a critical care master’s-level training can assume the role of
conference. The example of these two nurses clinical expert, collaborating with an experienced
exemplifies the circularity between practice that researcher in proposal development, data collec-
generates research and research that improves tion, data analysis, and interpretation. Nurses
practice. Systematic collection of data about a with master’s degrees enhance the quality and
clinical problem, such as the one identified by relevance of nursing research by providing not
Vini and Suzanne, contributes to the refinement only clinical expertise but also evidence-informed
and extension of nursing practice. knowledge about the way clinical services are
Registered nurses may participate in research delivered. Nurses with master’s-level training
projects as members of interdisciplinary or intra- also facilitate the investigation of clinical prob-
disciplinary research teams in one or more phases lems by enabling a climate that is open to nursing
of a project. For example, a staff nurse may work research and by engaging in evidence-informed
on a clinical research unit in which a particular practice projects. At the master’s level, nurses
type of nursing care is part of an established conduct research investigations to monitor the
research protocol (e.g., for pain management, quality of nursing in clinical settings and to help
prevention of falls, or treatment of urinary incon- others apply scientific knowledge to nursing
tinence). In situations such as these, the nurse practice.
administers care according to the format described To achieve the greatest expertise in appraising,
in the protocol. The nurse may also be involved designing, and conducting research, nurses must
in collecting and recording data relevant to the complete PhDs. Nurses with doctoral degrees
administration of, and the patient’s response to, develop theoretical explanations for phenomena
nursing care. relevant to nursing, develop methods of scientific
As important as the generation of research is inquiry, and use a variety of methods to modify
the sharing of research findings with colleagues. or extend existing knowledge so that it is relevant
Examples of such sharing include developing an to nursing (or to other areas of health care). In
article or presentation for a research or clinical addition to their role as researchers, nurses with
conference on the findings of a study and sharing doctoral-level training act as role models and
 CHAPTER 1 The Role of Research in Nursing 11

mentors to guide, stimulate, and encourage other

nurses who are developing their research skills. HISTORICAL PERSPECTIVE*
Nurses with doctoral degrees also collaborate and Nursing research has undergone many changes
consult with social, educational, and health care and developments since the 1970s. In addition,
institutions or governmental agencies in their the groundwork for the research that exists today
respective research endeavours. These nurses was laid in the late nineteenth century and
then disseminate their research findings to throughout the twentieth century. Box 1-1 shows
the scientific community, clinicians, and—as the major milestones in the development of
appropriate—the general public through scien- nursing research.
tific journal articles and presentations for nursing
research conferences. Nineteenth Century
Another essential responsibility of nurses, In the mid-nineteenth century, nursing as a formal
regardless of the level of education they achieve, discipline began to take root with the ideas and
is to pay special regard to the ethical principles practices of Florence Nightingale. Her concepts
of research, especially the protection of human have contributed to, and are congruent with, the
participants. For example, nurses caring for current priorities of nursing research. Promotion
patients who are participating in research on of health, prevention of disease, and care of the
antinausea chemotherapy must ensure that the sick were central ideas of her practice and publi-
patients have signed the informed consent form cations. Nightingale believed that systematic col-
and that all their questions are answered by the lection and exploration of data were necessary for
research team before they begin participation. nursing. Her collection and analysis of data on
Furthermore, if patients have an adverse reac- the health status of British soldiers during the
tion to the medication, nurses must not admin- Crimean War led to a variety of reforms in health
ister more doses until they have notified an care. Nightingale (1863) also noted the need for
appropriate member of the research team (see measuring the outcomes of nursing and medical
Chapter 6). Not all nurses must or should care and had expertise in statistics and epidemiol-
conduct research, but all nurses must play some ogy. Other than Nightingale’s work, little research
part in the research process. Nurses at all edu- was conducted during the early years of develop-
cational levels—whether they are consumers, ment of the field of nursing. Schools of nursing
researchers, or both—need to view the research had just begun to be established and were unequal
process as integral to the growing professional- in their ability to educate, whereas nursing leader-
ism in nursing. ship had just started to develop.
As a professional, you must take time to read
research studies and evaluate them, using the Twentieth Century
standards congruent with scientific research. In the twentieth century, research focused mainly
Also, you will need to use the critiquing process on nursing education, but some patient- and
in order to identify the strengths and weaknesses technique-oriented research was also evident (see
of each study. Bearing in mind that each study Box 1-1). In 1923, in the nited States, the
has its limitations, you should consider whether
sound and relevant evidence from one particular
This section (i.e., from pp. 11 to 15) is adapted with
study can be used in other settings as well. permission from Potter, P., Perry, A. G., Ross-Kerr, J.,
Chapter 20 will expand on how to bring research Wood, M. (Eds.), (2010). Canadian fundamentals of
into your nursing practice. nursing (4th ed., pp. 80–82). Toronto: Elsevier Canada.
12 PART ONE Research Overview

BOX 1-1
HISTORICAL MILESTONES IN THE DEVELOPMENT OF NURSING RESEARCH
1858 and 1863 Florence Nightingale publishes Notes on Matters Affecting the Health, Efficiency and Hospital
Administration of the British Army and Notes on Hospitals
1920 Public health courses are offered at the universities of British Columbia, Alberta, Toronto, McGill,
Dalhousie, and Western Ontario
1932 The Weir report, sponsored by the Canadian Nurses Association and the Canadian Medical
Association, calls for better nursing education and service
1952 The American Nurses Association first publishes Nursing Research
1959 The first Canadian nursing master’s degree program is launched at the University of Western Ontario
1964–1965 The first nursing research project is funded by a Canadian federal granting agency
International Journal of Nursing Studies and International Nursing Index are launched
1969–1970 Nursing Papers, the forerunner of the Canadian Journal of Nursing Research, is published at McGill
University
1971 McGill University launches the Centre for Nursing Research, and the first national Canadian conference
on nursing research is held; both are financed by the Department of National Health and Welfare
1978 Heads of university nursing schools and deans of graduate studies attend the Kellogg National
Seminar on Doctoral Education in Nursing
1982 The Alberta Foundation for Nursing Research, the first funding agency for nursing research, is
established
The Working Group on Nursing Research is established by the Medical Research Council of Canada
(MRC)
1985 The report of the Working Group on Nursing Research is released by the MRC
1988 The MRC and the National Health Research and Development Program establish a joint initiative to
structure nursing research grants
1991 The first fully funded Canadian nursing PhD programs are launched first at the University of Alberta,
followed by the University of British Columbia, McGill University, and the University of Toronto
1994 McMaster University launches its nursing PhD; MRC’s mandate includes health research
1999 The Nursing Research Fund is launched with a $25 million grant over 10 years; the Canadian Health
Services Research Foundation (CHSRF) administers the funds
The PhD nursing program is launched at the University of Calgary
2000 Five CHSRF/Canadian Institutes of Health Research (CIHR) Chairs Awards are granted to nursing
2002 The Office of Nursing Policy organizes a think tank called, “Pathfinding for Nursing Science in the 21st
Century,” which advocates for a coordinated voice for nursing science
2004 A forum on doctoral education is held in Toronto under the auspices of the Canadian Association of
Schools of Nursing to develop a national position paper on the PhD in nursing for Canada
The Canadian Consortium for Nursing Research and Innovation is established to develop a strategic
plan, build partnerships, and advocate for funding to support research programs and infrastructure
2003–2010 PhD programs in nursing were initiated at Dalhousie University, Queen’s University, Université Laval,
Université de Montréal, Université de Sherbrooke, University of Victoria, University of Western
Ontario, University of Ontario, and University of Saskatchewan
Adapted from Potter, P., Perry, A. G., Ross-Kerr, J., & Wood, M. (Eds.), (2010). Canadian fundamentals of nursing (4th ed., p. 81). Toronto: Elsevier
Canada.
 CHAPTER 1 The Role of Research in Nursing 13

Committee for the Study of Nursing Education were held annually or biennially for several years.
studied the preparation of educators, administra- Papers of the early conferences were published as
tors, and public health nurses and the clinical monographs. The number of nursing research
experiences of nursing students, publishing the conferences increased and began to be sponsored
results in the Goldmark report (Committee for the by professional, research, and academic organiza-
Study of Nursing Education, 1923). This report tions. In addition, nurses began to participate in
identified gaps in the educational background of the research conferences of interdisciplinary
nurses. As a result of the Goldmark report, the groups, such as the Canadian Association on
method for educating nurses changed, and more Gerontology.
university-based nursing curricula were devel- The two major factors in the development of
oped. A decade later in Canada, a similar compre- nursing research have been the establishment of
hensive study of nursing and nursing education research training through doctoral programs and
was carried out by Dr. George Weir, sponsored the establishment of funding to support nursing
jointly by the CNA and the Canadian Medical research. Throughout the 1970s and 1980s, uni-
Association. The so-called Weir report (Weir, versity faculties and schools of nursing built their
1932) documented serious problems in nursing research resources so that they could mount doc-
education and drew attention to the need for toral programs. The first provincially approved
changes to improve standards in education and doctoral nursing program was established at the
practice. The Weir report’s recommendation that niversity of Alberta Faculty of Nursing in 1991.
authority and responsibility for schools of nursing The niversity of British Columbia School of
be vested within provincial systems of education Nursing was established later that year, and pro-
was revolutionary at the time, and more than half grams at McGill niversity and the niversity of
a century passed before it was fully implemented Toronto followed in 1993. In the late 1990s and
across the country. early 2000s, other programs were launched,
Changes in the educational system for nurses bringing the total to 15.
were crucial for the development of nursing Growing awareness of the importance of
research. In Canada, the establishment of univer- nursing research gradually led to the availability
sity nursing courses in 1918, followed by mas- of research funds. In 1999, in response to inten-
ter’s degree programs in the 1950s and 1960s and sive lobbying by the CNA, the federal govern-
by doctoral programs in the 1990s and 2000s, was ment established the Nursing Research Fund,
key to the development of nursing research. budgeting 25 million for nursing research (i.e.,
The first nursing research journal, Nursing 2.5 million over each of the following 10 years),
Research, was established in the nited States in with the Canadian Health Services Research
1952. The first nursing research journal published Foundation (CHSRF) to administer the funds.
in Canada, Nursing Papers (later called the The research areas targeted for support included
Canadian ournal of Nursing Research), was nursing policies, management, human resources,
established at McGill niversity in 1969. Other and nursing care. Each year, 500,000 is desig-
journals were later established today, nurses nated for the Open Grants Competition, 500,000
publish their research, both within nursing and in for the Canadian Nurses Foundation for research
interdisciplinary fields, in dozens of journals. on nursing care, 500,000 for nursing chairs,
In 1971, McGill niversity established the 750,000 for training (postdoctoral fellowships
Centre for Nursing Research. In the same year, and student grants), and 250,000 for knowledge
the first National Conference on Nursing Research networks and dissemination activities. Five chairs
in Canada was held in Ottawa. These conferences in nursing research were funded by this initiative,
14 PART ONE Research Overview

representing excellence in nursing research across BOX 1-2

Canada (CHSRF, n.d.). The incumbents were NURSING RESEARCH NATIONAL
charged with developing research capacity in one ORGANIZATIONS*
particular area of nursing. Currently, more than • The Canadian Health Services Research Foundation
20 nurse researchers hold research chairs funded (CHSRF) provides evidence-informed decision
by the Canada Research Chairs and private funds making and change in Canada’s health care
system. Until 2009, the funding was distributed
(Jeans, 2005). Box 1-2 lists many of the national
through the Nursing Research Fund to develop
organizations and programs that provide funding nursing researchers and support research in areas
or coordination for nursing research in Canada. such as nursing recruitment, retention,
nfortunately, the Nursing Research Fund expired management, and emergent issues from health
service restructuring. CHSRF also supports the
in 2009. Several nursing organizations are cur- Research, Exchange, and Impact for System
rently lobbying intensively for a new government Support (REISS) program to maximize the effect of
fund of 55 million over another 10 years. nursing research on decision making within the
Canadian health care system.
Nursing research has focused progressively on • The Canadian Institutes of Health Research (CIHR)
evidence-informed practice in response to provide funding for several institutes, including the
demands to justify care practices and systems by Institute of Health Services and Policy Research,
that are dedicated to supporting innovative
improving patient outcomes and controlling costs.
research, capacity building, and knowledge transfer
The scope of nursing research has also broadened to improve the delivery of health care services.
to include historical and philosophical inquiry. • Partnership programs funded by CHSRF and CIHR
The establishment of the Centre for Philosophical include the Capacity for Applied and
Developmental Research and Evaluation (CADRE)
Nursing Research at the niversity of Alberta program and the Partnerships for Health System
exemplifies this new direction. Training centres Improvement. These programs provide awards to
have also been funded by the CHSRF to increase individuals and teams of researchers.
• The Canadian Nursing Foundation (CNF) depends
training for graduate nursing students and to on the contributions of the public and private
increase research capacity in nursing and related sectors to provide funding to support research
disciplines. For example, the Ontario Training addressing policy and to support clinical nursing
research. The CNF also provides scholarships for
Centre in Health Services and Policy Research
undergraduate and graduate education.
comprises six Ontario universities for the purpose • The Canadian Association for Nursing Research
of enhancing health services and policy research. (CANR) is a national organization whose mission is
The centre’s headquarters are located at McMaster to support research-based nursing practice and
provides awards to recognize the outstanding
niversity and involve collaboration with the ni- research contribution of nurses.
versity of Toronto, York niversity, niversity of • The Office of Nursing Policy at Health Canada has
Ottawa, Laurentian niversity, and Lakehead a mandate to improve the health of Canadians.
One of the priorities is to focus on strengthening
niversity (Web site: http://www.otc-hsr.ca). nursing sciences by describing issues and
Since the 1990s, some of the original nursing opportunities, increasing the number of researchers
research pathfinders and many new ones have and identifying actions, and identifying directions
to move the nursing science agenda forward
conducted research on a wide range of clinical
(Office of Nursing Policy, 2009).
topics, describing phenomena and testing inter- • The Canadian Consortium for Nursing Research
ventions. The existence of the many nursing jour- and Innovation (CNRI) was formed under the
nals shows the growth in the variety, quality, and leadership of the Office of Nursing Policy to gather
stakeholders from a variety of nursing
depth of research available for potential use in organizations to provide leadership and direction
practice. In a report funded by the CHSRF, an to advance nursing science in Canada.
analysis of the CIHR database revealed a wide *List of organizations from the Office of Nursing Policy (2006).
diversity of nursing research topics in Canada. Of
 CHAPTER 1 The Role of Research in Nursing 15

TABLE 1-1
RESEARCH THEMES IN CANADA
HEALTH ISSUES HEALTH SERVICE ORGANIZATIONS HEALTH PROMOTION

Chronic illness Delivery and structure of health care Tobacco use

Reproductive health Reducing disparities and improving access AIDS/HIV
Pain Hospital restructuring Abuse
End-of-life/palliative care Models of health care delivery Obesity
AIDS, acquired immune deficiency syndrome; HIV, human immunodeficiency virus.

the studies, 67 were conducted in the areas of FUTURE DIRECTIONS

health issues, health service organizations, and
health promotion. Not surprisingly, the research In the twenty-first century, the continuing expan-
on health issues was dominant the majority of sion of nursing research provides numerous
studies focused on (1) chronic illness (2) repro- opportunities for nurses to study important
ductive health (3) pain and (4) end-of-life/ research questions, promote health care, and ame-
palliative care (Table 1-1 Jeans Associates, liorate the side effects of illness and the conse-
2008). Nurses have made significant research quences of treatment while also optimizing the
contributions to health care. Naturally, nurses are health outcomes of patients and their families
far from finished with their quest for knowledge (Jennings McClure, 2004). Villeneuve and
and research, but the nursing leaders of the twen- MacDonald (2006), in To ard : Visions
tieth century have—by example—paved the for Nursing (p. 99), included the following
way for those who will emerge in the new prediction:
millennium.
. . . nurse researchers in 2020 [will] conduct studies
Nursing research capacity has expanded
that place less emphasis on nurses and nursing
greatly in Canada, as evidenced by increased processes than was the case 20 years [earlier],
numbers of doctoral programs and graduates, focusing instead on health, the needs of clients and
increased funding, and the development of communities, and providing sound evidence to
research teams and centres. However, the profes- guide policy and practice. Research conducted by
sion still faces many challenges. Despite the nurses [will be] interdisciplinary. Research to determine
more cost-effective ways to provide safe, high-quality
increased number of doctoral programs and health and illness services [will be] led by nurse
graduate students, research-prepared faculty to researchers.
supervise the research of the graduate students
are in short supply. In addition, the databases Major shifts in the delivery of health care
that researchers depend on for funding informa- include the following:
tion are not coordinated, and many do not include An emphasis on community-based care
classifications specifically for nursing (Jeans An emphasis on reducing disparities in
Associates, 2008). Moreover, the uptake of health care
research in practice has resulted in the prolifera- A focus on health promotion and risk
tion of best practice guidelines however, there reduction
is a dearth of documentation assessing the An increase in severity of illness in inpa-
impact of nursing research on health care ser- tient settings
vices and health policy (Jeans Associates, An increased incidence of chronic illness
2008, p. 21). An expanding population of older people
16 PART ONE Research Overview

An emphasis on provider accountability increasing emphasis on health services research

through a focus on quality and cost promotes scientific inquiry, which in turn leads to
outcomes a better understanding of health care delivery
The use of technology to serve human needs and issues that cross disciplines, such as resource
In accordance with these trends, nurse research- utilization and policymaking (Aiken, Clarke,
ers are beginning to focus on the development of Cheung, Sloane, Silber, 2003 Buerhaus,
quantitative and qualitative research programs Staiger, Auerbach, 2000).
and clinically based outcome studies. Strategies The work of Dr. Nancy Edwards, the director
that enhance nurses’ focus on outcomes manage- of the Community Health Research nit, provides
ment through evidence-informed quality improve- an excellent example of depth in nursing research.
ment activities and the use of research findings Working in partnership with Public Health and
for effective clinical decision making also are Long-Term Care–City of Ottawa, the niversity
being refined and identified as priorities (see of Ottawa, and a large team of co-researchers, Dr.
Chapter 20). Evidence-informed practice guide- Edwards is enhancing the scientific and evidence-
lines, standards, protocols, and critical pathways informed practice in public health. One of her
are becoming benchmarks for cost-effective, areas of study is the incidence, causes, and pre-
high-quality clinical practice. Also, nurse vention of falls among older adults. For example,
researchers and nurse leaders will become increas- in one study, investigators collected qualitative
ingly visible at the national level, functioning in data on older adults’ beliefs about falls and pre-
policymaking roles, representing the field of vention behaviour the key findings guided further
nursing on expert panels, and lobbying for more study on the use of fall-prevention devices, such
funding dollars. The extent of growth in nursing as grab bars in bathrooms (Community Health
research is evidenced by the establishment of the Research nit, 2005). In a more recent research
Nursing Research Fund, which will provide study, Dr. Edwards and colleagues (2006) focused
financial support for research opportunities iden- on the development and testing of tools to assess
tified by the community of nurse researchers. use of physical restraint.
Dr. Edwards’s work illustrates the value of
Promoting Depth in Nursing Research building replication studies into research pro-
Depth in nursing science becomes evident when grams. Stone and colleagues (2002) proposed that
research is replicated. Research programs that the adoption of research findings in practice, with
include a series of studies in a similar area, each their potential risks and benefits (including the
of which builds on a prior investigation, promote cost of implementation), should be based on a
depth in nursing science (Whittemore Grey, series of replicated studies that provide a body of
2002). Moreover, to maximize use of resources evidence, thereby increasing the degree to which
and to prevent duplication, researchers must the findings can be applied and generalized. When
develop intradisciplinary and interdisciplinary appropriate, a greater focus on generali ability is
networks in similar areas of study (Larson, 2003). important if the evolving science is to be consid-
Researchers from a variety of professions can ered reliable and usable in health care settings and
come together to delineate common and unique in health care policies. As such, future replication
aspects of patient care (i.e., medicine, nursing, studies will yield more credible results and will
and respiratory therapy). Cluster studies, multiple- play a crucial role in developing depth in nursing
site investigations, and programs of research science (Fahs, Stewart, Kalman, 2003).
facilitate the accumulation of evidence that sup- Dr. Edwards’s work also highlights why
ports or negates a theory. Furthermore, the research training is likely to become an
 CHAPTER 1 The Role of Research in Nursing 17

increasingly essential component of a research usually by a single researcher conducting research

career plan. A larger cadre of nurse researchers in his or her home country and then travelling to
who begin their research careers at a young age the other international sites. Because of the finan-
is important for the development of research pro- cial and logistical limitations to this method,
grams like Dr. Edwards’s. The goal is to increase however, the number of international collabora-
the longevity of research careers, enhance the tive research projects has increased. Nurse
discipline’s science development, promote men- researchers participating in collaborative interna-
toring opportunities, prepare the next generation tional research projects are well positioned to
of researchers, and provide leadership in health play a large role in improving health care globally
care for interdisciplinary health care debates. (Chiang-Hanisko, Ratchneewan, Ludwick,
Nurse researchers should also be committed to Martsolf, 2006).
developing publicly and privately funded research International organizations committed to the
programs and to subscribing to a life of periodic goal of health care for all help create natural
education and retraining supported by awards, research partnerships. For example, the World
grants, and fellowships. Research programs facil- Health Organization (WHO, n.d.) has established
itate growth in the depth and breadth of research a series of collaboration centres in order to
expertise and recognize that some researchers advance health care for the global community.
must be retrained as they develop or shift the One such centre works toward maximizing the
emphasis of their research, seek to broaden their contribution of nursing and midwifery and pro-
scientific background, and acquire new research vides relevant research and clinical training to
capabilities. nurses worldwide.

An International Perspective Research Priorities

The continuing development of a national and As the number of nurses with doctoral degrees
international research environment is essential increased, and as nursing chairs were created and
to the nursing profession’s mission to establish funding was expanded, a distinct set of Canadian
a global research community (Chang, 2000 nursing research priorities emerged (Box 1-3).
Hegyvary, 2004). The opportunities for cross- Funding agencies often determine research pri-
cultural and cross-national studies on subjects of orities on the basis of their particular needs and
common interest are consistent with the priority for interests. In 2007, the CHSRF’s priority research
global research (Hegyvary, 2004 Hinshaw, 2000). themes included (1) managing the delivery of
Because of nursing’s emphasis on the cultural high-quality services in the health care work-
aspects of care and the in uence of such factors place (2) managing the safe delivery of high-
on practice, international research is likely to quality services in the health care workplace (3)
increase. Access to multiple populations as a providing primary health care and (4) providing
function of globalization allows the testing of nursing leadership, organization, and policy.
nursing science from various perspectives. Inter- These themes were established after consultation
action with colleagues from other countries with health sector administrators, policymakers,
provides a rich context for the generation and and researchers across Canada. For example,
dissemination of research (Dickenson-Hazard, under the rubric of primary health care, policy-
2004 Ward, 2003). makers were particularly interested in research
International research projects are often dedicated to exploring which models of primary
focused on comparative research in which a phe- health care work best in specific contexts (CHSRF,
nomenon is studied in more than one country, 2007).
18 PART ONE Research Overview

BOX 1-3 maternal–neonatal mortality rate. Individuals of

CANADIAN NURSING RESEARCH PRIORITIES all ages who have sustained life-threatening ill-
nesses will live with the help of new life-
PRIORITY 1: NURSING PRACTICE
sustaining technology that will in turn create new
• Context (including determinants of health, health
reform, and ethical issues) demands for self-care and family support. Cancer,
• Populations (vulnerable groups and specific clinical heart disease, arthritis, asthma, chronic pulmo-
populations, particularly First Nations, Métis, and nary disease, diabetes, and Alzheimer’s disease,
Inuit)
• Interventions (wide range, from health promotion prevalent during middle age and later life, will be
to comfort measures) responsible for expenditures of large proportions
of the available health care resources. Human
PRIORITY 2: OUTCOMES
• Development of valid measures for multiple immunodeficiency virus/acquired immune defi-
dimensions ciency syndrome (HIV/AIDS), a chronic illness
• Links with clinical judgement that affects men, women, and children, will con-
PRIORITY 3: ENHANCED LINKS BETWEEN RESEARCH tinue to have a significant effect on health care
AND PRACTICE delivery.
• Development of a body of nursing knowledge Another vulnerable population, those with
Adapted from Potter, P., Perry, A. G., Ross-Kerr, J., & Wood, M.
mental health illness, will be served by a better
(Eds.), (2010). Canadian fundamentals of nursing (4th ed., p. 80). understanding of mental disorders, which will
Toronto: Elsevier Canada.
emerge as a result of advancements in psychobio-
logical knowledge and research initiatives. Mental
The CIHR’s nursing grants funded 11.6 health illnesses will continue to be a major public
million for 105 research projects in 2005. health issue depression has been cited as the
Although the availability of funding contributes number one mental illness worldwide (WHO,
to interest in nursing research, only 15 of 2001). Alcohol and drug abuse will continue to
nurses’ applications to the CIHR were successful be responsible for significant health care expenses.
in procuring funding. Finally, as national initiatives are launched to
In 2006, 50 Canadian nursing leaders identi- improve end-of-life care, the shortage of nurses—
fied additional priorities for nurses during a as clinicians and investigators—will become a
Nursing Leadership, Organization, and Policy major problem worldwide (WHO, 2001). Inves-
Network Day sponsored by the CHSRF. Several tigations that address quality-of-care outcomes
research themes were identified as requiring related to nursing are a top priority for psychiatric
further research, including human resources nurses in the twenty-first century (President’s
planning and management, nursing roles and New Freedom Commission on Mental Health,
scopes of practice, education, worklife, staff- 2003), as are nursing contributions to end-of-life
ing, leadership, and use of research through research (Matzo Sherman, 2001).
knowledge transfer and exchange (CHSRF, ndoubtedly, one of the most exciting research
2006, p. 7). opportunities for nurses in the future relates to
Reducing health disparities in underserviced genetics and the human genome, where study
communities and vulnerable populations is prospects range from basic biological topics to
another major topic that will shape the focus of clinical decision making and behavioural inter-
future nursing- and interdisciplinary-related ventions (J. K. Williams, Tripp-Reimer, Schutte,
research agendas. For example, the health Barnette, 2004).
concerns of mothers and infants will continue to In terms of the priority given to clinical
spur research that deals effectively with the research issues, the funding of investigations has
 CHAPTER 1 The Role of Research in Nursing 19

increasingly emphasized populations of interest. development of research priorities for interna-

For example, the historical exclusion of women tional nursing organizations such the Interna-
as participants in clinical research is well docu- tional Council of Nurses. The nursing research
mented minority women in particular were even priorities listed by this council include health,
more likely to be excluded from research studies. illness and care delivery services that emphasise
As a result, research data on the health of minor- quality and cost-effective care, community-based
ity women traditionally were extremely scarce. care, nursing workforce and health care reform
Nurse researchers such as Josephine Etowa in (International Council of Nurses, 2007, p. 2).
Canada are addressing this gap in research. Other types of research investigations (e.g.,
Along with national nursing research initia- those involving historical, feminist, or case study
tives, specific nursing interest groups frequently methods) embody the rich diversity of nursing
establish their own research priorities specific to research methods. The nursing profession must
their specialty. Some associations provide funds continue to value and promote creativity and
to support research initiatives. The Canadian diversity in research endeavours at all educational
Association of Critical Care Nurses, for example, levels as a way of empowering nurses for the
established a small research grant to support spe- future. As opportunities are recognized and gaps
cific research activities. The grant will be awarded in science are observed, nurses will conduct, cri-
to a researcher whose work is related directly to tique, and use nursing research in ways that pub-
the clinical practice of critical care nurses licly demonstrate how nursing care makes a
(Canadian Association of Critical Care Nurses, difference in patients’ lives.
2007). Nurse researchers will have an increasingly
International research priorities are in uenced strong voice in shaping public policy relating to
by social, political, and economic factors. Typi- health care. Shaver (2004) stated that disciplines
cally, international organizations such as the such as nursing—because of its focus on treat-
nited Nations and the WHO prioritize health ment of chronic illness, health promotion, inde-
research that will affect the most marginalized of pendence in health, and care of the acutely ill, all
the world’s population. As such, the eight goals of which are heavily emphasized values for the
identified by the nited Nations (2000) focus on future—will be central to the shaping of health
poverty, vulnerable populations (i.e., women and care policy in the future. Research evidence that
children), and the specific health issues. The supports or refutes the merit of health care needs
goals are as follows: and programs focusing on these issues will be
Eradicate extreme poverty and hunger timely and relevant. Thus, nursing and its science
Achieve universal primary education base are strategically placed to shape health
Promote gender equality and empower policy decisions (Fitzpatrick, 2004).
women Communication of nursing research has also
Reduce child mortality become increasingly important. Research find-
Improve maternal health ings continue to be disseminated in professional
Combat HIV/AIDS, malaria, and other arenas (e.g., international, national, regional, and
diseases local electronic and print publications and confer-
Ensure environmental sustainability ences), as well as in consultations and staff devel-
Develop a global partnership for develop- opment programs implemented on site through
ment Webinars and Web sites. Dissemination of
The global research priorities established by research findings in the public sector has also
the nited Nations and the WHO in uence the gained importance.
20 PART ONE Research Overview

Increasingly, nurse researchers are asked to

testify at governmental hearings and to serve on KEY POINTS
commissions and task forces related to health
care. Nurses are quoted in the media when health • Nursing research expands the unique body of
scientific knowledge that forms the foundation of
care topics are addressed, and their visibility has evidence-informed nursing practice. Research is the
expanded significantly. For example, Sue component that links education, theory, and
Johanson, RN, host of the Sunday Night Se practice.
• Nurses become knowledgeable consumers of
Sho , brought more than 35 years of experience research through educational processes and
in sex education to the public discourse on sex. In practical experience. As consumers of research,
addition to addressing all aspects of sex in a non- nurses must have a basic understanding of the
judgemental manner on her show, Ms. Johanson research process and must demonstrate critical
appraisal skills to evaluate the strengths and
has written several books and wrote a weekly col- weaknesses of research before applying the
umn on health for the Toronto Star Although re- research to clinical practice.
tired in 2008, Ms. Johanson still lectures to a • Nursing research blossomed in the second half of
the twentieth century: Graduate programs in
wide range of audiences. Dissemination of re- nursing expanded, research journals began to be
search through the public media provides excel- published, and funding for graduate education and
lent exposure of health-related information to nursing research increased dramatically.
• All nurses, whether they possess baccalaureate,
thousands of potential viewers, listeners, and master’s, or doctoral degrees, have a responsibility
readers. to participate in the research process.
Nurses have a research heritage to be proud of. • The role of the baccalaureate graduate is to be a
They also have a challenging and exciting future knowledgeable consumer of research. Nurses with
master’s and doctoral degrees are obliged to be
ahead of them. Both researchers and consumers researchers and sophisticated consumers of
of research need to engage in a united effort to research studies.
gather and assess research findings that make a • A collaborative research relationship within the
nursing profession will extend and refine the
difference in the care that is provided and in the scientific body of knowledge that provides the
lives that are touched by their commitment to grounding for theory-informed practice.
evidence-informed nursing practice. • The future of nursing research will focus on the
extension of scientific knowledge. Collaborative
research relationships between education and
practice will multiply. Programs of research studies
and replication of studies will become increasingly
CRITICAL THINKING CHALLENGES valuable.
• Research studies will emphasize clinical issues,
■ How can you access evidence-informed practice problems, and outcomes. Priority will be given to
research studies that focus on health promotion,
information when practising in the field? care for the health needs of vulnerable groups, and
■ What assumption underlies the recommendation the development of cost-effective health care
that the role of the nursing graduate in the systems.
research process be primarily that of a • Both consumers of research and nurse researchers
knowledgeable consumer? will engage in a collaborative effort to further the
growth of nursing research and accomplish the
■ What effects will evidence-informed patient profession’s research objectives.
outcome studies have on the practice of nursing?
Discuss how research will contribute to the
■
REFERENCES
development of intradisciplinary and
interdisciplinary networks. Aiken, L., Clarke, S. P., Cheung, R. B., Sloane, D. M.,
■ How are research priorities established by Silber, J. H. (2003). Education levels of hospital
organizations and professional bodies? nurses and patient mortality. ournal of American
Medical Association, , 1617-1623.
 CHAPTER 1 The Role of Research in Nursing 21

Buerhaus, P., Staiger, D., Auerbach, D. (2000). Committee for the Study of Nursing Education. (1923).
Implications of a rapidly aging nursing workforce. Nursing and nursing education in the nited States
ournal of the American Medical Association, Report of the committee and report of a survey by
(22), 2948-2954. osephine oldmar . New York: Macmillan.
Canadian Association of Critical Care Nurses. (2007). Community Health Research nit. (2005). Research
CACCN research grant. Retrieved from http:// projects funded in . Retrieved from http://
findarticles.com/p/articles/mi m5PTW/is 3 18/ai n31 aix1.uottawa.ca/ nedwards/chru/english/research2005.
153665/ tag content col1 html
Canadian Health Services Research Foundation. (2006). Dickenson-Hazard, N. (2004). Global health issues and
oo ing for ard, or ing together: Priorities for challenges. ournal of Nursing Scholarship, (1),
nursing leadership in Canada. Retrieved from 6-10.
http://www.chsrf.ca/Migrated/PDF/NLOP e.pdf Edwards, N., Danseco, E., Heslin, K., Ploeg, J., Santos,
Canadian Health Services Research Foundation. (2007). J., Stansfield, M., Davies, B. (2006). Development
Primary research themes. Retrieved from http:// and testing of tools to assess physical restraint use.
www.chsrf.ca/research themes/ph e.php orldvie s Evidence Based Nursing, (2), 73-85.
Canadian Health Services Foundation (2008). Fahs, P. S., Stewart, L. L., Kalman, M. (2003). A call
Annual report on the nursing research fund. for replication. ournal of Nursing Scholarship,
Retrieved from http://www.chsrf.ca/Programs/ (1), 67-72.
PastPrograms/NursingResearchFund.aspx Fitzpatrick, J. J. (2004). Translating clinical research
Canadian Nurses Association. (2002). Position state- into research policy. Applied Nursing Research,
ment: Evidence-based decision-ma ing and nursing (2), 71.
practice. Retrieved from http://www.cna-aicc.ca Forchuk, C., Martin, M. L., Chan, Y. L., Jensen, E.
Canadian Nurses Association. (2008). Code of ethics for (2005). Therapeutic relationships: From psychiatric
registered nurses. Retrieved from http://www. hospital to community. ournal of Psychiatric and
cna-nurses.ca/CNA/documents/pdf/publications/ Mental ealth Nursing, , 556-564.
Code of Ethics 2008 e.pdf Ford-Gilboe, M.,Wuest, J.,Varcoe, C., Merrit-Gray,
Canadian Nurses Association. (2009). Strengthening M. (2006). Developing an evidence-based health
Canada s health system by advancing health through advocacy intervention for women who have left an
nursing science. Retrieved from www.cna-nurses.ca/ abusive partner. Canadian ournal of Nursing
CNA/documents/pdf/MP leave behind e.pdf Research, , 147-167.
Canadian Nurses Association. (2010a). Canadian Hegyvary, S. T. (2004). Working paper on grand
Registered Nurse E amination: Competencies. challenges in improving global health. ournal of
Retrieved from http://www.cna-aiic.ca/CNA/nursing/ Nursing Scholarship, , 96-101.
rnexam/competencies/default e.aspx Hinshaw, A. S. (2000). Nursing knowledge for the 21st
Canadian Nurses Association. (2010b). Evidence- century: Opportunities and challenges. ournal of
informed decision-ma ing and nursing practice. Nursing Scholarship, , 117-123.
Retrieved from http://www.cna-aiic.ca/CNA/ International Council of Nurses. (2007). Nursing
documents/pdf/publications/PS113 Evidence research. Retrieved from http://www.icn.ch/images/
informed 2010 e.pdf stories/documents/publications/position statements/
Carper, B. A. (1978). Fundamental patterns of nursing. B05 Nsg Research.pdf
Advanced Nursing Science, (1), 13-24. Jeans, M. E. (2005). Shared leadership for nursing
Chang, W. Y. (2000). Priority setting for nursing research. Nursing eadership Toronto, ntario ,
research. estern ournal of Nursing Research, , (1), 20-23.
119-121. Jeans, M. E., Associates. (2008). Nursing research in
Chiang-Hanisko, L., Ratchneewan R., Ludwick, R., Canada: A status report. Retrieved from http://
Martsolf, D. (2006). International collaborations www.canr.ca/documents/NursingResCapFinalReport
in nursing research: Priorities, challenges and ENG Final.pdf
rewards. ournal of Research in Nursing, (4), Jennings, B. M., McClure, M. L. (2004). Strategies to
307-322. advance health quality. Nursing utloo , , 17-22.
College of Nurses of Ontario. (2009). National Larson, E. (2003). Minimizing disincentives for
competencies in the conte t of entry level registered collaborative research. Nursing utloo , ,
nurse practice. Toronto: Author. 267-271.
22 PART ONE Research Overview

Matzo, M., Sherman, D.W. (Eds.), (2001). Palliative Tarlier, D. (2005). Mediating the meaning of evidence
care nursing: uality care to the end of life. New through epistemological diversity. Nursing In uiry,
York: Springer. , 126-134.
McGinty, J., Anderson, G. (2008). Predictors nited Nations. (2000). Millennium development goals.
of physician compliance with American Heart Retrieved from http://www.un.org/millenniumgoals/
Association Guidelines for acute myocardial bkgd.shtml
infarction. Critical Care Nurse uarterly, (2), Villeneuve, M., MacDonald, J. (2006). To ard :
161-172. Visions for nursing. Ottawa: Canadian Nurses
Nightingale, F. (1863). Notes on hospitals. London: Association.
Longman Group. Ward, L. S. (2003). Race as a cross-variable in research.
Office of Nursing Policy. (2006). Nursing issues: Nursing utloo , (3), 120-125.
Research. Retrieved from http://www.hc-sc.gc.ca/ Weir, G. M. (1932). Survey of nursing education in
hcs-sss/pubs/nurs-infirm/onp-bpsi-fs-if/2006-res- Canada. Toronto: Macmillan. Note that this
eng.php publication is often referred to as the Weir report or
Office of Nursing Policy. (2009). f ce of Nursing the Weir survey.
Policy overvie . Retrieved from http://www. Whittemore, R., Grey, M. (2002). The systematic
hc-sc.gc.ca/ahc-asc/branch-dirgen/spb-dgps/onp-bpsi/ development of nursing interventions. ournal of
overview-apercu-eng.php Nursing Scholarships, (2), 115-120.
Paterson, B. (2002). An answer for Sandra. Canadian Williams, D. O. (2004). Treatment delayed is treatment
Nurse, (4), 14. denied. Circulation, , 1806-1808.
Potter, P., Perry, A. G., Ross-Kerr, J., Wood, M. Williams, J. K., Tripp-Reimer, T., Schutte, D.,
(Eds.), (2010). Canadian fundamentals of nursing Barnette, J. (2004). Advancing genetic nursing
(4th ed.). Toronto: Elsevier Canada. knowledge. Nursing utloo , (3), 73-79.
President’s New Freedom Commission on Mental World Health Organization. (2001). The orld health
Health. (2003). Achieving the promise: Transforming report mental health: Ne understanding, ne
mental health care in America: inal report DHHS hope. Geneva, Switzerland: Author.
Pub. No. SMA-03-3832 . Rockville, MD: .S. World Health Organization. (n.d.). Net or s of
Department of Health and Human Services. collaborating centres. Retrieved from http://
Sackett, D. L., Rosenberg, W., Gray, J. A., Haynes, R., www.who.int/collaboratingcentres/networks/
Richardson, W. (1996). Editorial: Evidence based networksdetails/en/index1.html
medicine: What it is and what it isn’t. British Medical
ournal, (7023), 71.
Shaver, J. (2004). Improving the health of
FOR FURTHER STUDY
communities: The position. Nursing utloo , , Go to Evolve at http://evolve.elsevier.com/
116-117. Canada/LoBiondo/Research for Audio Glossary, how-to
Stone, P. W., Curran, C. R., Bakken, S. (2002). instructions for Writing Proposals for Funding, and
Economic evidence for evidence-based practice. additional research articles for practice in reviewing
ournal of Nursing Scholarship, (3), 277-282. and critiquing.
C H A PTER 2

Theoretical Framework
Joan Samuels-Dennis | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Define key concepts in the philosophy of science.
• Identify and differentiate between theoretical/empirical, aesthetic, personal, sociopolitical, and
ethical ways of knowing.
• Identify assumptions underlying the post-positivist, critical, and interpretive/constructivist views of
research.
• Compare inductive and deductive reasoning.
• Differentiate between conceptual and theoretical frameworks.
• Describe how a framework guides research.
• Differentiate between conceptual and operational definitions.
• Describe the relationships among theory, research, and practice.
• Discuss levels of abstraction related to frameworks guiding research.
• Describe the points of critical appraisal used to evaluate the appropriateness, cohesiveness, and
consistency of a framework guiding research.

KEY TERMS
aim of inquiry epistemology post-positivism
concept hermeneutics post-positivist paradigm
conceptual definition hypothesis qualitative research
conceptual framework inductive reasoning quantitative research
constructivism methodology text
constructivist paradigm model theoretical framework
context ontology theory
critical social theory operational definition values
critical social thought paradigm variables
deductive reasoning philosophical beliefs worldview

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

23
24 PART ONE Research Overview

phenomenon of interest. no ledge generation

THE NATURE OF KNOWLEDGE next occurs, with the conduct of research that
ON A DAILY BASIS, NURSES DEVISE clinical questions provides answers to well-thought-out research
that, if answered, improve upon the care they questions. This no ledge is then distributed
provide to individuals, families, and communi- through journal articles, textbooks, and public
ties. A medical/surgical nurse may ask, How do presentations to nurses. Next, the no ledge is
pediatric and adult patients experience surgical adopted, as nurses alter their practice on the basis
pain, and how, in turn, can we enhance pain man- of published information or as health care orga-
agement for both groups A mental health nurse nizations develop policies and protocols that are
may ask, What is the relationship between the informed by newly generated knowledge. Finally,
use of specific psychotropic medication and heart no ledge is revie ed and revised as new health
disease among schizophrenic patients A public issues arise, advances in clinical practice occur,
health nurse might ask, What factors contribute or knowledge becomes outdated. In this chapter,
most significantly to the health of homeless we focus specifically on theoretical frameworks
youths living in a specific neighbourhood in and how they guide and inform no ledge
Ontario, Canada generation
Each question requires that clinicians/nurse Figure 2-2 outlines the various ways by which
researchers engage in a knowledge development nurses inform their practice. These include
process (Figure 2-1). The process begins with the theoretical/empirical, personal, experiential,
identification of no ledge gaps: the absence of ethical, aesthetic, and sociopolitical ways of
theoretical or scientific knowledge relevant to the knowing (Chinn Kramer, 2010 ander, 2007).

Knowledge Gap
• Nurses ask questions that require
answers from experts in the field.
• Absence of theoretical/empirical
knowledge.

Knowledge Review and Knowledge Generation
Revision
• Research questions are devised
• New health issues lead to the asking about a phenomenon.
of new questions. • Qualitative and quantitative
• Old knowledge is revised or excluded. methods are used to answer the
• New questions prompt the need for questions.
new research.

Knowledge Distribution
Knowledge Adoption
• Knowledge is shared with profession
• New knowledge is used to alter
through formal (presentation, journal
practice.
publications, reports) and informal
• New knowledge is used to
(media, Internet, social networks)
develop policies and protocols.
reporting methods.

FIGURE 2-1 Knowledge development process.

 • Knowledge that comes from the inner
experiences and maturation of the nurse
• Encompasses becoming a whole, aware,
genuine, and authentic self
• Continual life process that occurs with
• Scientific knowledge
deep reflection
• Development and testing of
• Allows for true interpersonal
theories/theoretical models Theoretical/ Personal relationships
• Observation and measurement of Empirical Knowledge
case-specific issues and phenomena Knowledge
• Quantitative and qualitative research
• Knowledge that comes from
repeated exposure to
• Contextual knowledge that situations that lead to the
moves beyond nurse– refining of earlier ideas and
patient relationship and the thoughts
profession Experiential • Involves movement along a
• Understanding of culture, Sociopolitical Knowledge continuum of novice →
society, and politics Knowledge competent → expert
• Awareness of how society practitioner
is organized and its • Expert practitioner is fully
implications for health engaged and attuned to
each situation and responds
on the basis of past learning
• The art of nursing
• The expressive, intuitive, and Ethical
Aesthetic • Moral component of nursing knowledge
creative aspect of nursing Knowledge
Knowledge • Encompasses the moral principles, codes,
• Practice that incorporates poetry, art,
music, and stories, allowing a deep and theories of ethical conduct
understanding of the human • The moment-to-moment judgement of
experience what is “right” and “responsible”
• Involves confrontation and resolution of
ethical conflicts

FIGURE 2-2 Nursing knowledge.

Based on information from Benner, P. (2004). Using the Dreyfus model of skill acquisition to describe and interpret skill acquisition and clinical judgment in
nursing practice and education. Bulletin of Science, Technology & Society, 24(3), 188; Benner, P., Tanner, X., & Chesla, C. (1992). From beginner to expert: Gaining
a differentiated clinical world in critical care nursing. Advances in Nursing Science, 14(3), 13; Chinn, P. L., & Kramer, M. K. (2010). Nursing’s fundamental paterns
CHAPTER 2 Theoretical Framework

of knowing. In P. Chinn & M. Kramer (Eds.), Integrated knowledge development in nursing (8th ed., pp. 1–17). St. Louis: Mosby; and Zander, P. E. (2007). Ways
of knowing in nursing: The historical evolution of a concept. Journal of Theory Construction and Testing, 11(1), 7–11.
25
26 PART ONE Research Overview

Theoretical/empirical knowledge is most com- that Erci and colleagues could observe these out-
monly referred to as scienti c no ledge In com- comes (e.g., blood pressure measured every 3
parison with all the other ways of knowing months) among patients who did and did not
outlined in Figure 2-2, theoretical/empirical receive care guided by Watson’s theory. These
knowledge has gained prominence in nursing and observations provided support for this hypothe-
currently serves as the guide for evidence- sis this support is an example of empirical
informed practice. Theoretical and empirical no ing
knowledge really cannot be separated however,
theoretical knowing is concerned with develop- PHILOSOPHIES OF RESEARCH
ing or testing theories or ideas that nurse research- Thus far, we have used a number of terms theo-
ers have about how the world operate. Theoretical retical, empirical, hypothesis that may be new to
knowing is informed by empirical knowing, you. Every specialty has characteristic terminol-
which involves observations of reality. Observa- ogy for communicating important features of the
tions may include the following: work of that specialty. Learning new terminology
1. Speaking with people about their life expe- is part of what nursing students do when they
riences (e.g., living with Alzheimer’s learn research methods and skills. Each research
disease) and using their responses to spe- method and all philosophies of science have spe-
cific and general questions to understand cialized language that nursing students will
the phenomenon encounter in the literature. Thus, to help you com-
2. Observing social or cultural interactions prehend the research you will read, it is important
(e.g., homeless individuals interacting with to clarify a few terms.
service providers) as they naturally occur, All research is based on hiloso hical belie s
interpreting what the interactions might about the world these beliefs are the motivating
mean for both parties, and using those inter- values, concepts, principles, and the nature of
pretations to develop theories about health human knowledge of an individual, group, or
service delivery for that population culture, and they are the basis of a orldvie or
3. Delivering an intervention (e.g., a school aradig Paradigm is from the Greek word
health program for obese children) and paradeigma, meaning pattern. Paradigms rep-
assessing changes in health care–related resent a sets of beliefs and practices, shared by
behaviours (e.g., type of foods consumed, communities of researchers that guide the
amount of daily exercise) after the delivery knowledge development process (Weaver
of the intervention Olson, 2006). Therefore, knowing and compre-
4. sing surveys or a questionnaire to ask a hending these beliefs and practices is important
large group of men and women questions in understanding and using research findings.
about experiences of violence and their These beliefs are not right or wrong rather, they
current symptom levels with regard to pain, represent different views of the world, and their
digestive problems, or depression. use the goals of the research.
Taking an example of published work, Erci Nursing research is guided by three research
and colleagues (2003) hypothesized that nursing paradigms: post-positivism, constructivism, and
care guided by Jean Watson’s theory of human social critical theory. These three paradigms are
caring would improve patient outcomes such as compared in Table 2-1 however, first you need
quality of life and blood pressure this hypothesis to understand the philosophical language used in
is an example of theoretical no ing However, the table. Ontology (from the Greek word onto,
it was only through developing an experiment meaning to be ) is the science or study of being
 TABLE 2-1
BASIC BELIEFS OF RESEARCH PARADIGMS
ITEM/QUESTIONS POST-POSITIVISM CRITICAL THEORY CONSTRUCTIVISM

ONTOLOGY
What can be said to exist? A material world exists Reality is constructed by those with Reality is constructed by individual
Into what categories can we Not all things can be understood, sensed, the most power at particular perception
sort existing things? or placed into a cause-and-effect points in history There exists no absolute truth or validity
relationship Reality is plastic and at all time Truth is relative and subjective and based
The senses provide us with an imperfect imperfectly understood on perception or some particular
understanding of the external/ Over time, reality is shaped by frame of reference
material world numerous social, political,
economical, cultural forces
Imperfectly shaped stories become
accepted reality

EPISTEMOLOGY
What is knowledge? Researchers are naturally biased Research is a transaction that occurs Research is a transaction that occurs
How is knowledge acquired? Objectivity (controlled bias) is the between the researcher and between the researcher and research
How do we know what we ultimate goal research participant participant
know? Objectivity encourages triangulation and The perceptions (standpoint) of the The perceptions (standpoint) of the
replication of findings across multiple researcher and the research researcher and the research
perspectives participants naturally influence participants naturally influence
Objectivity encourages intense scrutiny knowledge generation/creation knowledge generation/creation
of research findings from a larger Perceptions (standpoints) are Research emphasizes the meaning
community of scientists and the determined by context, and so ascribed to human experiences
rejection of poorly conducted research contextual awareness and its Context is not emphasized
relationship to the participants Objectivity as outlined by the post-
understanding of reality is the positivist is not a desired goal
focus of the research
Objectivity as outlined by the
post-positivist is not a desired goal

RESEARCHER’S VALUES
How do the researcher’s All attempts are made to exclude Researcher bias is recognized as Researcher bias is recognized as
values influence the researcher bias potentially influential potentially influential
knowledge development Influence denied Influence is limited with reflection Influence is limited with reflection and
CHAPTER 2 Theoretical Framework

process? and bracketing bracketing

Continued
27
 TABLE 2-1
28

BASIC BELIEFS OF RESEARCH PARADIGMS—cont’d

ITEM/QUESTIONS POST-POSITIVISM CRITICAL THEORY CONSTRUCTIVISM
METHODOLOGY
Within a particular discipline, Inquiry includes experimental, Inquiry requires dialogue between Inquiry requires dialogue between the
what principles, rules, and nonexperimental, and qualitative the investigator and research investigator and research participant
procedures guide the methods and is viewed as a series of participant Focus is on interpretation of written texts,
process through which logically related steps Dialogue is transformative or art, pictures, and videos
knowledge is acquired? Research questions/hypotheses are consciousness raising Interpretation brings to the forefront the
proposed and subjected to empirical Dialogue brings to the forefront the varying ways in which people
testing for the purpose of being proved historical context behind construct their understanding of their
incorrect rather than correct experiences of suffering, conflict, social world and how their
Research is characterized by careful and collective struggles interpretation shifts as they interact
PART ONE Research Overview

accounting and control of factors that Dialogue increases participants’ with others
may influence research findings awareness of actions required to
Research entails the use of qualitative insight change
methods to develop hypotheses about
people’s social world and the meaning/
purpose people assign to their actions

AIM OF INQUIRY
What is the goal of research? Explanation, prediction, and control Critique, change and reconstructed Understanding and reconstructed reality
reality, and emancipation

CONTEXT
What biographical, life, Focus is on biographical context and its Focus is on historical, social, and Focus is on life context, including
social, and political factors potential to influence research findings political context significant conditions and demands
may influence the Biographical context includes individual Context refers to the social and that provide greater understanding of
research findings? characteristics such as race, age, political climate in which an the phenomena being studied; focus
geographic location; it also accounts event or process occurred also emphasizes time and place
for past experiences in terms of both Social context highlights how
timing and content structural, economic,
representational, and institutional
factors of the past influence how
people understand an issue
today
Political context highlights how
political dialogue and opinions,
legal directives, and government
policies of the past influence
how people understand an issue.
From Handbook of qualitative research by Denzin, Norman K.; Lincoln, Yvonna S. Copyright 2000 Reproduced with permission of SAGE PUBLICATIONS INC BOOKS in the format Textbook via
Copyright Clearance Center.
 CHAPTER 2 Theoretical Framework 29

or existence and its relationship to nonexistence. understand our world are largely dependent on
Ontology addresses two primary questions: (1) our perception. Truth, as viewed by post-
What can be said to exist and (2) Into what positivists, is awed because truth is never abso-
categories can existing things be sorted E is lute. Truth and our understanding of the world are
te ology (from the Greek word epist m , determined by our life experiences, which in turn
meaning knowledge ) is the branch of philoso- inform how we view the world. Knowledge
phy that deals with what is known to be truth. development, from the perspective of the con-
Epistemology addresses three central questions: structivists, is not valuable if it is used simply to
(1) What is knowledge (2) How do we know prove or disprove theories. Rather, the value of
what we know and (3) What is the scope/ knowledge development lies in the ability to
limitation of knowledge ethodology refers to understand how people perceive their world.
discipline-specific principles, rules, and proce- Knowledge development occurs through obser-
dures that guide the process through which vation or dialogue with people, or both, and as a
knowledge is acquired. The ai o in uiry refers result of paying attention to the language people
to the goals or specific objectives of the research. use to describe life experiences. Constructivists
Conte t refers to the personal, social, and politi- value subjectivity (personal knowing) versus
cal environment in which a phenomenon of inter- objectivity (quantified knowing), inasmuch as the
est (that thing of interest ) occurs. The context aim of this form of research is to create an under-
of research studies can include physical settings, standing of people and their life experiences from
such as the hospital or home, or less concrete their point of view.
environments, such as the context that cultural Critical social thought is a philosophical ori-
understandings and beliefs bring to an experi- entation that suggests that reality and our under-
ence. alues are the personal beliefs of the standing of reality is constructed by people with
researcher. the most power at a particular point in history.
Post ositivis is a philosophical orientation Reality, and our understanding of the world, is
that suggests that a material world exists that is, always changing, and at all times we have an
things can be sensed (i.e., seen, touched, heard, imperfect understanding of our world. Critical
tasted). Furthermore, it is governed by the social thought places a strong emphasis on under-
expressed belief that although not all things can standing health and illness within the context of
be understood or explained, many things can be. history. This perspective supports the understand-
In fact, our world can be observed, events and ing that health and other aspects of reality are
phenomena can be categorized, and we can create shaped by numerous social, political, economical,
theories to explain why some things like illness and cultural factors. Such factors include gender,
and health occur or do not occur. Post-positivism social and economic status, minority versus
emphasizes the proving/disproving of theories for majority status, and even a country’s status as a
the purpose of explaining, predicting, and con- developed versus developing nation. A strong
trolling specific outcomes. Post-positivism values emphasis is placed on understanding how power
objectivity (e.g., observations from a neutral imbalances associated with these factors in u-
rather than a subjective position) and encourages ence health and well-being.
the intense scrutiny of research findings for the As in constructivism, objectivity is not a goal.
purpose of excluding knowledge that was not Rather, understanding people’s experiences from
developed through a rigorous process. their perspective is highly valued. In addition,
Constructivis is a philosophical orientation the goal of knowledge development is to pro-
that suggests reality and the way in which we vide evidence that will support change or the
30 PART ONE Research Overview

transformation of reality. Critical social thought question. An experiment would involve putting
incorporates feminist theory and what has come together two groups of at-risk women and then
to be known as action research. These approaches giving one group tamoxifen and the other group
to research examine how an individual’s or a a placebo (sugar pills) for a period of time (e.g.,
group’s position in society shapes that individu- 2 years). After 2 years, the nurse researcher would
al’s or group’s experiences and causes differential measure how many people in each group devel-
or unequal access to resources, power, autonomy, oped breast cancer. This approach is guided by
and privilege. the belief that reality can be imperfectly under-
stood through observation and measurement. The
aim of the research is to understand exactly how
Helpful Hint
the drug works and under what circumstances. In
All research is based on a paradigm; however, the
paradigm is rarely identified in a research report. addition, the goal is to predict who (which type
of patient) will benefit most from taking a specific
drug such as tamoxifen. The post-positivist
Table 2-1 provides an introduction to post- researcher recognizes that responses to tamoxifen
positivism, critical theory, and constructivism and will be in uenced by the biographical or personal
how each might in uence nursing research today. context (e.g., age, gender, genetic background,
The ost ositivist aradig is the basis of and smoking history), and the researcher will
most quantitative research and, to a smaller attempt to account for the potential effect these
extent, qualitative research. Grounded theory, as in uences might have on the research findings.
originally developed by Barney Glaser and Knowledge from this type of research will be
Anselm Strauss (1967), is very much based on used to promote evidence-informed practice,
the cause-and-effect philosophy of the post- health public policy, and the actions of cancer
positivist paradigm. The constructivist ara care advocates.
dig is the basis of most qualitative research that A researcher with a constructivist orientation
grants importance to her eneutics, which is the may be most interested in answering the question
interpretation of written, oral, and visual com- What is the lived experience of women who are
munication. In critical social theory both quali- being treated for breast cancer ualitative
tative and quantitative research are used to studies based on the constructivist paradigm are
highlight historical and current experiences of guided by the ontological view that multiple real-
suffering, con ict, and collective struggles. It is ities exist. As Olson (2006) stated, Phenomena
important to note that paradigms guide the devel- are studied through the eyes of people in their
opment of the research question and the methods lived situations (p. 461). For example, the
used to answer the questions. Some types of meaning of cancer for a young mother is probably
research are most congruent with the post- different from that for a grandmother. The
positivist paradigm others are most congruent meaning of cancer also may be different in
with the constructivist or social critical Canada than in Japan.
paradigm. Epistemology includes the view that truth
Consider the example of chemotherapy for varies and is subjective. Context is important, and
cancer. A researcher who has a post-positivist ori- description of the experience is vital. When
entation may be most interested in answering the seeking to understand patients’ experiences of a
question How effective is tamoxifen in reducing treatment, nurse researchers would expect that
breast cancer among at risk women This nurse what is important and true for one person may
researcher may use an experiment to address this not be so for another. Some of the differences
 CHAPTER 2 Theoretical Framework 31

may result from context. The experience may Findings from these studies would support the
well vary according to where the patient is treated researcher and the research participants in becom-
and the patient’s characteristics, such as age, ing change agents who advocate for transfor-
gender, and ethnicity. The experience of having mations in service delivery practices/policies that
cancer may be different for a patient whose intentionally or unintentionally support the
mother or father died a painful death from cancer unequal access to cancer treatment therapies.
than for a patient who knew people in whom
cancer was cured. The values of everyone Helpful Hint
involved are acknowledged in qualitative Values are involved in all research. For the post-
research. Again, the finding from this research positivist, it is vital that values not influence the results
of the research. However, for the critical social and con-
can be used to support evidence-informed prac- structivist researcher, values and their potential influ-
tice and patient-centred care. ences on the research results are accepted as a natural
A researcher who shares a critical social orien- part of the research process.
tation may be most interested in answering the
questions Does access to cancer treatment vary Another way of thinking about paradigms and
by racial/ethnic groups and What steps must linking them to research is illustrated in the Criti-
be taken to ensure equal access to cancer treat- cal Thinking Decision Path on p. 32. This algo-
ment such as tamoxifen Such a researcher may rithm demonstrates that beliefs lead to different
use a quantitative approach (e.g., mailing ques- questions, which in turn lead to the selection of
tionnaires to all people who received diagnoses different research approaches. ualitative and
of cancer in the previous 5 years and assessing quantitative research methods are associated with
whether race in uenced access to early treatment) different assumptions that are consistent with
to address the first question and a qualitative each method and are more specific than these
approach such as focus groups (e.g., speaking global worldviews (positivism, critical theory,
with groups of five individuals from various and constructivism). These beliefs and approaches
racial/ethnic groups about equal access) to address lead to different research activities, as illustrated
the second question. This approach is guided by in the decision path.
the ontological belief that reality is documented
by individuals with the most power at particular RESEARCH METHODS: QUALITATIVE
points in history. AND QUANTITATIVE
In addition, social critical researchers believe Research methods are the techniques, procedures,
that reality and people’s experiences are shaped and processes used by researchers to organize a
by numerous social, political, economical, cul- study in order for it to provide answers to the
tural forces. The goal of critical research is to research question. Research methods can be clas-
critique, change, and reconstruct reality (tell a sified into two major categories: qualitative and
different story) and to alleviate the experience of quantitative. A researcher chooses between these
social injustices/inequalities. The research process categories primarily on the basis of the question
creates change in study participants, the researcher, the researcher is asking. If a researcher wishes to
and society. This type of research, for example, test a cause-and-effect relationship, such as how
would attempt to highlight from a historical social support (cause) leads to high blood pres-
perspective how the positions of various group sure (effect), quantitative methods are most
(First Nations, Asians, Blacks) and the prestige appropriate. If, however, a researcher wishes to
granted to those groups in uence their access to discover and understand the meaning of an expe-
the appropriate treatments for breast cancer. rience or process such as death and dying, a
32 PART ONE Research Overview

CRITICAL THINKING DECISION PATH

Selecting a Research Process

If your beliefs are

Researcher Humans are biopsychosocial beings, or Humans are complex beings who attribute
beliefs known by their biological, unique meaning to their life situations.
psychological, and social They are known by their personal
characteristics. expressions.

Truth is objective reality that can be Truth is the subjective expression of reality
experienced with the senses and as perceived by the participant and shared
measured by the researcher. with the researcher. Truth is context laden.

then you will ask questions, such as

What is the difference in blood or

Example
questions pressure and heart rate between What is the structure of the lived experience
adolescents who are angry and those of anger for adolescents?
who are not angry?

and select approaches

Approaches QUANTITATIVE or QUALITATIVE

leading to research activities

Research Researcher selects a representative or Researcher selects participants who are

activities (of population) sample and experiencing the phenomenon of interest
determines size before collecting data. and collects data until saturation is reached.

Researcher uses an extensive Researcher uses an intensive approach to

approach to collect data. collect data.

Questionnaires and measurement Researcher conducts interviews and

devices are preferably administered participant or nonparticipant observation
in one setting by an unbiased in environments where participants
individual to control for extraneous usually spend their time. Researcher
variables. bias is acknowledged and set aside.

Primarily deductive analysis is used, Primarily inductive analysis is used, leading

generating a numerical summary to a narrative summary, which synthesizes
that allows the researcher to reject participant information, creating a description
or accept the null hypothesis. of human experience.
 CHAPTER 2 Theoretical Framework 33

qualitative approach would be appropriate. A a small group of people who met the following
researcher can also design a study that combines criteria: (1) they had received a diagnosis of a
both categories this approach is discussed later mild or moderate form of Alzheimer’s disease or
in this chapter. a related dementia, (2) they had talked about their
ualitative research is a systematic, interac- diagnosis with their physician, (3) they lived
tive, and subjective research method used to alone in the community, (4) they spent the night
describe and give meaning to life experiences. alone, (5) they were able to speak English, and
Figure 2-3 outlines the qualitative research (6) they were 55 years of age or older. These
process. A researcher would choose to conduct a criteria ensured that the participants of the study
qualitative research study if the question to be were actually experiencing the phenomenon of
answered concerns understanding the meaning of interest (living alone while suffering with
a human experience, such as grief, hope, or loss. Alzheimer’s or dementia) and would be able to
A study completed by de Witt and colleagues provide a rich description about their individual
(2010) demonstrates the qualitative research experiences (Step 2). de Witt and colleagues then
process. They conducted a study to understand conducted in-person interviews with eight par-
the meaning of living alone from the perspec- ticipants (Step 3).
tive of older people with Alzheimer’s disease or During the interviews, they asked the partici-
some other form of dementia. Having identified pants to share their thoughts on the following:
their purpose (Step 1), the authors then selected what it is like to live alone with memory loss

Step 1
Identify the research
purpose and question.

Step 7 Step 2
Summarize the findings Select a small group of
and describe the people who have
human experience. experienced the
phenomenon of interest.

Review of
literature

Step 3
Step 6
Conduct interviews about the
Conduct further phenomenon of interest or
interviews and observations until no observe the group
new themes occur—saturation. experiencing the phenomenon.

Step 4
Step 5
Analyze the data
Review of collected and look
literature for recurring
themes.

FIGURE 2-3 Qualitative research process.

34 PART ONE Research Overview

safety and living alone with memory loss what from qualitative studies help nurses understand
it is like to need help and ask for help and the experiences or phenomena that affect patients,
future. The interviews were recorded and typed and this information in turn leads to improved
out verbatim. Then all eight interviews were care and stimulates further research. Chapters 7
reviewed together, and common themes were and 8 provide an in-depth overview of the under-
identified (Step 4). Once the first sets of themes pinnings, designs, and methods of qualitative
were identified, a second interview was com- research.
pleted with six of the eight participant to ensure Whereas the purpose of qualitative research is
that the themes identified were correct and to to create meaning about a phenomenon, that of
further clarify the findings (Steps 5 and 6). Exist- uantitative research is to systematically
ing literature was also reviewed for four primary describe a phenomenon. A researcher would
reasons: (1) to assess what was already known choose to conduct a quantitative research study if
about the topic (2) to determine how consistent the question to be answered concerned testing for
these findings were with previously published the presence of specific relationships, assessing
research (3) to further contextualize the findings for group differences, clarifying cause-and-effect
and (4) to highlight any knowledge gaps that the interactions, or explaining how effective a nursing
study filled (Step 5). intervention was. uantitative methods entail the
de Witt and colleagues (2010) studied the phe- use of objective, precise, and highly controlled
nomenon of living alone with dementia in a way measurement techniques to gather information
totally different from that in any previous studies. that can be analyzed and summarized statistically.
Their findings revealed that the human experi- Figure 2-4 outlines the quantitative research
ence related to living alone with dementia was process. Like the qualitative research process, the
strongly linked to time: stored time, dreaded time, quantitative research process begins with the
holding on to time, and limited time (Step 7). development of a research question and a purpose
As this example demonstrates, qualitative statement that highlight a relationship between
methods emphasize understanding the meaning two things.
of an experience. The context of the experience A study completed by Stewart, Reutter, Letour-
also plays a role in qualitative research. In this neau, Makwarimba, (2009) demonstrates the
study, past experiences with other people with quantitative research process. They addressed the
dementia was a context for holding bac time or following research question: What are the effects
holding on to no when facing the risks of living of a support intervention on homeless youths with
alone with memory loss. As illustrated by this respect to the quality, composition, and size of
study, qualitative research is generally conducted social network satisfaction with the support
in natural settings (in this case, the homes of older received loneliness and isolation support-
adults with dementia), and data that are words seeking coping self-efficacy mental health and
te t rather than numerical data, are used to health care–related behaviours (Step 1). nlike
describe the experiences being studied. ualita- the qualitative research process, in which
tive data are also collected from a small number researchers conduct their review of the literature
of participants, which allows an in-depth study of after they have collected their data, the quantita-
a particular phenomenon. tive research process begins with a review of
Although the methods of qualitative research literature such as journal articles, books, govern-
are systematic, a subjective approach is used that ment documents, and even Internet sources to
is, the emphasis is on capturing the personal per- determine what is known about the phenomenon
ceptions of the study participants. Thus, data of interest and theories that explain the
 CHAPTER 2 Theoretical Framework 35

Step 1
Identify the research
purpose and question.

Step 6 Step 2
Analyze the data and report Review the literature to see
whether your hypotheses what is known about the
are likely to be true or false. concepts of interest.

Step 5 Step 3

Select a sample and measure the Identify a framework that best

concepts of interest. explains how the concepts relate
to one another.

Step 4
Decide on the most
suitable and rigorous
study design.

FIGURE 2-4 Quantitative research process.

phenomenon (Step 2). In their introduction and intervention and completed a questionnaire at
review of the literature, Stewart and colleagues three times: before the intervention, midway
outlined what is already known about the benefits through the intervention, and at the end of the
of support for homeless youths. They identified intervention. The questionnaire assessed the size
how support is related to each of the six variables and characteristics of their social network, their
identified previously in this chapter. They also satisfaction with the level of support received,
provided a diagram or framework that outlines coping behaviours, and depressive symptoms
for readers exactly how support in uences stress, (Step 5). These data were analyzed statistically,
coping, and health care–related behaviours among and Stewart and colleagues showed that, overall,
homeless youths (Step 3) (Figure 2-5). the authors suggest that the intervention improved
An important part of the quantitative research the health and well-being of the youths who par-
process is to decide which design is most ticipated in it (Step 6).
appropriate for answering the research question. As demonstrated in the article by Stewart and
The numerous choices include descriptive, colleagues (2009), quantitative research tech-
correlational, longitudinal, quasiexperimental, niques are systematic, and the methodology
and experimental designs. Stewart and collea- emphasizes control of the research process, the
gues (2009) chose a quasiexperimental design environment in which the study is conducted, and
(explained in Chapter 10) to test the effectiveness how each variable is measured. In contrast to
of the intervention with 70 homeless youths (Step qualitative approaches—in which a question is
4). To show how effective the intervention was, asked and the participant is responsible for
all 70 homeless youths participated in the providing an in-depth response—quantitative
36 PART ONE Research Overview

Social Support
• Source (peer, professional)
• Functions (informational,
affirmational, emotional,
instrumental)

Processes
• Social comparison
• Social exchange
• Social learning

Stress Health/
• Homelessness Coping Functioning
• Abuse/neglect • Problem focused • Loneliness
• Family • Support seeking • Depression
• Poverty • Emotion focused • Drug use
• Life stage • Health behaviours

FIGURE 2-5 Model of conceptual foundation.

From Stewart, M., Reutter, L., Letourneau, N., & Makwarimba, E. (2009). A support interven-
tion to promote health and coping among homeless youths. Canadian Journal of Nursing
Research, 41(2), 54–77.

responses are restricted to a preselected set of is placed on including a theoretical framework.

responses. Stewart and colleagues, for example, For quantitative research projects, a theoretical
did not ask each homeless youth how the framework offers researchers and consumers
intervention changed his or her coping behav- a way to understand how a phenomenon (that
iours rather, participants responded to a tool thing of interest) comes to exist. Frameworks
that measures two types of coping, which was can be abstract or concrete, but they provide a
selected for this study and used to assess how discipline-specific cause-and-effect explanation
participants’ coping behaviour scores changed of the phenomenon. Furthermore, a framework
over three times. guides the researcher in determining the ques-
When you read research articles, remember tions to be asked/answered by the research
that researchers may vary the steps slightly, project and in the development of the study
depending on the nature of the research problem, hypotheses.
but all of the steps should be addressed The social determinant of health perspective is
systematically. one example of a theoretical framework that is
often used in nursing (Chomik, 2001). This
INTRODUCTION TO FRAMEWORKS framework, developed by a panel of interdisci-
FOR RESEARCH plinary experts (policy developers and analysts,
When research projects are conducted, particu- community representatives, and health research-
larly within the nursing discipline, great emphasis ers), suggests that people’s health and well-being
 CHAPTER 2 Theoretical Framework 37

are determined by nine factors Health Canada begins with a theoretical framework, the findings
recognizes that the last two have a cross-cutting, of a qualitative study often lead to the creation of
in uential effect on all other health determinants a theoretical framework that conveys an under-
(Chomik, p. 13): standing of people’s lived experiences.
1. Socio-economic environment including As a follow-up exercise to this introduction to
income, income distribution, and social frameworks for research, read the following story
status social support networks education and consider its message for the practising nurse
employment and working conditions and who wishes to critique, understand, and conduct
social environments research.
2. Physical environment
3. Healthy child development Kate has worked in a coronary care unit (CC ) for
4. Personal health practices nearly 3 years since graduating from nursing school.
5. Individual capacity and coping skills She has grown more comfortable with her job over
time and now believes that she can readily manage the
6. Biology and genetic endowment complexities of patient care in the CC . Recently, she
7. Health and social services has observed the pattern of blood pressure change
8. Gender when health care professionals enter a patient’s room.
9. Culture/ethnicity This observation began when Kate noticed that one of
The determinant of health framework guides her patients, a 62-year-old woman who had continuous
arterial monitoring, showed dramatic increases in
the researcher in addressing the relationship
blood pressure, as much as 100 , each time the health
between health and any of the nine factors just care team made rounds in the CC . Furthermore, this
mentioned. The researcher can determine how elevation in blood pressure persisted after the team left
many factors to focus on and in what way the the patient’s room, and then her blood pressure slowly
factors are related to each other and, in turn, to decreased to preround levels within the following
health. In some cases, the researcher may develop hour. Conversely, when the nurse manager visited the
same patient on her usual daily rounds, the patient
a diagram or a pictorial representation of these engaged calmly in conversation and was often left
relationships. with lower blood pressure when the nurse manager
Methodological frameworks serve as a guide moved on to the next patient. Kate thought about what
for conducting qualitative research studies. Rather was happening and adjusted her work so that she could
than explain how the phenomenon of interest closely observe the details of this phenomenon over
several days.
comes to exist, the methodological framework
Team rounds were led by the attending cardiologist
identifies the principles, rules, and procedures and included nurses, pharmacists, social workers,
that guide the process through which knowledge medical students, and nursing students. The team dis-
is acquired. The human becoming basic research cussed the patient, and, occasionally, she was asked to
method (Cody, 1995 Parse, 2005) has been used respond to a question about her history of heart disease
extensively to guide nursing research. In this or her current experience of chest discomfort. Partici-
pants took turns listening to her heart, and the students
method, hermeneutics are used to discover the responded to questions related to her case. In contrast,
meaning people assign to their lived experiences the nurse manager’s visit was a one-on-one meeting
as expressed in text and art. The method consists in which the patient was given the nurse’s full atten-
of a dialogue between the researcher and text or tion. Kate noticed that the nurse manager was espe-
art form to answer research questions such as cially attentive to the patient’s experience. In fact, the
nurse manager usually sat and spent time talking to the
What does it mean to be human (Parse, 2001,
patient about how her day was going, what she was
p. 172). The analysis is completed in view of the thinking about while lying in bed, and what feelings
principles of the theory of human becoming were surfacing as she began to consider how life
(Parse, 1987). Whereas the quantitative project would be when she returned home.
38 PART ONE Research Overview

Kate decided to talk to the nurse manager about her connecting with self-in-relation as the patient re ected
observations. The nurse manager, Alison, was pleased on her experience in the moment and creating ease
that Kate had noticed these blood pressure changes in when she saw the patient’s blood pressure decrease
association with interaction with health care profes- after the nurse manager’s visit. Alison and Kate shared
sionals. She told Kate that she, too, noticed these an understanding that a relationship existed between
changes during her 8-year experience of working in the patient–health care professional interaction and the
the CC . Her observation led her to the theory of patient’s blood pressure. They discussed several pos-
attentively embracing story (Liehr Smith, 2000 sible issues that might be affecting this relationship
Smith Liehr, 2003), which seemed applicable to and made a list of research questions related to each
the observation. Alison had learned the theory as a issue (Table 2-2). Their list serves only as a re ection
first-year master’s-degree student and now was apply- of the complexity of the relationship other issues
ing it in practice and beginning plans to use the theory could generate a research question contributing to
to guide her thesis research. The theory of attentively understanding of the relationship between the patient–
embracing story proposes that intentional nurse- health care professional interaction and the patient’s
patient dialogue (communication for a specific blood pressure. The list developed by Kate and Alison
purpose), which engages the human story (encourages highlights the fact that the relationship cannot be
the patient to discuss her experience), enables connect- understood with one study rather, a series of studies
ing with self-in-relation (self-re ection) to create ease may enhance understanding and offer suggestions for
(Figure 2-6). As depicted by the theory model, the
central concept of the theory is intentional dialogue TABLE 2-2
(purposeful communication), which is what Kate first ISSUES AFFECTING BLOOD PRESSURE CHANGE
observed when she noticed Alison interacting with the AND RELATED RESEARCH QUESTIONS
patient.
ISSUES RESEARCH QUESTIONS
Alison was fully attentive to the patient, following
her lead in the conversation and pursuing what mat- Number of people Is there a difference in BP for
tered most to the patient. Alison seemed to obtain a lot in the patient’s patients in the CCU when
of information in a short time, and the patient seemed room interacting with one person in
willing to share information that she was not sharing comparison with interacting
with other people. According to the theory of atten- with two or more people?
Involvement of the For the patient in the CCU, what is
tively embracing story, three concepts—intentional
patient the relationship between BP and
dialogue, connecting with self-in-relation, and creat- the amount of time spent
ing ease—are intricately connected. Thus, when Kate listening to the health care
observed intentional dialogue, she also observed team’s discussion of personal
qualities during routine rounds?
What is the effect of the nurse-
Connecting with patient intentional dialogue on
Self-in-Relation BP within the hour after the
dialogue?
• Personal history Continuing effect of What is the BP pattern of patients
• Reflective awareness experience on in the CCU from the beginning
BP over the next of routine health care rounds
Intentional Dialogue hour until 1 hour after the completion
Nurse of rounds?
• True presence Patient Content of dialogue What is the relationship between
• Querying emergence issues discussed during
intentional dialogue and BP?
Creating Ease Meaning of What is the patient’s experience of
experience for being observed during routine
• Remembering disjointed the patient health care rounds?
story moments What is the patient’s experience of
• Flow in the midst of sharing personal matters with a
anchoring nurse while in the CCU?
FIGURE 2-6 Attentively embracing story. BP, blood pressure; CCU, coronary care unit.
 CHAPTER 2 Theoretical Framework 39

change. For instance, a thorough understanding may of different interactions. This logical process
lead to testing different approaches for conducting often generates the questions that make the most
team rounds.
sense for enhancing a patient’s well-being.
Another major theme in the story of Kate and
LINKS CONNECTING PRACTICE, THEORY, Alison can be found in each nurse’s approach to
AND RESEARCH the phenomenon of the relationship between the
Several important aspects of frameworks for patient–health care professional interaction and
research are embedded in the story of Kate and blood pressure. Each nurse was using a different
Alison. First, it is important for you to notice the approach to look at the situation, but both were
links among practice, theory, and research. Each systematically evaluating what was observed.
is intricately connected with the other to create This approach is the essence of science: system-
knowledge for the discipline of nursing (Figure atic collection, analysis, and interpretation of
2-7). A theory is a set of interrelated concepts data. Kate was using inductive reasoning a
that provides a systematic view of a phenomenon. process of starting with the details of experience
Theory guides practice and research practice and moving to a general picture. Inductive rea-
enables testing of theory and generates questions soning involves the observation of a particular set
for research and research contributes to theory of instances that belong to and can be identified
building and establishing practice guidelines. as part of a larger set. Alison told Kate that she,
Thus, what is learned through practice, theory, too, had begun with inductive reasoning but now
and research constitutes the knowledge of the was using deductive reasoning a process of
discipline of nursing. From this perspective, each starting with the general picture—in this case, the
reader is in the process of contributing to the theory of attentively embracing story—and
knowledge of the discipline. For example, if you moving to a specific direction for practice and
are a practising nurse, you can use focused obser- research. In deductive reasoning, the researcher
vation (Liehr, 1992), just as Kate did, to consider uses two or more related concepts that, when
the nuances of situations that matter to patient combined, enable the researcher to suggest rela-
health. Kate noticed the changes in blood tionships between the concepts.
pressure occurring with interactions and system- Inductive and deductive reasoning are basic in
atically began to pay close attention to the effect frameworks for research. Inductive reasoning is
the pattern of figuring out what is there from
Research the details of the nursing practice experience and
is the foundation for most qualitative inquiry.
Research questions related to the issue of the
meaning of experience for the patient (see Table
Theory
2-2) can be addressed with the inductive reason-
ing of qualitative inquiry. Deductive reasoning
begins with a structure that guides searching for
what is there. All but the last two research ques-
tions listed in Table 2-2 would be addressed with
the deductive reasoning of quantitative inquiry.
In view of Alison’s use of deductive reasoning
Practice guided by the theory of attentively embracing
FIGURE 2-7 Discipline knowledge: the theory-practice- story, we can assume that she has read and cri-
research connection. tiqued the literature on theoretical frameworks
40 PART ONE Research Overview

and has chosen attentively embracing story to pregnancy as an opportunity for change, safe
guide her master’s thesis research. For Kate to health care places and relationships, responsive
move on in her thinking about research to study care, and making interventions safe and respon-
the way changes in blood pressure are related to sive. These concepts were described in the context
the patient–health care professional interaction, of the participants’ stories and relevant literature
she needs to become well-versed in the impor- thus, a conceptual structure that could be mod-
tance of theoretical frameworks. As she reads the elled was created.
literature and reviews research studies, she will A odel is a symbolic representation of a set
critique the theoretical frameworks guiding those of concepts that is created to depict relationships.
studies. By critiquing existing frameworks, she Figure 2-5 shows Stewart and colleagues’ (2009)
will develop the knowledge and understanding model of social support. It highlights the process
needed to choose an appropriate framework for through which support from peers and profes-
research. As a beginning, Kate is reading this sionals in uences the stressful life situations,
chapter, recognizing that she is critiquing nursing coping behaviours, and health care–related
research. behaviours of homeless youths. In this model,
arrows are used to depict a process that explains
how social support is related to the social network,
Helpful Hint
stress, and health functioning. For example, the
Investigators may not always provide a detailed,
explicit statement of the one or more observations that arrow from social support to processes sug-
led them to their conclusions when using inductive rea- gests that social support has an effect on social
soning. Likewise, you will not always find a clear expla- network comparison, exchange, and learning.
nation of the structure guiding a study in which
deductive reasoning is used. Whether this is positive or negative is unknown
however, the social network then in uences
coping behaviours (problem focused, support
seeking, and emotion focused), which in turn
FRAMEWORKS AS STRUCTURES in uence health care–related functioning (loneli-
FOR RESEARCH ness, depression, drug use, and health behav-
Whether you are evaluating a qualitative or a iours). This model could be the basis for deductive
quantitative study, look for the framework that reasoning. An example of a deductive question
guided the study. In general, in an article in which that could be derived from the model is as follows:
the researcher is using qualitative inquiry and What is the difference in social comparison [an
inductive reasoning methods, the framework is indicator of the quality of the social network] for
described at the end of the publication, in the homeless youths who participate in a supportive
discussion section. From the study’s findings, the intervention, and how does this influence their
researcher builds a structure for moving forward. problem-focused coping skills [one indicator of
coping]?
For example, in their study on women and vio-
lence in improving pregnancy and parenting care In an article by a researcher who uses quantita-
for Aboriginal families, D. Smith and colleagues tive inquiry and deductive reasoning methods, the
(2006) investigated experiences from community framework is described at the beginning of the
members. These stories were analyzed, and the article, before a discussion of study methods.
findings were synthesized at the theoretical level.
The researchers moved from the particulars of the The Ladder of Abstraction
experiences of pregnancy and parenting care to a The ladder of abstraction is a way for you to gain
general structure of concepts that included a perspective when reading and thinking about
 CHAPTER 2 Theoretical Framework 41

frameworks for research. When you critique the support framework (see Figure 2-5). The frame-
framework of a study, imagine a ladder (Figure work is used to identify specific concepts of inter-
2-8). The highest level on the ladder, the world- est, form several hypotheses about how the
view, includes beliefs and assumption or the para- concepts are related, and logically structure the
digm to which the research belongs. The middle study’s general orientation.
portion of the ladder includes the framework, At the lowest level on the ladder of abstraction
theories, and concepts that the researcher uses to are variables. ariables are the elements that can
articulate the problem, purpose, and structure for be observed through the senses. The key empiri-
the research. cal aspects of a study—its concepts and
For example, as stated by Stewart and associ- variables—are generally articulated through con-
ates (2009), the purpose of their study was to pilot ceptual and operational definitions. A conce tual
test a comprehensive support intervention for de nition is much like a dictionary definition,
homeless youths that was intended to optimize conveying the general meaning of the concept.
peer in uence, reduce loneliness and isolation, However, the conceptual definition goes beyond
and enhance coping skills. From their review of the general meaning found in the dictionary the
the literature, which involved reading published concept is defined as it is rooted in the theoretical
research articles about stress and coping among literature. The o erational de nition specifies
this population and book chapters that theorized how the concept will be measured: that is, what
about stress and coping, they developed a social instruments will be used to capture the essence of
the variable.

Helpful Hint
Worldview Abstract Some research reports embed conceptual defini-
tions in the literature review. The reader should find the
conceptual definitions so that the logical fit between the
conceptual and the operational definitions can be
determined.
Framework

Frameworks, Theories, and Concepts

Frameworks, theories, and concepts can be com-
Theories
pared with each other from the perspective of
abstraction, with concepts being the lowest on the
middle portion of the ladder and frameworks
the highest. Frameworks, such as the social
Concepts
determinants of health perspective, provide a
general orientation to understanding a phenome-
non of interest and identify what factors are most
significant as we examine various aspects of
Variables
health, for example. However, a framework
Concrete does not necessarily reveal how every possible
factor relates to one another. The theories specify
FIGURE 2-8 Ladder of abstraction for evaluating research the relationship between the components of a
frameworks. framework.
42 PART ONE Research Overview

Theories using an animal model. These definitions and

As stated earlier, a theory is a set of interrelated associated images of the concept of pain incorpo-
concepts that serves the purpose of explaining or rate different experiential and knowledge compo-
predicting phenomena. A theory is like a blue- nents, all with the same label: pain Therefore, it
print, a guide for modelling a structure. A blue- is important to know the meaning of the concept
print depicts the elements of a structure and the to the person. For a reader, it is important to know
relationship of each element to the other, just as the meaning that the researcher gives to the con-
a theory depicts both the concepts that compose cepts in a research study.
it and how they are related. Chinn and Kramer
(1999, p. 63) defined a theory as an expression HYPOTHESIS AND VARIABLES. Chinn and Kramer
of knowledge . . . a creative and rigorous structur- (1999) defined a hy othesis as a tentative state-
ing of ideas that project a tentative, purposeful, ment of relationship between two or more vari-
and systematic view of phenomena (p. 268). For ables that can be empirically tested (p. 254).
example, consider the relationship between A hypothesis is a best guess or prediction about
healthy child development, individual capacity what one expects to find about the variable out-
and coping, and mental well-being. Several theo- lined in the study. The variables represent the
ries that account for the relationship between most concrete components of the ladder of
these variables have been developed by nurse abstraction and are derived from the identified
researchers (i.e., Neuman’s 1995 systems concepts of a theory. They represent the measur-
model). Such models address the ways in which able properties (what can be observed, measured,
stressful events deplete people’s coping resources or both) of a quantitative study.
and, in turn, how both in uence the mental health As you read this text, you will learn how to
of people (children or adults). articulate a hypothesis, just as Kate formulated
a hypothesis about the relationship between
Concepts patient–health care professional interaction and
At the lower end of the ladder of abstraction, a blood pressure. Although Kate did not label her
conce t is an image or symbolic representation idea as a hypothesis, it was a best guess based on
of an abstract idea. Chinn and Kramer (1999, p. observation. Take a minute to recall an experience
51) defined a concept as a complex mental for- from your own nursing practice that has provoked
mulation of experience. Concepts are the major confusion, and state it as a hypothesis. A mis-
components of theory and convey the abstract match between what is known or commonly
ideas within a theory. In this chapter, the concepts accepted as fact and what we experience creates a
of the theory of attentively embracing story — hypothesis-generating moment. Every nurse ex-
intentional dialogue, connecting with self-in- periences such moments. Cultivating hypothesis-
relation, and creating ease—have been defined, generating moments requires noticing them,
and their relationships have been modelled. Each focusing one’s observation on specific details and
concept creates a mental image that is explained the relationships between them, and allowing time
further through the conceptual definition. For for creative thinking and dialogue (Liehr, 1992)
example, the concept of pain creates a mental this leads to possibilities for creating a low-level
image that is based on experience. Its experiential microrange theory or hypothesis.
meaning is different for a child who has just
fallen off a bike, for an older adult with rheuma- Frameworks
toid arthritis, and for a nurse with a doctoral The Critical Thinking Decision Path on p. 43
degree who is studying pain mechanisms by takes you through the thinking of a researcher
 CHAPTER 2 Theoretical Framework 43

who is about to begin conducting research. You prevention and cessation programs. Therefore,
can expect to find some, but not all, of the phases the point of the literature review is to build a case
of decision making addressed in a research pub- for doing the research. Researchers do not provide
lication. Beginning with the worldview, the a framework for the study because they are plan-
highest rung on the ladder of abstraction, the ning an inductive approach to study the problem.
researcher is inclined to approach a research Conversely, researchers who use deductive
problem from the perspective of inductive or reasoning must choose between a conceptual and
deductive reasoning. Researchers who pursue an a theoretical framework. In the theory literature,
inductive reasoning approach generally do not these terms are used interchangeably (Chinn
present a framework before beginning the discus- Kramer, 1999) however, in the case presented in
sion of methods. This is not to say that the litera- the Critical Thinking Decision Path, each term is
ture will not be reviewed before the methods are distinguished from the other on the basis of
introduced. For example, Bottorff and colleagues whether the researcher is creating the structure or
(2004) were interested in understanding how ado- whether the structure has already been created by
lescents view cigarette addiction because an someone else. In general, each of these terms
understanding of how adolescents view smoking refers to a structure that provides guidance for
and addiction can lead to more effective smoking research. A conce tual ra e or is a structure

CRITICAL THINKING DECISION PATH

Choosing a Theoretical Path

Highest level
Guided by a view of the world, the researcher uses of discourse

Deductive reasoning or Inductive reasoning

To create a structure To identify a structure To begin to piece

to guide research to guide research together data to address
a research question
Conceptual framework Theoretical framework

Grand theory
Midrange theory
Microrange theory

The researcher poses the research question

and the conceptual definitions of study variables
and
uses operational definitions to articulate
measurement of the study variables Lowest level
of discourse
44 PART ONE Research Overview

of concepts, theories, or both that is used to critically appraise a framework for research (see
construct a map for the study. It presents a theory, Critiquing Criteria box on the facing page).
which explains why the phenomenon being The first question posed is whether a frame-
studied exists. Generally, a conceptual framework work is presented. Sometimes, a structure may
is constructed from a review of the literature or guide the research, but a diagrammed model
is developed as part of a qualitative research is not included in the report. You must then look
project. A theoretical ra e or may also be for the study structure in the description of the
defined as a structure of concepts, theories, or study concepts. When the framework is identi-
both that is used to construct a map for the study. fied, consider its relevance for nursing. A nurse
However, it is based on a philosophical or theo- does not have to create the framework, but the
rized belief or understanding of why the phenom- importance of the study’s content for nursing
enon under study exists. should be clear. The question of how the frame-
work depicts a structure congruent with nursing
Helpful Hint should be addressed. Sometimes frameworks
When researchers have used conceptual frame- from very different disciplines, such as physics
works to guide their studies, you can expect to find a or art, may be relevant to nursing. The author
system of ideas, synthesized for the purpose of organiz- must clearly articulate the meaning of the frame-
ing thinking and providing study direction.
work for the study and link the framework to
nursing.
From the perspective of the Critical Thinking Once the meaning and relationship to nursing
Decision Path, theoretical frameworks can incor- are articulated, you will be able to determine
porate grand, midrange, or microrange theories. whether the framework is appropriate to guide the
Whether the researcher is using a conceptual or a research. For instance, a blatant mismatch occurs
theoretical framework, conceptual and then oper- if a researcher is studying students’ responses to
ational definitions will emerge from the frame- the stress of being in the clinical setting for the
work. The decision path moves down the ladder first time but presents a framework of stress
of abstraction from the philosophical to the related to recovery from chronic illness. Such
empirical level, tracking thinking from the most obvious mismatches do not generally arise
abstract to the least abstract for the purposes of however, subtle versions of mismatch do occur.
planning a research study and accruing evidence So you will need to look closely at the framework
to guide nursing practice and research. to determine whether it is appropriate and the
best fit for the research question and proposed
APPRAISING THE EVIDENCE study design.
Next, focus on the concepts being studied.
The Framework Do you know which concepts are being studied
The framework for research provides guidance and how they are defined and translated into
for the researcher as study questions are fine- measurable variables Does literature exist to
tuned, methods for measuring variables are support the choice of concepts Concepts
selected, and analyses are planned. Once data are should clearly re ect the area of study for
collected and analyzed, the framework is used as example, if in a study the general concept of
a basis for comparison. Did the findings coincide stress is used but the concept of anxiety is more
with the framework If discrepancies exist, can appropriate to the research focus, difficulties
they be explained by means of the framework will arise in defining variables and determining
The reader of research needs to know how to methods of measurement. These issues relate
 CHAPTER 2 Theoretical Framework 45

to the logical consistency within the framework, framework for continuing study and thus focus
the concepts being studied, and the methods of the direction of future research.
measurement. Evaluating frameworks for research requires
Throughout the entire critiquing process, from skill that can be acquired only through repeated
worldview to operational definitions, you are critique and discussion with other nurses who
evaluating whether the theoretical framework is have critiqued the same publication. The novice
appropriate. At the end of a research article, you reader of research must be patient while these
can expect to find a discussion of the findings as skills are developed. With continuing education
they relate to the model. This final point enables and a broader knowledge of potential frame-
readers to evaluate the framework for use in works, you will build a repertoire of knowledge
further research. The discussion may suggest nec- to enable you to judge the foundation of a research
essary changes to enhance the relevance of the study, the framework for research.

CRITIQUING CRITERIA

1. Is the framework for research 4. Are the concepts and variables 6. Is there a logical, consistent link
clearly identified? clearly and appropriately between the framework, the
2. Is the framework consistent with defined? concepts being studied, and the
a nursing perspective? 5. Did the study present sufficient methods of measurement?
3. Is the framework appropriate to literature to support the selected 7. Are the study findings examined
guide research on the subject of concepts? in relation to the framework?
interest?

■ Nurse researchers contend that a theoretical

CRITICAL THINKING CHALLENGES framework is essential for systematically
identifying the relationship among the chosen
■ Explain the difference between research that is
variables. If this is true, why do nonnursing
based on a constructivist paradigm and research
research studies not identify theoretical
that is based on a positivist paradigm.
frameworks?
■ Discuss how a researcher’s values can influence
■ As a consumer of research, how would you use
the results of a study. Include an example in
computer databases to verify tools for measuring
your answer.
operational definitions?
■ You are taking an elective course in advanced
■ How would you argue against the following
pathophysiology. The professor compares the
statement: “As a beginning consumer of
knowledge of various disciplines and states that
research, it is ridiculous to expect me to
nursing is an example of a nonscientific
determine whether a researcher’s study has an
discipline, declaring in support of this position
appropriate theoretical framework; I’ve only had
that nursing’s knowledge has been generated
Nursing Theory 101.”
with unstructured methods, such as intuition,
■ Is it possible for a research study’s
trial and error, tradition, and authority. What
theoretical framework and variables to be
assumptions has this professor made? Would
you counter or support this position? the same?
46 PART ONE Research Overview

Chinn, P. L., Kramer, M. K. (1999). Theory and

KEY POINTS nursing: Integrated no ledge development.
St. Louis: Mosby.
• The scientific approaches used to generate nursing Chinn, P. L., Kramer, M. K. (2010). Nursing’s
knowledge reflect both inductive and deductive fundamental patterns of knowing. In P. Chinn
reasoning. M. Kramer (Eds.), Integrated no ledge development
• The interaction among theory, practice, and in nursing (8th ed., pp. 1-17). St. Louis: Mosby.
research is central to knowledge development in Chomik, T. A. (2001). The population health template:
the discipline of nursing. ey elements and actions that de ne a population
• Conceptual frameworks are created by the health approach (Discussion paper). Ottawa:
researcher, whereas theoretical frameworks are Population and Public Health Branch.
identified in the literature. Cody, W. K. (1995). Of life immense in passion, pulse,
• The use of a framework for research is important
and power: Dialoguing with Whitman and Parse—A
as a guide to systematically identify concepts and
to link appropriate study variables with each hermeneutic study. In R. R. Parse (Ed.), Illumina-
concept. tions: The human becoming theory in practice and
• Conceptual and operational definitions are critical research (pp. 269-307). New York: National League
in the evolution of a study, regardless of whether for Nursing Press.
they are explicitly stated. de Witt, L., Ploeg, J., Black, M. (2010). Living alone
• In developing or selecting a framework for with dementia: An interpretive phenomenological
research, knowledge may be acquired from other study with older women. ournal of Advanced
disciplines or directly from nursing. In either case, Nursing, (8), 1698-1707.
that knowledge is used to answer specific nursing Erci I. B., Sayan, A., Tortumuoglu, G., Kilic, D., Sahub,
questions.
O., G ng rmus, . (2003). The effectiveness of
• Theory is distinguished by its scope. Grand theories
are the broadest in scope and at the highest level Watson’s caring model on the quality of life and
of abstraction, whereas microrange theories are the blood pressure of patients with hypertension. ournal
most narrow in scope and at the lowest level of of Advanced Nursing, (2), 130-139.
abstraction. Midrange theories are in the middle. Glaser, B. G., Strauss, A. L. (1967). The discovery of
• Midrange theories are at a level of abstraction that grounded theory: Strategies for ualitative research.
enhances their usefulness for guiding practice and New York: Aldine.
research. Hamric, A. B., Spross, J. A., Hanson, C. M. (2005).
• When you critique a framework for research, Advanced nursing practice (3rd ed.). Philadelphia:
examine the logical, consistent link between the W. B. Saunders.
framework, the concepts for study, and the
Liehr, P. (1992). Prelude to research. Nursing Science
methods of data collection.
uarterly, , 102-103.
Liehr, P., Smith, M. J. (2000). sing story to guide
nursing practice. International ournal of uman
Caring, (2), 13-18.
Lincoln, Y., Guba, E. (2000). Paradigmatic controver-
REFERENCES sies, contradictions, and emerging con uences. In
Benner, P. (2004). sing the Dreyfus model of skill N. K. Denzin Y. S. Lincoln (Eds.), andboo of
acquisition to describe and interpret skill acquisition ualitative research (2nd ed., pp. 163-188).
and clinical judgment in nursing practice and Thousand Oaks, CA: Sage.
education. Bulletin of Science, Technology Society, Neuman, B. (Ed.). (1995). The Neuman systems model
(3), 188. (3rd ed.). San Mateo, CA: Appleton Lange.
Benner, P., Tanner, C., Chesla, C. (1992). From Olson, J. (2006). nderstanding paradigms used for
beginner to expert: Gaining a differentiated clinical nursing research. ournal of Advanced Nursing, ,
world in critical care nursing. Advances in Nursing 459-469.
Science, (3), 13. Parse, R. R. (1987). Nursing science: Major paradigms,
Bottorff, J. L., Johnson, J. L., Moffat, B., Grewal, J., theories, and criti ues. Philadelphia: W. B.
Ratner, P., Kalaw, C. (2004). Adolescent construc- Saunders.
tions of nicotine addiction. Canadian ournal of Parse, R. R. (2001). ualitative in uiry: The path of
Nursing Research, , 22-39. sciencing. Sudbury, MA: Jones Bartlett.
 CHAPTER 2 Theoretical Framework 47

Parse, R. R. (2005). The human becoming modes of Weaver, K., Olson, J. K. (2006). nderstanding
inquiry: Emerging sciencing. Nursing Science paradigms used for nursing research. ournal of
uarterly, , 297-300. Advanced Nursing, (4), 459-469.
Smith, D., Edwards, N., Varcoe, C., Martens, P. J., ander, P. E. (2007). Ways of knowing in nursing: The
Davies, B. (2006). Bringing safety and responsive- historical evolution of a concept. ournal of Theory
ness into the forefront of care for pregnant and Construction and Testing, (1), 7-11.
parenting aboriginal people. Advances in Nursing
Science, (2), E27-E44.
Smith, M. J., Liehr, P. (2003). Middle range theory
FOR FURTHER STUDY
for nursing. New York: Springer. Go to Evolve at http://evolve.elsevier.com/
Stewart, M., Reutter, L., Letourneau, N., Canada/LoBiondo/Research for Audio Glossary, how-to
Makwarimba, E. (2009). A support intervention to instructions for Writing Proposals for Funding, and
promote health and coping among homeless youths. additional research articles for practice in reviewing
Canadian ournal of Nursing Research, (2), 54-77. and critiquing.
C H A PTER 3

Critical Reading Strategies:

Overview of the Research Process
Geri LoBiondo-Wood | Judith Haber | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Identify the steps that researchers use to conduct quantitative and qualitative research.
• Identify the importance of critical thinking and critical reading for the reading of research articles.
• Identify the steps associated with critical reading.
• Use the steps of critical reading to review research articles.
• Use identified strategies to critically read research articles.
• Use identified critical thinking and critical reading strategies to synthesize critiqued articles.
• Identify the format and style of research articles.

KEY TERMS
abstract critical reading critique
assumptions critical thinking critiquing criteria

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

48
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 49

AS YOU READ THIS TEXT, YOU will learn how the TABLE 3-1
steps of the research process unfold. The steps STEPS OF THE RESEARCH PROCESS AND
are systematic and orderly, and they relate to the JOURNAL FORMAT: QUALITATIVE RESEARCH
development of nursing knowledge. nderstand- RESEARCH PROCESS STEPS USUAL LOCATION IN JOURNAL
ing the step-by-step process that researchers use OR FORMAT ISSUES HEADING OR SUBHEADING
will help you develop the critiquing skills neces- Identification of the In abstract, introduction, or
sary to judge the soundness of research studies phenomenon both
Purpose of research In abstract, at beginning or
you will encounter in the literature. Throughout study question end of introduction, or in
the chapters in this book, research terms per- more than one of these
tinent to each step are identified, defined, and locations
illustrated with many examples from the research Literature review In introduction, discussion, or
both
literature. Four published research studies are Design In abstract, “Introduction”
featured in the appendixes, and they used as section, “Methods”
examples to illustrate significant points in each subsection titled “Design,”
“Methods” section in
chapter. Judging not only a study’s soundness general, or more than one
but also evaluating a study’s applicability to of these locations
practice is a key skill. Sample In “Methods” subsection titled
“Sample,” “Subjects,” or
Before you can judge a study, you need to “Participants”
understand the differences between and among Legal-ethical issues In section on data collection,
studies. As you read the chapters and the appen- in “Procedures” section, or
dixes, you will encounter many different study in description of sample
Data-collection procedure In “Data Collection” or
designs, as well as standards for critiquing the “Procedures” section
soundness of each step of a study and for judging Data analysis In “Methods” subsection titled
both the strength of evidence provided by a “Data Analysis” or “Data
Analysis and Interpretation”
study and its application to practice. The steps Results In abstract (briefly), In
of the qualitative research process generally separate section titled
proceed in the order outlined in Table 3-1. Table “Results” or “Findings”
Discussion and In separate “Discussion” or
3-2 outlines the highlights of the general steps recommendation “Discussion and
associated with quantitative research. Remember Implications” section
that a researcher may vary the steps slightly, References At end of article
depending on the nature of the research problem,
but all of the steps should be addressed
systematically. CRITICAL THINKING AND CRITICAL
This chapter provides an overview of critical READING SKILLS
thinking, critical reading, and critiquing skills. To develop an expertise in evidence-informed
The chapter also introduces the overall format of practice, you need to be able to critically read all
a research article and provides an overview types of research literature. As you read articles,
of subsequent chapters in the book. These com- you may notice the difference in style or format
ponents of the chapter are designed to help you between research articles and theoretical or clini-
read research articles more effectively and with cal articles. The terms in a research article may
greater understanding. You will learn about the be new to you, and the focus of its content is dif-
research process so that you will be able to prac- ferent. Reading research articles can be difficult
tise from a base of evidence to improve patient and frustrating at first, but the best way to become
outcomes. a knowledgeable consumer of research is to use
50 PART ONE Research Overview

TABLE 3-2
STEPS OF THE RESEARCH PROCESS AND JOURNAL FORMAT: QUANTITATIVE RESEARCH
RESEARCH PROCESS STEPS OR FORMAT ISSUES USUAL LOCATION IN JOURNAL HEADING OR SUBHEADING

Research problem In abstract, introduction (not labelled as a research problem), or separate

subsection titled “Problem”
Purpose In abstract or introduction, or both; at end of literature review or
discussion of theoretical framework; or in separate section titled
“Purpose”
Literature review At end of introduction but not labelled as a literature review; in separate
section titled “Literature Review,” “Review of the Literature,” or
“Related Literature”
Variables reviewed may appear as titles of sections or subsections
Theoretical framework, conceptual In “Literature Review” section (combined) or in separate sections titled
framework, or both “Theoretic Framework” and “Conceptual Framework”; or each concept
or definition used in theoretical or conceptual framework may appear
as title of separate section or subsection
Hypothesis/research questions Stated or implied near end of “Introduction” section, which may be
labelled; in separate sections or subsection titled “Hypothesis” or
“Research Questions”; or, for first time, in “Results” section
Research design In abstract or introduction (stated or implied) or in section titled
“Methods” or “Methodology”
Sample: type and size “Size”: may be stated in abstract, in “Methods” section, or in separate
“Methods” subsection as “Sample,” “Sample/Subjects,” or
“Participants”
“Type”: may be implied or stated in any of previous headings described
under size
Legal-ethical issues In section titled “Methods,” “Procedures,” “Sample,” “Subjects,” or
“Participants” (in all cases, stated or implied)
Instruments (measurement tools) In section titled “Methods,” “Instruments,” or “Measures”
Validity and reliability In section titled “Methods,” “Instruments,” ”Measures,” or “Procedures”
(specifically stated or implied)
Data-collection procedure In “Methods” subsection titled “Procedure” or “Data Collection” or in
separate section titled “Procedure”
Data analysis In “Methods” subsection under subheading “Procedure” or “Data
Analysis”
Results In separate section titled “Results”
Discussion of findings and new findings Combined with results or in separate section titled “Discussion”
Implications, limitations, and Combined with discussion or presented in separate or combined major
recommendations sections
References At end of article
Communicating research results In research articles, poster, and paper presentations

critical thinking and critical reading skills when interpretation will be based on your current
you read research articles. As a student, you are knowledge, experience, and understanding.
not expected to completely understand a research Remember that becoming a competent critical
article. It is also understood that you will find it thinker and consumer of research, like learning
challenging to critique research articles until you the steps of the research process, takes time,
obtain repeated experience doing so. Nor are you patience, and experience.
expected to develop critiquing skills on your own. Critical thin ing is the rational examination
An essential objective of this book is to help of ideas, inferences, assumptions, principles,
you acquire critical thinking and critical read- arguments, conclusions, issues, statements,
ing skills. No perfect critique exists your beliefs, and actions (Paul Elder, 2008). As
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 51

applied to critically reading research, this means BOX 3-1

that you are engaged in the following:
HIGHLIGHTS OF CRITICAL READING
Systematic understanding of the research PROCESS STRATEGIES
process Photocopy the article to be critiqued, and make
Thinking that displays a mastery of the cri- notations directly on the copy.
teria for critiquing research and evidence-
STRATEGIES FOR PRELIMINARY UNDERSTANDING
informed practice
• Keep a research textbook and a dictionary by your
The art of being able to make your thinking side.
better (i.e., clearer, more accurate, or more • Review the chapters in the textbook on the various
defensible) by clarifying what you under- steps of the research process.
• Highlight or underline on the photocopy any new
stand and what you do not know terms, unfamiliar terms, and significant sentences.
In other words, being a critical thinker means that • Look up the definitions of new terms, and write
you are consciously thinking about your own them on the photocopy.
• Highlight or underline identified steps of the
thoughts and what you say, write, read, or do, as research process.
well as what other people say, write, or do. While
thinking about all of this, you are also questioning STRATEGIES FOR COMPREHENSIVE UNDERSTANDING
the appropriateness of the content, applying stan- • Identify the main idea or theme of the article; state
it in your own words in one or two sentences.
dards or criteria, and seeing how the information • Continue to clarify terms that may be unclear on
measures up. subsequent readings.
Developing the ability to critically evaluate • Before critiquing the article, make sure you
understand the main points of each reported step
research articles requires both critical thinking of the research process that you identified.
skills and critical reading skills. Critical reading
is an active, intellectually engaging process in STRATEGIES FOR ANALYSIS UNDERSTANDING
• Using the critiquing criteria, determine how well
which the reader participates in an inner dialogue the study meets the criteria for each step of the
with the writer (Paul Elder, p. 461). A critical process.
reader actively looks for assu tions (accepted • Determine which level of evidence fits the study.
truths), key concepts and ideas, reasons and jus- • Write cues, relationships of concepts, and
questions on the photocopy.
tifications, supporting examples, parallel experi- • Ask fellow students to analyze the same study,
ences, implications and consequences, and any using the same criteria, and then compare their
other structured features of the text so as to inter- results with yours.
• Consult faculty members about your evaluation of
pret and assess the text accurately and fairly (Paul the study.
Elder, 2008).
Critical reading is a process that involves the STRATEGIES FOR SYNTHESIS UNDERSTANDING
• Review your notes on the article, and determine
following levels of understanding and allows you how each step discussed in the article compares
to critically assess a study’s validity: with the critiquing criteria.
Preliminary: familiarizing yourself with the • In your own words, type a one-page summary of
content (skimming the article) the reviewed study.
• Cite article references at the top according to the
Comprehensive: understanding the research- American Psychological Association (2010) style
er’s purpose or intent manual or another reference style.
Analysis: understanding the parts of the study • In your own words, and using the critiquing
criteria, briefly summarize each reported research
Synthesis: understanding the whole article and step.
each step of the research process in a study • In your own words, briefly describe the study’s
Box 3-1 provides more in-depth strategies for strengths and weaknesses.
attaining these levels of understanding.
52 PART ONE Research Overview

Critical thinking and critical reading skills

can be further developed by learning the research Helpful Hint
process. You will find that critical thinking If you still have difficulty understanding a research
study after using the strategies related to skimming and
and critical reading skills used in the nursing comprehensive reading, make another copy of your
process can be transferred to understanding the marked-up research article, include your specific ques-
research process and reading research articles. tions or area of difficulty, and ask your professor to read
it. Comprehensive understanding and synthesis are nec-
You will gradually be able to read an entire essary for analyzing a research article. Understanding
research article and re ect on it by identifying the author’s purpose and methods for the study reflects
and challenging assumptions, identifying key critical thinking and facilitates the evaluation of the
study.
concepts, questioning methods, and determining
whether the conclusions are based on the study’s
findings. Once you have obtained this compe-
tency in critiquing research, you will be ready to STRATEGIES FOR CRITIQUING
synthesize the findings of multiple research RESEARCH STUDIES
studies to use in developing evidence-informed The evaluation of a research article requires an
practice. appraisal or critique of the published study. The
Critiquing a research study requires several criti ue is the process of critical appraisal in
readings. At minimum, you should read it three which a person objectively and critically evalu-
or four times. The first strategy is to keep your ates a research report’s content for scientific
research textbook at your side as you read. sing validity or merit and application to practice. It
this book while you read a study may help you requires some knowledge of the subject matter,
do the following: as well as knowledge of how to read critically and
Identify the steps of the research process use critiquing criteria. Criti uing criteria are the
and how the study was conducted standards, appraisal guides, or questions used to
Clarify unfamiliar concepts or terms judge (assess) an article. Guidelines for conduct-
uestion assumptions and rationale ing a critique are presented in the following chap-
Assess the study for validity ters of this book:
As you analyze and synthesize an article, you General principles of qualitative and quan-
are ready to begin the appraisal process that will titative research in Chapters 1 through 6
help determine a study’s worth. An illustration of Criteria for critiques of qualitative research
how to use critical reading strategies is provided in Chapters 7 and 8
by the example in the Practical Application box, In-depth exploration of criteria needed for
which contains excerpts from the abstract, intro- evaluation of quantitative research in
duction, literature review, theoretical framework Chapters 9 through 11
literature, and Methods and Procedure sec- Exploration of critiquing criteria for the
tions of the quantitative study by Wong and asso- process related to research such as sam-
ciates (2010 see Appendix B). Note that this pling, data collection, rigour, data analysis
particular article contains both a literature review and presenting the findings in Chapters 12
and a discussion of the theoretical framework that through 17
clearly supports the objectives and purpose of the Summarized examples of critiquing criteria
study. Also note that in the Practical Application for qualitative studies and an example of a
box, parts of the text of these sections from the qualitative critique in Chapter 18
article were deleted so that the examples could be Summarized critiquing criteria and exam-
as concise as possible. ples of a quantitative critique in Chapter 19
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 53

Practical Application
EXAMPLE OF CRITICAL APPRAISAL READING STRATEGIES*
Introductory paragraphs, Health care restructuring in the 1990s in Canada and the Unites States contributed to significant changes
study’s purpose and in 122 senior nurse leader (SNL) roles, including expansion of their decision-making responsibilities
aims (Murray et al., 1998; Mass et al., 2006; Smith et al., 2006). In some organizations, nurse executives
were added to senior executive teams, and in others, their scope of participation in organizational
decisions related to budget, strategic planning, quality of care, and a host of challenging
organizational issues greatly increased. However, surprisingly little is actually known about the
patterns of SNL participation in decision-making (PDM) at the senior executive level of health care
organizations and, in particular, the consequences of organizational changes on nurse executive
decision-making.
Our aim was to describe SNL decision-making processes in terms of the scope and degree of their
involvement in strategic and tactical decisions at the executive management level in organizations
across Canada.

Literature review: Concepts

SNL role changes with Substantial changes in the health care system have contributed to new role expectations, higher
restructuring knowledge requirements, and increased responsibility and accountability for nurse leaders, including:
quality and effective coordination of patient services; managing many clinical areas with a broadened
span of control; operating merged facilities and decentralized structures; and decision-making in
finance, human resources and quality and safety of patient services across the continuum of care
(Anthony et al., 2005; Arnold et al., 2006; Duffield et al., 2001; Kleinman, 2003; Upenieks, 2003).
SNL role in Of the published literature on nurse leader decision-making, there is little coherence in topics such as
organizational risk propensity (Smith & Friedland, 1998), ethical decision-making (Berggren & Severinsson, 2003;
decision-making Fonville, 2002), manager role in facilitating staff participation in decision-making (Krairiksh, 2000),
middle manager involvement in organizational strategic decision-making (Ashmos et al., 1998) and
personality type and decision-making styles (Freund, 1988). Only a small body of research focused
on SNLs’ organizational decision-making influence (Dwore et al., 2000; Havens, 1998; Banaszak-Holl
et al., 1999; Wangsness, 1991).

Theoretical framework In our framework, PDM…by SNLs in executive management teams is viewed as creating new
organizational connections and mechanisms for exchanging information and enriching interpretation
of issues that ultimately influence the quality of management decisions. The scope of decision-
making is enhanced by involving SNLs at the beginning of decision-making stages (timing), and the
breadth of content expertise is expanded by their clinical and professional knowledge. The intensity
of PDM is a function of the number and range of decision-making activities involving SNLs. Any
decision-making process entails several different fundamental information processing actions from
raising issues, clarifying problems, generating and evaluating solutions to making a final choice
(Anderson & McDaniel, 1999). The greater the scope and intensity of SNL PDM, the greater the
likelihood that they and others perceive them as having an influence on decisions. Last, we propose
that decision-making influence is related to the quality of final management decisions reached.

Hypothesis 1) The scope (timing and breadth) and intensity (number of decision activities) of SNL participation in
executive decision-making processes positively predicts the degree of SNL decision influence.
2) SNL decision influence positively predicts perceived quality of operational management decisions.

Design Data were collected by mailed survey from 63 SNLs and 49 chief executive officers (CEOs) in 66 health
care organizations in 10 Canadian provinces.

Instrument We used the Participation in Strategic Decision-Making Scale (Banaszak-Holl et al., 1999) to measure SNL
decision-making processes.

*For references cited in this box, please see Appendix B.

Adapted from Wong, C. A., Laschinger, H., Cummings, G. G., Vincent, L., & O’Connor, P. (2010). Decisional involvement of senior nurse leaders in
Canadian acute care hospitals. Journal of Nursing Management, 18, 122-133.
54 PART ONE Research Overview

In analyzing a research report, you must evalu-

ate each step of the research process and ask OVERCOMING BARRIERS: USEFUL
whether each step of the process meets the crite- CRITIQUING STRATEGIES
ria. For instance, the critiquing criteria in Chapter Throughout this text, you will find special fea-
5 (p. 109) are as follows: The literature review tures that will help refine and develop your com-
identifies gaps and inconsistencies in the litera- petence as a research consumer. A critical thin ing
ture about a subject, concept, or problem and decision path related to each step of the research
all of the concepts and variables are included in process will sharpen your decision-making skills
the review. These two criteria relate to critiquing as you critique research articles. Look for Internet
the research question and the literature review resources in chapters that will enhance your
components of the research process. The Practical research consumer skills. Critical thin ing chal-
Application box shows several examples in which lenges, which appear at the end of each chapter,
Wong and associates (2010) identified gaps in are designed to reinforce your critical thinking
the literature and how they intended to fill these and critical reading skills in relation to the steps
gaps by conducting a study for the stated objec- of the research process. elpful hints, designed
tive and purpose (see Appendix B for the com- to reinforce your understanding and critical think-
plete study). Remember that when you are doing ing, appear at various points throughout the chap-
a critique, you are pointing out strengths, as well ters. Evidence-informed practice tips, which will
as weaknesses. Developing critical reading skills help you apply evidence-informed practice strate-
will enable you to complete a critique success- gies in your clinical practice, are also provided in
fully. To review the appraisal strategies that each chapter. Finally, Practical Application bo es
facilitate the understanding gained by reading offer examples of translating principles and
for analysis, see Box 3-1. Standardized critical methods of nursing research into real-life nursing
appraisal tools, such as the Critical Appraisal situations and interventions.
Skills Programme (CASP http://www.phru.nhs. Once you complete a research critique or two,
uk/Pages/PHD/CASP.htm), can also be used by you will be ready to discuss your critique with
students and clinicians to systematically appraise your fellow students and professor. Best of all,
the strength and quality of evidence provided in you can enjoy discussing the points of your
research articles (see Chapter 20). appraisal because your critique will be based on
Critiquing can be thought of as looking at a objective data, not just personal opinion. As you
completed jigsaw puzzle. Does it form a compre- continue to use and perfect critical analysis skills
hensive picture, or is a piece out of place What by critiquing studies, remember that these very
is the level of evidence provided by the study and skills are an expected clinical competency for
its findings What is the balance between the delivering evidence-informed nursing care.
risks and benefits of the findings that contribute
to clinical decisions How can the evidence be EVIDENCE-INFORMED PRACTICE
applied in the treatment of a patient or a patient AND RESEARCH
population or in a specific setting In the case of Along with gaining confidence while reading and
reading several studies for synthesis, you need to critiquing research studies, you must undertake a
consider how interrelated each of the studies are final step of reading and appraising the research
and determine the overall strength and quality of literature: deciding how, when, and whether to
evidence and its applicability to practice. Reading apply a study or studies to your practice so that
for synthesis is essential in critiquing research your practice is evidence informed. Evidence-
studies. informed practice allows you to systematically
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 55

use the best available evidence with the integra- following section introduces you to the steps of
tion of individual clinical expertise, as well as the research process as presented in published
the patient’s values and preferences, in making articles.
clinical decisions (Sackett, Straus, Richardson, Once you read an article, you will need to
Rosenburg, Hayes, 2000). Evidence-informed decide which level of evidence a research article
practice has processes and steps that are followed, provides and how well the study was designed
as does the research process. These steps are pre- and executed. Figure 3-1 depicts a model for
sented throughout the text. determining the levels of evidence associated
When you use evidence-informed practice with the design of a study, ranging from system-
strategies, the first step is to be able to read a atic reviews of randomized clinical trials to expert
research article and understand how each section opinions. The rating system or evidence hierarchy
is linked to each step of the research process. The model presented here is just one of many. Many

Rights were not granted to include this figure

in electronic media.
Please refer to the printed publication.

FIGURE 3-1 Levels of evidence: Hierarchy for rating levels of evidence, associated with a
study’s design. Evidence is assessed at a level according to its source.
From Melnyk, B. M., & Fincoult-Overholt, E. (2005). Evidence-based practice in nursing &
literature: A guide to best practice. Philadelphia, PA: Lippincott, Williams & Wilkins.
56 PART ONE Research Overview

hierarchies for assessing the relative worth of dif- chapters on quantitative research. For example,
ferent types of research literature for both the Wong and associates’ (2010) study (see Appendix
qualitative and quantitative research literature are B) is level IV because of its descriptive design,
available. whereas the study by Seneviratne and colleagues
You will note from Figure 3-1 that research (2009 see Appendix A) is level VI because of its
evidence is traditionally categorized from weakest qualitative design. Remember, as discussed
to strongest, with an emphasis on support for the earlier, that the level by itself does not reveal the
effectiveness of interventions. The concept of full worth of a study but is another tool that helps
levels of evidence tends to dominate the evidence- you think about the strengths and weaknesses of
informed practice literature, rendering unclear the a study and the nature of the evidence provided
merit of qualitative studies. Chapter 2 suggests in the findings and conclusions. The chapters on
that different research methods provide different qualitative research provide an understanding of
types and levels of evidence, all of which inform how qualitative studies can be assessed for use in
practice. practice. You will use the evidence hierarchy
Although evidence provided by qualitative presented in Figure 3-1 throughout the book as
studies seems to rank lower in the hierarchy of you develop your research consumer skills, and
evidence presented (that is, levels V and VI), so it is important to become familiar with its
Sandelowski (2004) noted that hierarchies are content.
used under the assumption that randomized clini- This rating system represents an evidence hier-
cal trials are the gold standard of research archy for judging the strength of a study’s design,
this assumption devalues qualitative research. which is just one level of assessment that in u-
However, qualitative research has increased and ences how confident the reader is about the con-
thrived over the years. Thousands of reports of clusions drawn by the researcher. Assessing
well-conducted qualitative studies exist on topics the strength of scientific evidence or potential
such as (1) personal and cultural constructions of research bias provides a vehicle to guide nurses
disease, prevention, treatment, and risk (2) living in evaluating research studies for their applicabil-
with disease and managing the physical, psycho- ity in clinical decision-making. In addition to
logical, and social effects of multiple diseases and identifying the level of evidence needed to grade
their treatment (3) decision-making experiences the strength of a body of evidence, there are the
with beginning and end of life, as well as assistive three domains of quality, quantity, and consis-
and life-extending, technological interventions tency (Agency for Healthcare Research and
and (4) contextual factors favouring and mitigat- uality, 2002):
ing against quality care, health promotion, pre- uality: the extent to which a study’s
vention of disease, and reduction of health design, implementation, and analysis mini-
disparities (Sandelowski, 2004). The answers mizes bias
provided by qualitative data re ect important uantity: the number of studies in which
evidence that may offer valuable insights about the research question has been evaluated,
a particular phenomenon, patient population, or including overall sample size across studies,
clinical situation. It is important to remember that as well as the strength of the findings from
researchers, in choosing which research method- the data analyses
ology to use, base their decision primarily on the Consistency: the degree to which similar
question they are trying to answer. findings are reported from investigations of
The meaningfulness of an evidence rating the same research question in studies that
system will become clearer to you as you read the have similar and different designs
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 57

Assess
Ask Gather Act Evaluate
Appraise

FIGURE 3-2 Evidence-informed practice steps.

Evidence-informed practice has specific pro- emphasis to the method, results, and discussion
cesses and steps that are followed, as does the of implications than to the details of assumptions,
research process. The steps of the process are to hypotheses, or definitions of terms. Decisions
ask, gather, assess and appraise, act, and evaluate about the amount of material presented for each
(Figure 3-2). Chapter 20 provides an overview of step of the research process are constrained by the
evidence-informed practice and introduces you to following:
the steps and strategies associated with evidence- A journal’s space limitations
informed practice. A journal’s author guidelines
The type or nature of the study
RESEARCH ARTICLES: FORMAT An individual researcher’s evaluation of
AND STYLE what is the most important component of
Before you consider reading research articles, it the study
is important to have a sense of their organization The following discussion provides a brief
and format. Many journals publish either only overview of each step of the research process and
research articles or research in addition to clinical how it might appear in an article (refer to Tables
or theoretical articles. Although many journals 3-1 and 3-2). It is important to remember that the
have some common features, they also have format of a quantitative research article will differ
unique characteristics. All journals have guide- from that of a qualitative research article.
lines for manuscript preparation and submission
these guidelines are published by each journal. A Abstract
review of these guidelines will give you an idea An abstract is a short, comprehensive synopsis
of the format of articles that appear in specific or summary of a study at the beginning of an
journals. article. An abstract quickly focuses the reader
It is important to remember that even though on the main points of a study. A well-presented
each step of the research process is discussed at abstract is accurate, self-contained, concise, spe-
length in this text, you may find only a short cific, nonevaluative, coherent, and readable.
paragraph or a sentence in the research article that Abstracts vary in length from 50 to 250 words.
gives the details of the step in a specific study. The length and format of an abstract are dictated
Because of the journal’s publishing guidelines, by the journal’s style. Both quantitative and quali-
the published study that appears in a journal is a tative research studies have abstracts that provide
shortened version of the complete work carried a succinct overview of the study. An example
out by the researcher or researchers. You will also of an abstract can be found at the beginning of
find that some researchers devote more space in the study by Sobieraj and colleagues (2009 see
an article to the results, whereas others present a Appendix C). Their abstract follows an outline
longer discussion of the methods and procedures. format that highlights the major steps of the study.
Since the 1990s, most authors have given more It reads in part as follows:
58 PART ONE Research Overview

The purpose of this quasi-experimental study was to end of the Literature Review or Conceptual
test an intervention on the use of music during simple Framework section. The study’s purpose may or
laceration repair to promote parent-led distraction in may not be labelled as such, or it may be referred
children aged 1 to 5.
to as the study’s aim or objective. Thorne and
The remainder of the abstract provides a synopsis associates (2009 see Appendix D) described the
of the background of the study and the methods, purpose of the study in the last paragraph in the
results, and conclusions. All of the other studies section on background to the literature:
in the appendixes have abstracts. The general problem that this article addresses is the
continuing prevalence of poor communication during
Helpful Hint the diagnostic window of time within the cancer
A journal abstract is usually a single paragraph experience. Specifically, we report findings associated
that provides a general reference to the research with those communication encounters that patients
purpose, research questions, or hypothesis, or a combi- experience as problematic during the diagnostic period
nation of these aspects, and highlights the methodology within the context of a longitudinal cohort study of
and results, as well as the implications for future practice patient perspectives of cancer communication across
or research.
the illness trajectory.

In contrast, Seneviratne and colleagues (2009)

Introduction had a section titled Aim (see Appendix A):
Early in a research article, in a section that may The aim of the study was to uncover nurses’
or may not be titled Introduction, the researcher perceptions of the contexts of caring for acute stroke
presents a background picture of the area survivors.
researched and its significance to practice. In the
study by Thorne and associates (2009 see In a section titled Purpose, Sobieraj and col-
Appendix D) on communication during the cancer leagues (2009) stated, The purpose of this study
diagnostic period, the reader can find the basis of was to test an intervention by using children’s
the research question early in the report: songs to promote parent-led distraction during
Although it is well known that communication will simple laceration repair in children aged 1 to 5
have a strong impact on both the experience and (see Appendix C).
psychosocial outcomes of cancer . . . poor
communication remains a significant problem facing Literature Review and
cancer patients across the spectrum of service. . . . In Theoretical Framework
part, this is because of the complexity of
communication and the variability of the human Authors of studies and journal articles present the
experience of cancer, both of which create significant literature review and theoretical framework in
challenges for the development of a solid empirical different ways. In many research articles, the lit-
science within this field. erature review is merged with the discussion of
Another example is found in the study by Sen- the theoretical framework. The resulting section
eviratne and associates (2009 see Appendix A): includes the main concepts investigated and may
Nurses have long believed that they play an essential
be titled Review of the Literature, Literature
role in stroke care, but they remain uncertain about Review, Theoretical Framework, Related
the nature of their contributions. Literature, Background, or Conceptual
Framework or it may not be a separate section
Definition of the Purpose at all. In reviewing Appendixes A through D, you
The purpose of the study is defined either at the will find differences in the headings used. Wong
end of the researcher’s initial introduction or at the and associates (2010) had both a Literature
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 59

Review and a Theoretical Framework section the rigour of the critiquing criteria addressed does
(see Appendix B) in the articles in Appendixes not substantially change, some of the terminology
A and D, headings that re ect the theoretical con- of the questions differs for qualitative and quan-
cepts of the study were used in the Background titative studies. For instance, in regard to the
sections, but the authors did not call such discus- study by Wong and associates (2010 see Appen-
sion a literature review or a framework. Finally, dix B), you might ask whether the hypotheses
Appendix C has a Literature Review section. were generated from the theoretical framework or
One style is not better than another all of the literature review and whether the design chosen
studies in the appendixes contain all of the critical was appropriate and consistent with the study’s
elements but present the elements differently. questions and purpose. With a qualitative study
such as that by Thorne and associates (2009 see
Hypothesis or Research Question Appendix D), however, you might be asking
A study’s research questions or hypotheses can whether the researchers conducted the study in a
also be presented in different ways. Research manner consistent with the principles of qualita-
reports in journals often do not have separate tive research and therefore focused on the identi-
headings for reporting the Hypotheses or fication of the themes of knowledge and choice.
Research uestion. They are often embedded Do not get discouraged if you cannot easily
in the Introduction or Background section or determine the design. More often than not, the
not labelled at all (e.g., as in the studies in the specific design is not stated or, if an advanced
appendixes). uantitative research studies have design is used, the details are not spelled out. One
hypotheses or research questions. If a researcher of the best strategies is to review the chapters in
uses hypotheses in a study, the researcher may this text that address designs and to ask your
report whether the hypotheses were or were not professors for assistance. The following tips will
supported such reporting occurs toward the end help you determine whether the study you are
of the article, in the Results or Findings reading employs a quantitative design:
section. Wong and associates (2010 see Appen- Hypotheses are stated or implied.
dix B) list the hypotheses in a section titled The terms control and treatment group
Hypotheses, and Sobieraj and colleagues (2009 appear.
see Appendix C) discuss the hypothesis in the The term survey, correlational, or e post
Purpose section. ualitative research studies facto is used.
do not have hypotheses but do have research The term random or convenience is men-
questions and purposes. tioned in relation to the sample.
Variables are measured by instruments or
Research Design scales.
The type of research design can be found in the Reliability and validity of instruments are
abstract, within the purpose statement, or in the discussed.
introduction to the Procedures or Methods Statistical analyses are used.
section, or it may not be stated at all. For example, In contrast, qualitative studies do not usually
the studies in Appendixes A, B, C, and D all focus on numbers. In some articles about quali-
identify the design type in both the abstract and tative studies, standard quantitative terms (e.g.,
the body of the study report. subjects) may be used rather than qualitative
One of your first objectives is to determine terms (e.g., informants or participants). Deciding
whether the study is qualitative or quantitative so on the type of qualitative design can be confus-
that the appropriate criteria are used. Although ing one of the best strategies is to review this
60 PART ONE Research Overview

text’s chapters on qualitative design, as well as to

critique qualitative studies. Begin trying to link Helpful Hint
the study’s design with the level of evidence asso- Remember that not all research articles include
headings related to each step or component of the
ciated with that design as illustrated in Figure 3-1. research process, but each step is presented at some
This will give you a context for evaluating the point in the article.
strength and consistency of the findings and their
applicability to practice. Although many studies In some cases, researchers do not report the
may not specify the particular design used, all reliability and validity of commonly used, estab-
studies inform the reader of the specific method- lished instruments in an article and may refer you
ology used, which can help you decide the type to other references. Ask for assistance from your
of design used to guide the study. instructor if you are in doubt about the validity or
reliability of a study’s instruments. ualitative
Sampling researchers typically report on the validity of the
The population from which the sample was drawn findings and on how potential biases are acknowl-
is discussed in the section titled Methods or edged and worked with (see Chapter 7). For
Methodology under the subheadings of Sub- example, Seneviratne and colleagues (2009 see
jects, Participants, or Sample. For example, Appendix A) described how they dealt with
Seneviratne and colleagues (2009) discuss the potential biases in a section titled Re exivity.
sample in a section titled Participants and
Setting that was separate from Methodology Procedures and Data-Collection Methods
(see Appendix A). However, Thorne and associ- The procedures used to collect data or the step-
ates (2009) present the sample selection criteria by-step way in which the researcher used the
in a section titled Study Method (see Appendix measures (instruments or scales) is generally
D). Researchers should describe both the popula- described in the Procedures section. In each of
tion from which the sample was chosen and the the studies in Appendixes A through D, the
number of participants who took part in the study, researchers indicated how they conducted the
and they should also mention whether partici- study in detail in subsections titled Data Collec-
pants dropped out of the study. The authors of all tion, Sample and Data Collection Procedures,
of the studies in the appendixes discuss their and Study Method. Notice that the researchers
samples in enough detail so that who the partici- in each study in Appendixes A through D also
pants were and how they were selected are quite stated that the studies were approved by an insti-
clear. tutional review board (see Chapter 6), thereby
ensuring that each met ethical standards.
Reliability and Validity
The discussion related to instruments used to Data Analysis/Results
measure the variables of a study is usually The data analysis procedures (i.e., the statistical
included in a Methods subsection titled Instru- tests used and the results of descriptive and infer-
ments or Measures. The researcher usually ential tests applied in quantitative studies) are
describes the particular measure (i.e., instrument presented in the section titled Results or
or scale) used by discussing its reliability and Findings. Although qualitative studies do not
validity. Wong and associates (2010 see Appen- involve the use of statistical tests, the procedures
dix B) describe the reliability of the instrument to for analyzing the themes, concepts, and observa-
measure participation in decision making in a tional or print data are usually described in the
Methods subsection titled Instrument. Methods or Data Collection section and
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 61

reported in the Results, Findings, or Data

Analysis section. Seneviratne and colleagues References
(2009) reported on the methods of qualitative All of the references cited in a research article are
data analysis in their Data Analysis section and included at the end of the article. The main
the results of their analysis in their Findings purpose of the reference list is to support the
sections (see Appendix A). The article by Wong material presented by identifying the sources in a
and associates (2010) has several sections, one manner that allows for easy retrieval by the
titled Data Analysis, which describes how the reader. Journals have various referencing styles
data were analyzed, and four subsections under to organize references.
Results, which describe the results of the
hypotheses tested (see Appendix B). Communicating Results
Communicating the results of a study can take the
Discussion form of a research article, poster, or paper presen-
The last part of an article about a research study tation. All are valid ways of providing nurses with
is the Discussion section. In this section, the the data and the ability to provide high-quality
researchers explain how all of the parts of the patient care that is based on research findings.
study are related and analyze the study as a whole. Evidence-informed nursing care plans and
The researchers refer to the literature reviewed practice protocols, guidelines, or standards are
and discuss how their study is similar to or dif- outcome measures that effectively indicate com-
ferent from other studies. Researchers may report municated research.
the results and discussion in one section but As you develop critical thinking and reading
usually report their results in separate Results skills by using the strategies presented in this
and Discussion sections (see Appendixes A, B, chapter, you will become more familiar with the
C, and D). One way is not better than another. research and appraisal processes. Your ability to
Journal and space limitations determine how read and critique research articles will gradually
these sections are handled. Any new findings or improve. You will be well on your way to becom-
unexpected findings are usually described in the ing a knowledgeable user of research from nursing
Discussion section. and other scientific disciplines for application in
nursing practice.
Recommendations and Implications
In some articles, a separate section titled Con- Helpful Hint
clusions describes the implications of the find- When writing a paper on a specific concept or
ings for practice and education, as well as related topic that requires you to critique and synthesize the
findings from several studies, you might find it useful
limitations based on the findings, and future to create an evidence table of the data. Include the
studies may be recommended (see Appendixes A, following information: author, date, type of study,
B, and D) in other articles, this information design, level of evidence, sample, data analysis, find-
ings, and implications.
appears in several sections with such titles as
Discussion, Limitations, Nursing Implica-
tions, Implications for Research and Practice, SYSTEMATIC REVIEWS: META-ANALYSES,
and Summary. Sobieraj and colleagues (2009 INTEGRATIVE REVIEWS, AND
see Appendix C) concluded their study with a META-SYNTHESES
discussion subsection titled Practice Recom- Another variety of articles that is appearing more
mendations. Again, one way is not better than frequently in the literature and is very important
another—only different. for understanding evidence-informed practice are
62 PART ONE Research Overview

systematic reviews. Systematic reviews include Conclusions summary: In the conclusions

meta-analyses, integrative reviews, and meta- or summary, the strength, quality, and con-
syntheses. The authors of these articles investi- sistency of the data are described as they
gate a number of studies related to a specific apply to practice. This section contains rec-
clinical question and, using a specific set of cri- ommendations about which aspects of prac-
teria and methods, evaluate those articles as a tice are supported by the data in the articles
whole. The methods detailed here are not pre- and for which aspects further research is
scriptive but serve as a general outline of how you needed to more fully answer the question
will find these articles formatted. Overall, posed in the review.
although they vary somewhat in their approach, For example, Purc-Stephenson and Thrasher
these reviews are intended, in essence, to better (2010) reviewed qualitative studies on telephone
inform practice and develop evidence-informed triage and advice (1) to understand nurses’ experi-
practice. ences with telephone triage and advice and (2) to
The components of these types of articles are identify factors that facilitate or impeded the
as follows: nurses’ decision-making processes. In searching
Bac ground: The introduction covers the peer-reviewed research literature databases
content related to the background of the from 1980 to 2008, the authors found 144 poten-
clinical question and clarifies the specific tial studies. After the inclusion criteria were
question that the review answers. The arti- applied (to determine which studies were rele-
cle’s authors clarify the definitions of the vant) and a quality appraisal review, 16 studies
concepts in the question so that the reader were chosen for the final sample. Each of the
understand the concepts that were used in reviewers then extracted key themes from each
assessment. of the studies and then compared them with
Method: The methods used for searching one another. Once the themes were refined, the
the literature are detailed. The exact elec- researchers synthesized the findings to identify
tronic databases, the dates, and the key new interpretations.
words used to conduct the search are pro-
vided. In addition, the article details the CLINICAL GUIDELINES
inclusion and exclusion criteria by which Clinical guidelines are systematically developed
the literature was chosen to review and cri- statements or recommendations that serve as a
tique. If a number of articles were found guide for practitioners. Guidelines have been
and not used, the authors detail why articles developed to assist in bridging practice and
were excluded from the review. research. Guidelines are developed by profes-
Appraisal of the literature: The articles that sional organizations, government agencies, insti-
are included in the literature review are dis- tutions, and convened expert panels. Guidelines
cussed in the body of the article, and an provide clinicians with an algorithm for clinical
evidence table is used to present the high- management or for decision making with regard
lights of each article. The author uses the to specific diseases (e.g., colon cancer) or treat-
evidence table to compare and contrast the ments (e.g., pain management). Not all guidelines
articles, critique them for scientific validity, are well developed and, like research, must be
and discuss how well they answer the clini- assessed before implementation. Clinical guide-
cal question. If the author uses a meta- lines, although they are systematically developed
analysis format, a summary of the data is and make explicit recommendations for practice,
presented. may be formatted differently. Guidelines should
 CHAPTER 3 Critical Reading Strategies: Overview of the Research Process 63

clearly present scope and purpose of the practice,

detail who contributed to the development of the KEY POINTS
guidelines, demonstrate scientific rigour, demon-
strate clinical applicability, and demonstrate • Critical thinking and critical reading skills will
enable you to question the appropriateness of the
editorial independence. The Appraisal of Guide- content of a research article, apply standards or
lines Research and Evaluation (AGREE http:// critiquing criteria to assess the study’s scientific
www.agreecollaboration.org) has developed an merit for use in practice, and consider alternative
ways of handling the same topic.
instrument for assessing the quality of clinical • Critical reading involves active interpretation and
guidelines. objective assessment of an article and searching for
As you venture through this textbook, you will key concepts, ideas, and justifications.
be challenged to think about not only reading and • Critical reading requires four stages of
understanding: preliminary (skimming),
understanding research studies but also applying comprehensive, analysis, and synthesis. Each stage
the findings to your practice. Nursing has a rich is characterized by specific strategies to increase
legacy of research that has grown in depth and your critical reading skills.
• Critically reading for preliminary understanding is
breadth. Producers of research and clinicians accomplished by skimming or quickly and lightly
must engage in a joint effort to translate findings reading an article in order to familiarize yourself
into practice that will make a difference in the with its content and to provide you with a general
sense of the material.
care of patients and families. • Critically reading for a comprehensive
understanding is designed to increase your
understanding both of the concepts and research
terms in relation to the context and of the parts of
the study in relation to the whole study, as
presented in the article.
CRITICAL THINKING CHALLENGES • Critically reading for analysis understanding is
designed to divide the content into parts so that
■ The critical reading of research articles may each part of the study is understood. The critiquing
require a minimum of three or four readings. Is process begins at this stage.
this always the case? What assumptions underlie • Critical reading to reach the goal of synthesis
this claim? understanding combines the parts of a research
■ Why is it necessary to reach an analysis stage of study into a whole. During this final stage, the
reader determines how each step of the research
critical reading before you can critique a study? process relates to all the other steps, how well the
■ To synthesize a research article, what questions study meets the critiquing criteria, and the
must you first be able to answer? usefulness of the study for practice.
■ If registered nurses are not expected to conduct • Critiquing is the process of objectively and critically
research, how can nursing students be expected evaluating the strengths and weaknesses of a
research article for scientific merit and application
to critique each step of the nursing process, an to practice, theory, and education. The need for
entire study, or several studies? more research on the topic or clinical problem is
■ Discuss several strategies that might motivate also addressed at this stage.
practising nurses to critically appraise research • Each article should be reviewed for level of
articles. evidence as a means of judging the application of
the findings to practice.
■ What level of evidence is presented in each of • Research articles have different formats and styles,
the articles that appear in Appendixes A, B, C, depending on journal manuscript requirements and
and D? Justify your answers. whether they are quantitative or qualitative studies.
• The basic steps of the research process are
presented in journal articles in various ways.
Detailed examples of such variations can be found
in chapters throughout this text.
• Evidence-informed practice begins with the careful
reading and understanding of research articles.
64 PART ONE Research Overview

REFERENCES practice. ournal of Advanced Nursing, (9),

1872-1881. doi: 10.1111/j.1365-2648.2009.05053.x
Agency for Healthcare Research and uality. (2002). Sobieraj, G., Bhatt, M., LeMay, S., Rennick, J.,
Systems to rate the strength of scienti c evidence Johnston, C. (2009). The effect of music on parental
(File inventory, Evidence Report/Technology participation during pediatric laceration repair.
Assessment No. 47, AHR Publication No. Canadian ournal of Nursing Research, (4), 68-82.
02-E016). Rockville, MD: Author. Thorne, S., Armstrong, E., Harris, S. R., Hislop, T. G.,
American Psychological Association. (2010). Publica- Kim-Sing, C., Oglova, V., . . . Stajduhar, K. L.
tion manual of the American Psychological Associa- (2009). Patient real-time and 12-month retrospective
tion (6th ed.). Washington, DC: Author. perceptions of difficult communications in the cancer
Paul, R., Elder, L. (2008). The miniature guide to diagnostic period. ualitative ealth Research,
critical thin ing concepts and tools. Dillon Beach, (10), 1983-1394. doi: 10.1177/1049732.309348382
CA: Foundation for Critical Thinking Press. Wong, C. A., Laschinger, H., Cummings, G. G.,
Purc-Stephenson, R. J., Thrasher, C. (2010). Tele- Vincent, L., O’Connor, P. (2010). Decisional
phone triage and advice: A meta-ethnography. involvement of senior nurse leaders in Canadian
ournal of Advanced Nursing, (3), 482-494. doi: acute care hospitals. ournal of Nursing Management,
10.1111/j.1365-2648.2010.05275.x , 122-133. doi: 10.1111/j.1365-2834.2010.01053.x
Sackett, D. L., Straus, S. E., Richardson, W. S.,
Rosenburg, W., Hayes, R. B. (2000). Evidence-
based medicine: o to practise and teach EBM.
London: Churchill Livingstone.
FOR FURTHER STUDY
Sandelowski, M. (2004). sing qualitative research. Go to Evolve at http://evolve.elsevier.com/
ualitative health research, (10), 1366-1386. Canada/LoBiondo/Research for Audio Glossary, how-to
Seneviratne, C. C., Mather, C. M., Then, K. L. instructions for Writing Proposals for Funding, and
(2009). nderstanding nursing on an acute stroke additional research articles for practice in reviewing
unit: Perceptions of space, time and interprofessional and critiquing.
C H A PTER 4

Developing Research Questions,

Hypotheses, and Clinical
Questions
Judith Haber | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Discuss the purpose of developing a research question.
• Describe how the research question and hypothesis are related to the other components of the
research process.
• Describe the process of identifying and refining a research question.
• Identify the criteria for determining the significance of a research question.
• Discuss the appropriate use of the purpose, aim, or objective of a research study.
• Discuss how the purpose, research question, and hypothesis suggest which level of evidence is to
be obtained from the findings of a research study.
• Identify the characteristics of research questions and hypotheses.
• Describe the advantages and disadvantages of directional and nondirectional hypotheses.
• Compare the use of statistical hypotheses with that of research hypotheses.
• Discuss the appropriate use of research questions versus hypotheses in a research study.
• Discuss the differences between a research question and a clinical question in relation to evidence-
informed practice.
• Identify the criteria used for critiquing a research question and a hypothesis.
• Apply the critiquing criteria to the evaluation of a research question and a hypothesis in a research
report.

KEY TERMS
clinical question nondirectional hypothesis research question
dependent variable population statistical hypothesis
directional hypothesis problem statement testability
hypothesis purpose testable
independent variable research hypothesis variable
65
66 PART ONE Research Overview

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

AS YOU READ EACH CHAPTER, REMEMBER that each that emerge from practice. These are often articu-
step of the research process is defined and dis- lated in a roble state ent such as the follow-
cussed as to how that particular step relates to ing, posed by Thorne and associates (2009): The
evidence-informed practice. All research studies general problem that this article addresses is the
begin with questions and/or hypotheses. The first continuing prevalence of poor communication
step in developing evidence-informed practice is during the diagnostic window of time within the
also to ask a question, but it is a clinical question. cancer experience (see Appendix D).
The purpose of research questions and hypothe- For an investigator conducting a study, the
ses in a research study discussed in the beginning research question or the hypothesis is a key
of this chapter are different from those of the preliminary step in the research process. The
clinical questions found in a project concerning research uestion presents the idea that is to be
evidence-informed practice. In a research study, examined in the study and is the foundation of the
the research question and hypothesis lead to the research study. Once the research question is
development of a research study, whereas in a clear, the researcher selects the most appropriate
project on evidence-informed practice, the clini- research design. If the research question is pri-
cal question is what leads to the study. At the marily explorative, descriptive, or theory generat-
beginning of this chapter, you will learn about ing, the researcher opts for qualitative methods.
research questions and hypotheses from the per- In these studies, a hypothesis is not formulated.
spective of the researcher, which, in the second For studies in which the researcher is seeking a
part of this chapter, will help you to generate your specific answer to a research question, however,
own clinical questions that you will use to guide a hypothesis is generated and tested.
the development of evidence-informed practice Hypotheses can be considered intelligent
projects. From a clinician’s perspective, you must hunches, guesses, or predictions that help
understand how the research question and hypoth- researchers seek the solution or answer to the
esis align with the rest of the study. In your role research question. Hypotheses are a vehicle for
as a practising nurse, the clinical questions you testing the validity of the theoretical framework
will develop (see Chapter 20) represent the first assumptions and provide a bridge between theory
step of the evidence-informed practice process. and actuality. In the scientific world, researchers
When nurses ask certain questions, they are derive hypotheses from theories and subject
often well on their way to developing a research them to empirical testing. A theory’s validity is
question or hypothesis. Such questions include not directly examined. Instead, through testing
What is happening in this situation What hypotheses, researchers can evaluate the merit of
are the patient’s experiences Why are things a theory.
being done this way I wonder what would For a clinician making an evidence-informed
happen if . . . What characteristics are associ- decision about a patient care issue, a clinical
ated with . . . and What is the effect of . . . question—such as whether chlorhexidine or
on patient outcomes Research questions are povidone-iodine is more effective in preventing
usually generated from situations or problems infections in central catheters—would guide the
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 67

nurse in searching for and retrieving the best research study should be able to discern that the
available evidence. This evidence, combined with researcher has done the following:
clinical expertise and patient preferences, would 1. Defined a specific topic area
provide an answer on which to base the most 2. Reviewed the relevant scientific literature
effective decision about patient care for the 3. Examined the question’s potential signifi-
affected population. cance in nursing
Research questions or hypotheses often appear 4. Pragmatically examined the feasibility of
at the beginning of research articles. However, studying the research question
because of space constraints or stylistic consider-
ations in journal publications, the research ques- Defining the Research Question
tion or hypothesis may be embedded in the Brainstorming with teachers, advisers, or col-
purpose, aims, goals, or even the results section leagues may provide valuable feedback to help
of the research report. Both the consumer and the the researcher focus on a specific question area.
producer of research need to understand the For example, suppose a researcher told a col-
importance of research questions and hypotheses league that an area of interest was whether men
as the foundational elements of a research study. and women recovered differently after cardiac
This chapter provides methods of developing surgery. The colleague may have said, What is
research questions and hypotheses, the standards it about the topic that specifically interests you
for writing them, and a set of criteria for evaluat- Such a conversation may have initiated a train of
ing them. It also highlights the importance of thought that resulted in a decision to explore the
clinical questions and how to develop them. recovery processes and gender differences. Box
4-1 illustrates how a broad area of interest was
DEVELOPING AND REFINING A RESEARCH narrowed to a specific research topic.
QUESTION: STUDY PERSPECTIVE
A researcher spends a great deal of time refining Evidence-Informed Practice Tip
a research idea or problem into a research ques- A well-developed research question guides a
tion. nfortunately, the evaluator of a research focused search for scientific evidence about assessing,
diagnosing, treating, or assisting patients with under-
study is not privy to this creative process because standing their prognosis with regard to a specific health
it occurs during the study’s conceptualization. problem.
The final research question usually does not
appear in the research article unless the study is
qualitative rather than quantitative. Although this Beginning the Literature Review
section does not teach you how to formulate a The literature review should reveal a collection
research question, it does provide an important of relevant individual studies and systematic
glimpse into the researcher’s process of develop- reviews that have been critically examined (see
ing a research question. Chapter 5). Concluding sections in such articles—
Research questions or topics do not arise spon- that is, the recommendations and implications for
taneously. As shown in Table 4-1, research ques- practice—often identify remaining gaps in the
tions should indicate that practical experience, literature, the need for replication, or the need for
critical appraisal of the scientific literature, or extension of the knowledge base about a particu-
interest in an untested theory was the basis for the lar research focus.
generation of a research idea. The research ques- ualitative and quantitative researchers con-
tion should re ect a refinement of the researcher’s duct literature reviews differently. For qualitative
initial thinking. The evaluator of a nursing researchers, the value of the literature review
68 PART ONE Research Overview

TABLE 4-1
HOW PRACTICAL EXPERIENCE, SCIENTIFIC LITERATURE, AND UNTESTED THEORY INFLUENCE THE
DEVELOPMENT OF A RESEARCH IDEA
AREA INFLUENCE EXAMPLE

Practical experience Clinical practice provides a wealth of Of the 98,500 emergency visits by children, 25% are
experience from which research for the treatment of lacerations and open wounds.
problems can be derived. The nurse Although the treatment is relatively painless with
may observe the occurrence of a the use of topical anaesthesia, the fear, anxiety, and
particular event or pattern and distress associated with the experience is significant.
become curious about why it occurs, Several techniques, such as distraction, have proved
as well as its relationship to other to have a positive effect on procedural distress. Can
factors in the patient’s environment. music be a useful tool to distract the child and
involve the parent in positive behaviour (Sobieraj,
Bhatt, LeMay, Rennick, & Johnston, 2009)?
Critical appraisal of the The critical appraisal of research studies Several studies have been conducted on the needs of
scientific literature that appear in journals may family members of patients in the ICU. A subset
indirectly suggest a problem area by focused on the informational needs of families.
stimulating the reader’s thinking. Families benefitted from informational interventions,
Nurses may observe the outcome as evidenced by improved comprehension,
data from a single study or a group decreased anxiety, and increased satisfaction. The
of related studies that provide the researchers recognized a need for better
basis for developing a pilot study or understanding of the family members’ (1) perception
quality improvement project to of informational support, (2) anxiety levels, and (3)
determine the effectiveness of this satisfaction with care and the relationships among
intervention in their own practice. these. The overall objective was to further refine the
informational program and to initiate a formal
evaluation program (Bailey, Sabbagh, Loiselle,
Boileau, & McVey, 2010).
A research idea may also be suggested Workplace bullying is prevalent in Canada, affecting
by a critical appraisal of the literature millions of women every year. Being bullied at work
that identifies gaps and suggests is a devastating life experience with many negative
areas for future study. Research ideas consequences particularly related to physical,
also can be generated by research emotional, social, and economic well-being.
reports that suggest the value of Although many of the consequences have been
replicating a particular study to studied extensively, absence because of sickness has
extend or refine the existing body of not been extensively explored (O’Donnell,
scientific knowledge. MacIntosh, & Wuest, 2010).
Verification of an untested nursing Health care structuring in Canada has resulted in
theory provides a relatively considerable role changes for senior nurse leaders
uncharted territory from which (SNLs), providing the opportunity for nurse leaders
research questions can be derived. to leverage their leadership skills and play a greater
Inasmuch as theories themselves are role in decision making at the senior level. Little is
not tested, a researcher may think known about the patterns of SNL decision making.
about investigating a particular Using an adapted theoretical framework on health
concept or set of concepts related to care professionals’ participation in strategic decision
a particular nursing theory. The making in health care organizations, Wong and
deductive process would be used to associates (2010) described the scope and degree of
generate the research question. The involvement of SNLs in executive level decisions in
researcher would pose questions acute care organizations across Canada (see
such as, “If this theory is correct, Appendix B).
what kind of behaviour will I expect
to observe in particular patients and
under which conditions?” or “If this
theory is valid, what kind of
supporting evidence will I find?”
ICU, intensive care unit
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 69

BOX 4-1 usually start with a very cursory or general review

DEVELOPMENT OF A RESEARCH HYPOTHESIS of the literature to help focus the study, whereas
quantitative researchers begin their study with
IDEA EMERGES
an extensive review of the literature on their
• Senior nurse leaders’ (SNLs’) work post-restructuring
research questions and related topics (Streubert
BRAINSTORMING Carpenter, 2011). The literature review also helps
• Do the changes subsequent to the health care researchers determine whether their study can
restructuring provide new opportunities for
leadership and greater roles in decision making? contribute to the field of nursing. ualitative
• Do the changes result in diminished authority and researchers conduct a literature review during the
a decrease in representation at the policymaking data analysis or discussion of the findings to tell
level?
• What do nursing leaders perceive about their the reader how the findings fit into what is already
decision making? known about the phenomenon (Streubert
• What are the organizational outcomes? Carpenter, 2011, p. 26).
LITERATURE REVIEW The databases that researchers use for the litera-
• In many cases the restructuring resulted in ture review—for example, Cumulative Index to
opportunities to provide leadership and to play a Nursing and Allied Health Literature (CINAHL),
greater role in decision-making. PsycINFO, MEDLINE, and PubMed—contain
• However, in other organizations without discipline-
based nursing service, SNLs reported decreased relevant articles that have been critically exam-
direct supervision of nurses. ined. Concluding sections in such articles (i.e.,
• Decision-making involvement can be measured, the recommendations and implications for prac-
and there is a connection between involvement in
decision activities, perceived influence over tice) often identify remaining gaps in the litera-
decisions, and organizational outcomes. ture, the need for replication, or the need for
• The scope and intensity of SNL involvement in extension of the knowledge gleaned on a particular
strategic decision-making is related to their
research focus.
perceptions of influence in the organization.
In the example about decision making among
VARIABLES senior nurse leaders (SNLs see Box 4-1), the
• Independent (predictor) variable researchers may have conducted a preliminary
• Scope (timing and breadth) and intensity of
participation (number of decision activities) review of books and journals for theories and
• Dependent variable research studies on the changes in the health care
• Decision-making influence system, governance structures, and organizational
RESEARCH QUESTION models and factors related to decision making.
• Does the scope and intensity of SNL participation These factors, termed variables in the language
in executive decision-making processes predict the of research, should be potentially relevant, of
degree of SNL decision influence?
interest, and measurable.
From Wong, C. A., Laschinger, H., Cummings, G. G., Vincent, L., & The search for relevant factors to SNL deci-
O’Connor, P. (2010). Decisional involvement of senior nurse leaders
in Canadian acute care hospitals. Journal of Nursing Management,
sion making that are mentioned in the literature
18, 122–133. doi; 10.1111/j.1365-2834.2010.01053.x might begin with an exploration of the scope and
See Appendix B.
degree of SNLs’ contributions to executive-level
decisions in acute care organizations across
is controversial. Many researchers believe that Canada. Wong and associates (2010) investigated
an extensive literature review causes investiga- participation in strategic organizational decision
tors to develop biases or beliefs that limit their making as a theoretical framework for describing
openness to exploring the phenomenon under such participation with the use of information
study. As a general rule, qualitative researchers processing and complexity science theories. The
70 PART ONE Research Overview

researcher can then use this information to further lenge an existing theory, or clarify a con-
define the research question, to address a gap in ict in the literature.
the literature, and to extend the body of knowl- The findings will potentially enable profes-
edge related to decision making among SNLs. At sionals to formulate or alter nursing prac-
this point, the researcher could write the follow- tices or policies.
ing tentative research question: What is the If the research question has not met any of these
scope and degree of SNL involvement after criteria, the researcher needs to extensively revise
restructuring in acute care organizations across the question or discard it. For example, in the
Canada After reading this question, you should research question Does the scope and intensity
be able to envision the interrelatedness of the of SNL participation in executive decision-
initial definition of the research question, the lit- making processes predict the degree of SNL deci-
erature review, and the refined research question. sion in uence (see Box 4-1), the significance
Readers of research reports examine the end of the question includes the following facts:
product of this process in the form of a research Health care restructuring has contributed to
question, hypothesis, or both. Thus, readers significant changes in SNLs’ roles.
need an appreciation of how the researcher for- New governance structures and organiza-
mulates the final research question directing the tional models have radically changed nurs-
study. ing leadership structures.
SNLs’ participation in decision making is
important for an organization’s strategic
Helpful Hint
decisions.
Reading the literature review or theoretical frame-
work section of a research article helps you trace the Participation in organizations’ strategic
development of the implied research question, hypoth- decisions is associated with reductions in
esis, or both. hospital costs and improvement in patient
outcomes.

Examining Significance
When considering a research question, it is crucial Evidence-Informed Practice Tip
that the researcher has examined the question’s Without a well-developed research question, the
researcher may search for incorrect, irrelevant, or unnec-
potential significance to nursing. The research essary information. Such information is a barrier to iden-
question should have the potential to contribute tifying the potential significance of the study.
to and extend the scientific body of nursing
knowledge. Guidelines for selecting research
questions should meet the following criteria:
Patients, nurses, the medical community Determining Feasibility
in general, and society will potentially The feasibility of a research question must be
benefit from the knowledge derived from examined pragmatically. Regardless of how sig-
the study. nificant or researchable a question may be, prag-
The results will be applicable for nursing matic considerations—such as time availability
practice, education, or administration. of participants, facilities, equipment, and money
The results will be theoretically relevant. experience of the researcher and any ethical
The findings will lend support to untested considerations—may render the question inap-
theoretical assumptions, extend or chal- propriate because it lacks feasibility.
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 71

THE FULLY DEVELOPED RESEARCH Evidence-Informed Practice Tip

QUESTION The answers to questions generated by qualita-
tive data reflect evidence that may provide the first
As discussed previously, qualitative researchers insights about a phenomenon that has not been studied
develop and refine a research question that out- previously.
lines a general topic area. Examples include the
following: Variables
What is the impact of an education program Researchers call the properties that they study
about acquired immune deficiency syn- variables Such properties take on different
drome (AIDS) on the lives of a group values. Thus, a variable is, as the name suggests,
of gandan nurses and nurse-midwives something that varies. Properties that differ from
(Harrowing Mill, 2010) each other, such as age, weight, height, religion,
What is the effect of a brief, focused edu- and ethnicity, are examples of variables. Research-
cational intervention on the quality of verbal ers attempt to understand how and why differ-
interactions between nursing staff and ences in one variable relate to differences in
patients in a chronic care facility (Boscart, another variable. For example, a researcher may
2009) be concerned about the rate of pneumonia in post-
What is it about immunizing children who operative patients on ventilators in critical care
strongly resist needle injection that is a units. This rate is a variable because not all criti-
problem for public health nurses (Ives cally ill postoperative patients on ventilators have
Melrose, 2010) pneumonia. A researcher may also be interested
As a quantitative researcher finalizes a research in what other factors can be linked to ventilator-
question, the following three characteristics acquired pneumonia (VAP). Clinical evidence
should be evident: suggests that elevation of the head of the bed is
1. The variables under consideration are also associated with VAP. You can see that these
clearly identified. factors are also variables that need to be consid-
2. The population being investigated is ered in relation to the development of VAP in
specified. postoperative patients.
3. The possibility of empirical testing is When speaking of variables, the researcher is
implied. essentially asking, Is related to What is the
Because each of these elements is crucial in effect of on How are 1 and 2 related to
the formulation of a satisfactory research ques- The researcher is asking a question about
tion, the criteria are discussed in greater detail in the relationship between one or more independent
the following sections. These elements can often variables and a dependent variable .
be found in the introduction of the published An inde endent variable usually symbolized
article however, they are not always stated in an by , is the variable that has the presumed effect
explicit manner. on the dependent variable. In experimental
research studies, the researcher manipulates the
Helpful Hint independent variable. For example, a nurse may
Remember that research questions are used to
guide all types of research studies, but they are most Note: In cases in which multiple independent or depen-
often used in exploratory, descriptive, qualitative, or
dent variables are present, subscripts are used to indicate
hypothesis-generating studies.
the number of variables under consideration.
72 PART ONE Research Overview

study how different methods of administering between age and negative attitudes about patients
pain medication affect the patient’s perception of with hepatitis C: that is, the older the nurses were,
pain intensity. The researcher may manipulate the the more negative were their attitudes about
independent variable (i.e., the method of admin- patients with hepatitis C. This example highlights
istering pain medication) by using nurse- versus the fact that causal relationships are not necessar-
patient-controlled administration of analgesics. In ily implied by the independent and dependent
nonexperimental research, the independent vari- variables rather, only a relational statement with
able is not manipulated and is assumed to have possible directionality is proposed.
occurred naturally before or during the study. For Table 4-2 presents a number of examples to
example, the researcher may be studying the rela- help you learn how to write research questions.
tionship between gender and the perception Practise substituting other variables for the exam-
of pain intensity. The independent variable— ples in the table. You will be surprised at the skill
gender—is not manipulated it is presumed to you develop in writing and critiquing research
exist and is observed and measured in relation to questions.
pain intensity. Although one independent variable and one
The de endent variable represented by , is dependent variable were used in the examples
often referred to as the consequence or the pre- just given, there is no restriction on the number
sumed effect that varies with a change in the of variables that can be included in a research
independent variable. The dependent variable is question. Remember, however, that questions
not manipulated. It is observed and assumed to should not be unnecessarily complex or unwieldy,
vary with changes in the independent variable. particularly in beginning research efforts.
Predictions are based on how changes to the inde- Research questions that include more than one
pendent variable will affect the dependent vari- independent or dependent variable may be divided
able. The researcher is interested in understanding, into more concise subquestions.
explaining, or predicting the response of the Finally, note that variables are not inherently
dependent variable. For example, a researcher independent or dependent. A variable that is clas-
might assume that the perception of pain (i.e., the sified as independent in one study may be consid-
dependent variable) will vary according to the ered dependent in another study. For example, a
person’s gender (i.e., the independent variable). In nurse may review an article about sexual behav-
this case, the researcher is trying to explain the iours that are predictive of the risk for HIV infec-
perception of pain in relation to the gender: that is, tion or AIDS. In this case, HIV/AIDS is the
male or female. Although variability in the depen- dependent variable. In another article in which
dent variable is assumed to depend on changes in the relationship between HIV/AIDS and maternal
the independent variable, this assumption does not parenting practices is considered, HIV/AIDS
imply that a causal relationship exists between status is the independent variable. Whether a vari-
and or that changes in cause to change. able is independent or dependent depends on the
In a study about nurses’ attitudes toward role it plays in a particular study.
patients with hepatitis C, the researcher discov-
ered that older nurses had a more negative atti- Population
tude about such patients than did younger nurses. The o ulation (a well-defined set that has
The researcher did not conclude that the nurses’ certain properties) is either specified or implied
attitudes toward patients with hepatitis C were in the research question. If the scope of the ques-
negative because of their age however, it is tion has been narrowed to a specific focus and the
apparent that there was a directional relationship variables have been clearly identified, the nature
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 73

TABLE 4-2
RESEARCH QUESTION FORMAT
TYPE FORMAT EXAMPLE

QUANTITATIVE EXPERIMENTAL
Correlational Is there a relationship between X Is there a relationship between the effectiveness of pain
(independent variable) and Y management strategies and quality of life?
(dependent variable) in the specified
population?
Comparative Is there a difference in Y (dependent Is there a difference in prevention of osteoporosis in at-risk
variable) between people who have survivors of breast cancer who receive a combination of
characteristic X (independent variable) long-term progressive strength training exercises,
and those who do not have alendronate, calcium, and vitamin D, in comparison
characteristic X? with those who do not receive this treatment?
Quantitative Is there a difference in Y (dependent What is the difference in physical, social, and emotional
variable) between Group A, which adjustment in women with breast cancer (and their
received X (independent variable), and partners) who have received phase-specific standardized
Group B, which did not receive X? education by video versus phase-specific telephone
counselling?

QUALITATIVE
Phenomenological What is or was it like to have X? How do older adults learn to live with early-stage
dementia?

of the population is evident to the reader of the

research report. For example, a research question Testability
may be Is there a relationship between the type The research question must be phrased in such
of discharge planning for older adults hospital- a way that there is a specific issue that needs
ized with heart failure and the outcomes for par- to be answered. In many cases, the question is
ticipating patients and their caregivers This testable that is, measurable by quantitative
question suggests that the population under con- methods. For example, the research question
sideration includes older adults hospitalized for Should postoperative patients control how much
heart failure and their caregivers. The question pain medication they receive is stated incor-
also implies that some of the older adults and rectly for a variety of reasons. One reason is that
their caregivers were involved in a professional- the question is not testable it represents a value
patient partnership model of discharge planning, statement rather than a relational problem state-
in contrast to other older adults who received the ment. A scientific or relational question must
usual discharge planning. The researcher or propose a relationship between an independent
reader will have an initial idea of the composition variable and a dependent variable in such a way
of the study population from the outset. that the variables can be measured. Many inter-
esting and important questions are not valid
Evidence-Informed Practice Tip research questions because they are not amenable
Make sure that the population of interest and the to testing.
setting have been clearly described so that if you plan The question Should postoperative patients
to replicate the study, you will know exactly who the control how much pain medication they receive
study population needs to be.
could be revised from a philosophical question to
74 PART ONE Research Overview

a research question that implies testability. TABLE 4-3

Examples of the revised research question are as COMPONENTS OF THE RESEARCH QUESTION
follows: AND RELATED CRITERIA
Is there a relationship between patient-
Testability Differential effect of pain intensity and
controlled analgesia versus nurse- number of painful sites on functional
administered analgesia and the perception disability (physical and social functioning)
of postoperative pain Population Black and white older adults
Variables Independent Variables:
What is the effect of patient-controlled Pain intensity
analgesia on pain ratings provided by post- Pain sites
operative patients Race
Health (number of limiting diagnoses)
These examples illustrate the relationship between
Dependent Variables:
the variables, identify the independent and depen-
Management effectiveness
dent variables, and imply the testability of the Functional status
research question.
Now that the elements of the formal research
question have been presented in greater
Helpful Hint
detail, this information can be integrated by for-
Remember that research questions are often not
mulating a formal research question. Earlier in explicitly stated. The reader must infer the research
this chapter, the following unrefined research question from the report’s title, the abstract, the intro-
question was formulated: What is the scope duction, or the purpose.
Using your focused question, search the literature for
and degree of SNL involvement after restruc- the best available answer to your clinical questions.
turing in acute care organizations across
Canada
This research question was originally derived STUDY PURPOSE, AIMS, OR OBJECTIVES
from a general area of interest: understanding Once the research question is developed and the
the new role of SNLs in acute care organizations. literature review is critiqued in terms of the level,
The topic was more specifically defined by delin- strength, and quality of evidence available for the
eating a particular research question. The ques- particular research question, the purpose, aims, or
tion crystallized further after a preliminary objectives of the study become focused, and the
literature review and emerged in the unrefined researcher can decide whether a hypothesis
form just given. It is now possible to propose a should be tested or a research question answered.
refined research question in which the problem The ur ose of the study encompasses the
is stated specifically in question form and the aims or objectives the investigator hopes to
relationship of the key variables in the study, achieve with the research, not the question to be
the population being studied, and the empirical answered. Purpose, aims, and objectives are syn-
testability of the question are specified: Does the onymous terms. For example, a nurse working
scope and intensity of SNL participation in exec- with patients with bladder dysfunction who are in
utive decision-making processes predict the rehabilitation may be disturbed by the high inci-
degree of SNL decision in uence (see Appen- dence of urinary tract infections. The nurse may
dix B). propose the following research question: What
As another example, Table 4-3 lists the com- is the optimum frequency of changing urinary
ponents of the research question regarding vari- drainage bags in patients with bladder dysfunc-
ance in the perception of pain in relation to a tion to reduce the incidence of urinary tract infec-
person’s race. tion If this nurse were to design a study, its
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 75

purpose might be to determine the differential

Evidence-Informed Practice Tip
effect of 1-week and 4-week schedules of chang-
The purpose, aims, or objectives often provide
ing urinary drainage bags on the incidence of the most information about the intent of the research
urinary tract infections in patients with bladder question and hypotheses and suggest the level of evi-
dysfunction. dence to be obtained from the findings of the study.
The purpose communicates more than just the
nature of the question. Through the researcher’s
selection of verbs, the purpose statement suggests DEVELOPING THE RESEARCH HYPOTHESIS
the manner in which the researcher sought to Like the research question, hypotheses are often
study the question. Verbs such as discover, not stated explicitly in a research article. The
e plore, or describe suggest an investigation of a hypotheses are often embedded in the data analy-
relatively underresearched topic that might be sis, results, or discussion section of the research
more appropriately guided by research questions report. You then need to discern the nature of the
than by hypotheses. In contrast, verbs such as hypotheses being tested. Similarly, the population
test (testing the effectiveness of an intervention) may not be explicitly described but is identified
or compare (comparing two alternative nurs- in the background, significance, and literature
ing strategies) suggest a study with a better- review. It is then up to you to discern the nature
established body of knowledge that is hypothesis of the hypotheses and population being tested.
testing in nature. Box 4-2 provides other exam- For example, in the study by Meneses and col-
ples of purpose statements. leagues (2007), the hypotheses are embedded in
the Methods section of the article: the overall
BOX 4-2 hypotheses are to determine the effect of the
EXAMPLES OF PURPOSE STATEMENTS breast cancer psychoeducational intervention
“The purpose of this study was to describe satisfactory (BCEI) on overall quality of life . . . and on the
and unsatisfactory experiences of postpartum nursing individual quality of life . . . and to examine
care from the perspective of adolescent mothers.” whether the effects of the intervention were
(Peterson, Sword, Charles, & DiCenso, 2007, p. 201)
“The aim of this phenomenological study was to
durable over time. You must interpret this state-
understand the meaning of living alone from the ment as representing the hypotheses that test the
perspective of older people with Alzheimer disease or effect of the BCEI on quality of life in survivors
a related dementia.” (de Witt, Ploeg, & Black, 2010, p. of breast cancer. Therefore, it is important for you
1700)
“The aim of this meta-ethnography [a method to be acquainted with the components of hypoth-
used to synthesize qualitative research findings] was eses, how they are developed, and the standards
to explore nurses’ experience with telephone triage for writing and evaluating them.
and advice within the primary-care sector and to
understand the factors that facilitate or impede their Hypotheses ow from the research question,
decision-making process.” (Purc-Stephenson & literature review, and theoretical framework.
Thrasher, 2010, p. 483) Figure 4-1 illustrates this ow. A hy othesis is a
“The purpose of this study was to test an
intervention using children’s songs to promote
statement about the relationship between two or
parent-led distraction during simple laceration repair in more variables that suggests an answer to the
children aged 1 to 5?” (Sobieraj, Bhatt, LeMay, research question. A hypothesis converts the
Rennick, & Johnston, 2009, p. 70) question posed by the research question into a
“Specifically, this study set out to explore nurses’
perceptions of the impact of the 15-minute interview declarative statement that predicts an expected
on the hospital admission process and on their family outcome. It explains or predicts the relationship
nursing practice.” (Martinez, D’Artois, & Rennick, or differences between two or more variables in
2007, p. 162)
terms of the expected results or outcomes of a
76 PART ONE Research Overview

relationship between two or more variables. This

implies a systematic relationship between an
independent variable and a dependent vari-
able The direction of the predicted relation-
Hypothesis
ship is also specified in this statement. Phrases
Theoretical
framework such as greater than less than positively
related, negatively related, or curvilinearly
related and difference in connote the direc-
tionality that is proposed in the hypothesis. The
following is an example of a directional hypoth-
Research
question esis: the rate of continuous smoking abstinence
Literature dependent variable at 6 months post partum,
review according to self-report and biochemical valida-
tion, will be significantly higher in the treatment
group receiving postpartum counselling inter-
vention than in the control group independent
FIGURE 4-1 Interrelationships of the research question,
literature review, theoretical framework, and hypothesis. variable . The dependent and independent vari-
ables are explicitly identified, and the relational
aspect of the prediction in the hypothesis is con-
study. Hypotheses are formulated before the tained in the phrase significantly higher than.
study is actually conducted they provide direc- The nature of the relationship, either causal or
tion for the collection, analysis, and interpretation associative, is also implied by the hypothesis. A
of data. causal relationship is one in which the researcher
can predict that the independent variable
Helpful Hint causes a change in the dependent variable In
When hypotheses are not explicitly stated by the
research, it is rare that a definitive stand can be
author at the end of the “Introduction” section or before assumed about a cause-and-effect relationship.
the “Methods” section, they are embedded or implied For example, a researcher might hypothesize that
in the “Results” or “Discussion” section of a research relaxation training would have a significant effect
article.
on the physical and psychological health status of
patients who have suffered myocardial infarction.
Characteristics The researcher would have difficulty predicting a
Nurses who are conducting research or critiquing strong cause-and-effect relationship, however,
published research studies must have a working because the multiple intervening variables (e.g.,
knowledge of what constitutes a good hypoth- age, medication, and lifestyle changes) might also
esis. Such knowledge provides a standard for in uence the participant’s health status.
evaluating their own work and the work of others. Variables are more commonly related in non-
The following discussion about the characteris- causal ways that is, the variables are related but
tics of hypotheses presents criteria to be used in an associative way. This means that variables
when a hypothesis is formulated or evaluated. change in relation to each other. For example,
because strong evidence exists that asbestos
Relationship Statement exposure is related to lung cancer, a researcher
The first characteristic of a hypothesis is that it is may be tempted to state a causal relationship
a declarative statement identifying the predicted between asbestos exposure and lung cancer.
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 77

However, not all individuals exposed to asbestos framework, and to the hypotheses. For example:
develop lung cancer and, conversely, not all indi- The scope (timing and breadth) and intensity
viduals who have lung cancer have been exposed (number of decision activities) of SNLs’ partici-
to asbestos. Thus, a position advocating a causal pation in executive decision-making processes is
relationship between these two variables would positively predictive of the degree of SNLs’ in u-
be scientifically unsound. Instead, only an asso- ence on decisions (see Appendix B). This example
ciative relationship exists between the variables makes clear that an explicitly developed, relevant
of asbestos exposure and lung cancer, with a body of scientific data provides the theoretical
strong systematic association between the two grounding for the study.
phenomena.
Wording the Hypothesis
Testability As you become more familiar with the scientific
The second characteristic of a hypothesis is its literature, you will observe that a hypothesis can
testability The variables of the study must lend be worded in various ways. Regardless of the
themselves to observation, measurement, and specific format used to state the hypothesis, the
analysis. The hypothesis is either supported or not statement should be worded in clear, simple, and
supported after the data have been collected and concise terms. If this criterion is met, the reader
analyzed. The predicted outcome proposed by the will understand the following:
hypothesis is or is not congruent with the actual The variables of the hypothesis
outcome when the hypothesis is tested. Hypoth- The population being studied
eses advance scientific knowledge by confirming The predicted outcome of the hypothesis
or refuting theories. Information about hypotheses may be further
A hypothesis may fail to meet the criteria of clarified in the Instruments, Sample, or
testability because the researcher has not made a Methods section of a research report.
prediction about the anticipated outcome, because
the variables are not observable or measurable, or Statistical Versus Research Hypotheses
because the hypothesis is couched in terms that Readers of research reports may observe that
are value laden. a hypothesis is further categorized as either a
research or statistical hypothesis. A research
Helpful Hint hy othesis also known as a scienti c hypothesis,
When a hypothesis is complex (i.e., contains consists of a statement about the expected rela-
more than one independent or dependent variable), it tionship of the variables. A research hypothesis
is difficult for the findings to indicate unequivocally that
the hypothesis is supported or not supported. In such indicates what the outcome of the study is
cases, the reader must infer which relationships are sig- expected to be. A research hypothesis is also
nificant from the “Findings” or “Discussion” section. either directional or nondirectional. If the
researcher obtains statistically significant find-
Theory Base ings for a research hypothesis, the hypothesis is
A sound hypothesis is consistent with an existing supported. For example, in a study of the effec-
body of theory and research findings. Whether a tiveness of a home-based nursing intervention in
researcher arrives at a hypothesis inductively or reducing parenting stress in three groups of fami-
deductively, the hypothesis must be based on a lies with irritable infants, the research hypothesis
sound scientific rationale. Readers should be able was that mothers who received the home-based
to identify the ow of ideas from the research nursing intervention (REST—reassurance,
question to the literature review, to the theoretical empathy, support, and time-out) for infant
78 PART ONE Research Overview

irritability will report less parenting stress than state that it is more exact and conservative statis-
the mothers who did not receive the intervention tically and that failure to reject the statistical
(Keefe, Kajrlsen, Lobo, Kotzer, Dudley, 2006). hypothesis implies that the evidence to support
Because the findings for this hypothesis were not the idea of a real difference is insufficient. You
statistically significant, the hypothesis was not will note that research hypotheses are generally
supported, thereby indicating that the REST inter- used more often than statistical hypotheses
vention did not significantly reduce parenting because they are more desirable for stating the
stress for parents with irritable infants. The exam- researcher’s expectation. Readers then have a
ples in Table 4-4 represent research hypotheses. more precise idea of the proposed outcome. In
According to a statistical hy othesis (also any study that involves statistical analysis, the
known as a null hypothesis), there is no relation- underlying statistical hypothesis is usually
ship between the independent and dependent assumed without being explicitly stated.
variables. The examples in Table 4-5 illustrate
statistical hypotheses. If, in the data analysis, a Directional versus Nondirectional
statistically significant relationship emerges Hypotheses
between the variables at a specified level of sig- Hypotheses can be formulated directionally or
nificance, the statistical hypothesis is rejected. nondirectionally. A directional hy othesis speci-
Rejection of the statistical hypothesis is equiva- fies the expected direction of the relationship
lent to acceptance of the research hypothesis. For between the independent and dependent vari-
example, Simonson and colleagues (2007) sought ables. The reader of a directional hypothesis may
to identify differences in the rates of anaesthetic observe not only that a relationship is proposed
complications in hospitals whose obstetric anaes- but also the nature or direction of that relation-
thesia is provided solely by certified registered ship. The following is an example of a directional
nurse anaesthetists (CRNAs) in comparison with hypothesis: The scope (timing and breadth) and
hospitals with only anaesthesiologists. The statis- intensity (number of decision-making processes)
tical hypothesis—that there would be no differ- positively predicts the degree of SNL decision
ences in anaesthetic complication rates between in uence (Wong et al., 2010, p. 125 see Appen-
the hospitals that relied on different anaesthesia dix B). Sobieraj and colleagues (2009) hypoth-
providers—was supported. Because the differ- esized that parents in the intervention group
ence in outcomes was not greater than that would demonstrate a greater degree of parent-led
expected by chance, the statistical hypothesis was distraction than those in the control group (see
accepted. To further differentiate between a sta- Appendix C). Examples of directional hypotheses
tistical hypothesis and a research hypothesis, con- can also be found in examples 2 to 7 of Table 4-4.
sider the following hypotheses: Whereas a nondirectional hy othesis indi-
Research hypothesis: Hospitals with higher cates the existence of a relationship between the
nurse-to-patient ratios will have fewer variables, it does not specify the anticipated
adverse patient events. direction of the relationship. The following is an
Statistical null hypothesis: There is no differ- example of a nondirectional hypothesis there
ence in the number of adverse patient events will be a difference in the level of fatigue expe-
in hospitals with higher nurse-to-patient rienced by two groups of caregivers of preterm
ratios. infants (infants on apnea monitors versus those
Some researchers refer to the statistical hypoth- not on apnea monitors) during three time periods:
esis as a statistical contrivance that obscures a prior to discharge, 1 week after discharge and 1
straightforward prediction of the outcome. Others month after discharge.
 TABLE 4-4
EXAMPLES OF HOW TO WORD A HYPOTHESIS
VARIABLES HYPOTHESIS TYPE OF DESIGN; LEVEL OF EVIDENCE SUGGESTED

1. There are significant differences in self-reported cancer pain, symptoms accompanying pain, and functional status
according to self-reported ethnic identity.
Independent Nondirectional, research Nonexperimental; level IV
Ethnic identity
Dependent
Self-reported cancer pain
Symptoms accompanying pain
Functional status
2. Individuals who participate in usual care plus blood pressure telemonitoring will have a greater reduction in blood
pressure from baseline to 12-month follow-up than would individuals who receive only usual care.
Independent Directional, research Experimental; level II
Telemonitoring
Usual care
Dependent
Blood pressure
3. There will be a greater decrease in state anxiety scores for patients receiving structured informational videos before
abdominal or chest tube removal than for patients receiving standard information.
Independent Directional, research Experimental; level II
Preprocedure structured videotape
information
Standard information
Dependent
State anxiety
4. The incidence and degree of severity of participants’ discomfort will be lower after administration of medications by
the Z-track intramuscular injection technique than after administration of medications by the standard intramuscular
injection technique.
Independent Directional, research Experimental; level II
Z-track intramuscular injection
technique
Standard intramuscular injection
technique
Dependent
Participant discomfort
5. Nurses with high levels of social support from coworkers have low perceived job stress.
Independent Directional, research Nonexperimental; level IV
Social support
Dependent
Perceived job stress
6. There will be no difference in rates of complications from anaesthetics between hospitals in which anaesthetics are
administered primarily by certified registered nurse anaesthetists (CRNAs) and hospitals in which anaesthetics are
administered primarily by anaesthesiologists (MDs).
Independent Nondirectional; null Nonexperimental; level IV
Type of anaesthesia provider (CRNA or
MD)
Dependent
Anaesthesia complication rate
7. There will be no significant difference in the duration of patency of a 24-gauge intravenous lock in a neonatal
patient when flushed with 0.5 mL of heparinized saline (2 U/mL), standard practice, in comparison with 0.5 mL of
0.9% normal saline.
Independent Nondirectional; null Experimental; level II
Heparinized saline
Normal saline
Dependent
Duration of patency of intravenous lock
80 PART ONE Research Overview

TABLE 4-5
EXAMPLES OF STATISTICAL (NULL) HYPOTHESES
HYPOTHESIS VARIABLES TYPE OF DESIGN SUGGESTED

Oxygen inhalation by nasal cannula of up to 6 L/min Independent: Oxygen inhalation by Experimental

does not affect oral temperature measurement nasal cannula
taken with an electronic thermometer. Dependent: Oral temperature
There will be no difference in performance accuracy Independent: Nurse practitioner Nonexperimental
between adult nurse practitioners (ANPs) and (ANP or FNP) category
family nurse practitioners (FNPs) in formulating Dependent: Diagnosis and
accurate diagnoses and acceptable interventions for intervention performance accuracy
suspected cases of domestic violence.

Nurses who are learning to critique research You should note that nondirectional hypoth-
studies should be aware that both the directional eses may also be deduced from a theory
and nondirectional forms of hypothesis state- base. Because of the exploratory nature of
ments are acceptable. There are definite advan- many studies for which the hypotheses are
tages and disadvantages that pertain to each form. nondirectional, in contrast, the theory base
Proponents of the directional hypothesis argue may not be as developed.
that researchers naturally have hunches, guesses, Directional hypotheses provide a specific
or expectations about the outcome of their theoretical frame of reference within which
research. It is the hunch, the curiosity, or the the study is being conducted.
guess that initially leads them to speculate about They suggest that the researcher believes
the question. The literature review and the con- that the evidence is indicative of a particular
ceptual framework provide the theoretical foun- outcome, and as a result, the analyses of
dation for deriving the hypothesis. For example, data can be accomplished in a statistically
the theory (e.g., self-efficacy theory) provides a more sensitive way.
critical rationale for proposing that relationships The important point about the directionality of
between variables have particular outcomes. the hypotheses is whether the rationale for the
When there is no theory or related research on choice the researcher has proposed is sound.
which to base a rationale, or when findings in
previous research studies are ambivalent, a non-
directional hypothesis may be appropriate. As RELATIONSHIP AMONG THE HYPOTHESIS,
you read research articles, you will note that THE RESEARCH QUESTION, AND THE
directional hypotheses are much more commonly RESEARCH DESIGN
used than nondirectional hypotheses. Regardless of whether the researcher uses a sta-
In summary, when you evaluate a hypothesis, tistical or research hypothesis, there is a sug-
note that directional hypotheses have several gested relationship among the hypothesis, the
advantages that make them appropriate for use in research question, the research design of the
most studies: study, and the level of evidence provided by
Directional hypotheses indicate that a theory the results of the study. The type of design,
base was used to derive the hypotheses and experimental or nonexperimental, in uences the
that the phenomena under investigation have wording of the hypothesis. For example, when
been critically examined and interrelated. an experimental design is used, the research
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 81

consumer would expect to see hypotheses that

re ect relationship statements, such as the DEVELOPING AND REFINING A CLINICAL
following: QUESTION: A CONSUMER’S PERSPECTIVE
1 is more effective than 2 on Practising nurses, as well as students, are chal-
The effect of 1 on is greater than that of lenged to keep their practice up to date by search-
2 on ing for, retrieving, and critiquing research articles
The incidence of will not differ in partici- that apply to practice issues that they encounter
pants receiving 1 and 2 treatments. in their clinical setting (Cullum, 2000). Practitio-
The incidence of will be greater in par- ners strive to use the current best evidence from
ticipants after 1 than after 2. research in making clinical and health care deci-
Such hypotheses indicate that an experimental sions. Although research consumers are not
treatment (i.e., independent variable ) will be conducting research studies, their search for
used and that two groups of participants, experi- information from practice is also converted into
mental and control groups, are being used to test focused, structured clinical questions that are the
whether the difference in the outcome (i.e., foundation of evidence-informed practice. Clini-
dependent variable ) predicted by the hypoth- cal questions often arise from clinical situations
esis exists. Hypotheses re ecting experimental for which there are no ready answers. You have
designs also concern the effect of the experimen- probably had the experience of asking, What is
tal treatment (i.e., independent variable ) on the the most effective treatment for . . . or Why
outcome (i.e., dependent variable ). do we still do it this way
In contrast, hypotheses related to nonexperi- Framed according to criteria similar to those
mental designs re ect associative relationship related to framing a research question, focused
statements such as the following: clinical questions are used as a basis for searching
will be negatively related to the literature to identify supporting evidence
A positive relationship will exist between from research. A clinical uestion has five
and components:
Thus, in a study in which the hypotheses 1. Population
were associative relationship statements, the 2. Intervention
evidence provided by the results of that investiga- 3. Comparison
tion have level IV strength (nonexperimental 4. Outcome
design). 5. Time
The Critical Thinking Decision Path will help The five components, known as PICOT, consti-
you determine both the type of hypothesis pre- tute a format that is effective in helping nurses
sented in a study and the study’s readiness for a develop searchable clinical questions (Melnyk
hypothesis-testing design. Fineout-Overholt, 2011). Box 4-3 presents each
component of the clinical question.
The significance of the clinical question
Evidence-Informed Practice Tip becomes obvious as the research evidence from
Think about the relationship between the wording the literature is critiqued. The research evidence
of the hypothesis, the type of research design sug- is used side by side with clinical expertise and the
gested, and the level of evidence provided by the find-
ings of a study with each kind of hypothesis. The patient’s perspective to develop or revise nursing
research consumer may want to consider which type of standards, protocols, and policies that are used to
hypothesis potentially will yield the strongest results plan and implement patient care (Cullum, 2000
applicable to practice.
Sackett, Straus, Richardson, Rosenberg,
82 PART ONE Research Overview

CRITICAL THINKING DECISION PATH

Determining the Type of Hypothesis or Readiness for Hypothesis Testing

Examine the literature review and theoretical framework

Assess the primary purpose of the study and the explicit

or implied research question

Is the primary purpose Is the primary purpose

exploratory, descriptive, testing causal or associative
or hypothesis generating? relationships?

Yes No Yes No

Formulate Formulate Assess

research hypothesis for presence
question(s) (hypotheses) of IV and DV,
Research Statistical
a predictive
or scientific or null
statement,
hypothesis hypothesis
testability, and
(hypotheses) (hypotheses)
appropriate
theory
base

Directional Nondirectional
hypothesis hypothesis

DV, dependent variable; IV, independent variable.

 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 83

BOX 4-3 TABLE 4-6

PICOT COMPONENTS OF A CLINICAL QUESTION ELEMENTS OF A CLINICAL QUESTION
COMPARISON
Population: The individual patient or group of patients
POPULATION INTERVENTION INTERVENTION OUTCOME
with a particular condition or health care problem
(e.g., adolescents aged 13 to 18 with type 1 People with Pain diaries No pain Increased
insulin-dependent diabetes) advanced diaries pain
Intervention: The particular aspect of health care that cancer control
is of interest to the nurse or the health team; for
example, a therapeutic intervention (inhaler or
nebulizer for treatment of asthma), a preventive
intervention (pneumonia vaccine), a diagnostic
intervention (measurement of blood pressure), or
an organizational intervention (implementation of The situation is the patient or problem being
a bar coding system to reduce medication errors) addressed. This can be a single patient or a
Comparison intervention: Standard care or no group of patients with a particular health
intervention (e.g., antibiotic or ibuprofen for
children with otitis media); a comparison of two problem (palliative care of patients with
treatment settings (e.g., rehabilitation centre or cancer).
home care) The intervention is the dimension of
Outcome: Improved outcome (e.g., improved
glycemic control, decreased hospitalizations, health care interest, and the question is
decreased medication errors) often about whether a particular interven-
Time: time involved to demonstrate an outcome (e.g., tion (in this case, pain diaries) is a useful
weight loss maintained over a period of 2 years)
treatment.
The outcome encompasses the effect of the
treatment (intervention) for this patient or
Practical Application the patient population in terms of quality
With regard to the example of pain, albeit (e.g., decreased pain perception) and cost
from a different perspective, a nurse working
in a palliative care setting wondered whether (low cost). It essentially answers whether
completing pain diaries was useful for patients with the intervention makes a difference for the
advanced cancer who were receiving palliative care. patient population.
She wondered whether they were spending time
developing something that had previously been The individual parts of the question are vital
shown to be useless or even harmful. After all, it is pieces of information to remember when you
conceivable that monitoring one’s pain in a diary search for evidence in the literature. One of the
actually heightens one’s awareness and experience of
pain. To focus her search of the literature, the nurse
easiest ways to do this is to use a table, such as
developed the following question: “Does the use of Table 4-6. Examples of clinical questions are
pain diaries in the palliative care of patients with highlighted in Box 4-4. Chapter 5 provides exam-
cancer lead to improved pain control?” ples of how to effectively search the literature to
find answers to questions posed by researchers
Haynes, 2000 Thompson, Cullum, McCaughan, and research consumers.
Sheldon, Raynor, 2004). Issues or questions
can arise from multiple clinical and managerial
situations (see Practical Application box).
Evidence-Informed Practice Tip
Sometimes it is helpful for nurses who develop
You should formulate clinical questions that arise
clinical questions from a consumer’s perspective from your clinical practice. Once you have developed a
to consider three elements—(1) the situation, (2) focused clinical question by using the PICOT format,
the intervention, and (3) the outcome—as they search the literature for the best available evidence to
answer your clinical question.
frame their focused question:
84 PART ONE Research Overview

BOX 4-4
EXAMPLES OF CLINICAL QUESTIONS
• In overweight or obese people with type 2 diabetes, • What is the effect of arch supports on balance,
does an intensive lifestyle intervention reduce weight functional mobility, back pain, and lower extremity
and cardiovascular disease risk factors? (Look AHEAD joint pain in older adults? (Mulford, Taggart, Nivens,
Research Group, Pi-Sunyer, Blackburn, et al., 2007) & Payrie, 2008)
• Is diet, exercise, or both effective for weight • Is there a significant difference in the effect of
reduction in postpartum women? (Amorim, Linne, different body positions on blood pressure in healthy
& Lourenco, 2007) young adults? (Eser, Korshid, Gunes, & Demir, 2007)
• In patients who require mechanical ventilation for • In people with impaired glucose tolerance, do
longer than 48 hours, is oral decontamination with lifestyle or pharmacological interventions prevent or
chlorhexidine or with chlorhexidine plus colistin delay onset of type 2 diabetes? (Gillies, Abrams,
effective for reducing the incidence of ventilator- Lambert, Cooper, Sutton, Hsu, & Khunti, 2007)
associated pneumonia? (Koeman, van der Ven, • What are the experiences of middle-aged people
Hak, Kaasjager, de Smet, Dormans, . . . Benton, living with chronic heart failure? (Nordgren, Asp, &
2006) Fagerberg, 2007)

study). For example, the significance may be to

CRITIQUING THE RESEARCH QUESTION answer a question encountered in the clinical area
The Critiquing Criteria box on p. 87 provides and thereby improve patient care, to resolve a
several criteria for evaluating this initial phase of con ict in the literature regarding a clinical issue,
the research process: the research question. or to provide data supporting an innovative form
Because the research question represents the of nursing intervention that is more effective and
basis for the study, it is usually introduced at the is also cost effective.
beginning of the research report to indicate Sometimes readers find that the research ques-
the focus and direction of the study. Readers are tion is not clearly stated at the conclusion of the
then in a position to evaluate whether the rest introduction. In some cases, the author only hints
of the study logically pertains to this basis. The at the research question, and the reader is chal-
author often begins by identifying the background lenged to identify it. In other cases, the author
and significance of the issue that led to crystalliz- embeds the research question in the introductory
ing development of the unanswered question. The text or purpose statement. To some extent, where
clinical and scientific background, significance, or whether the author states the research question
or both are summarized, and the purpose, aim, or depends on the style of the journal. Nevertheless,
objective of the study is identified. Finally, the the evaluator must remember that the main
research question and any related subquestions research question should be implied if it is not
are proposed before or after the literature review. clearly identified in the introductory section—
The purpose of the introductory summary of even if the subquestions are not stated or implied.
the theoretical and scientific background is to When critiquing the research question, the
provide the reader with a contextual glimpse of reader looks for the presence of the three key
how the author critically thought about the devel- elements described on p. 171:
opment of the research question. The introduction Does the research question express a rela-
to the research question places the study within tionship between two or more variables or,
an appropriate theoretical framework and begins at least, between an independent variable
the description of the study. This introductory and a dependent variable
section should also include the significance of the Does the research question specify the
study (i.e., why the investigator is conducting the nature of the population being studied
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 85

Does the research question imply the pos- strengths and weaknesses of the hypotheses in a
sibility of empirical testing research report:
You will use these three elements as criteria 1. When reading a research study, you may
for judging the soundness of a stated research find the hypotheses clearly delineated in a
question. If the variables, the population, and the separate hypothesis section of the research
implications for testability are unclear, then the article (i.e., after the literature review or
remainder of the study will probably falter. For theoretical framework section or sections).
example, a research study on anxiety during the In many cases, the hypotheses are not
perioperative period contained introductory mate- explicitly stated and are only implied in the
rial on anxiety in general, anxiety as it relates to results or discussion section of the article.
the perioperative period, and the potentially ben- In such cases, you must infer the hypotheses
eficial in uence of nursing care in relation to from the purpose statement and the type of
anxiety reduction. The author concluded that the analysis used. You should not assume that
purpose of the study was to determine whether if hypotheses do not appear at the beginning
selected measures of patient anxiety could be of the article, they do not exist in the par-
shown to vary when different approaches to ticular study. Even when hypotheses are
nursing care were used during the perioperative stated at the beginning of an article, they are
period. The author did not state the research ques- re-examined in the results or discussion
tions. A restatement of the problem in question section as the findings are presented and
form might be as follows: discussed.
2. If a hypothesis or set of hypotheses is pre-
sented, the data analysis should answer the
What is the difference in patient anxiety level in hypotheses directly. Because the hypothesis
relation to different approaches to nursing care should re ect the culmination and ex-
during the perioperative period pression of this conceptual process, its
If this process of developing a research ques- placement in the research report logically
tion is clarified at the outset of a research study, follows the literature review and the theo-
the report that follows can develop logically. retical framework discussion. It should be
Readers will have a clear idea of what the report consistent with both the literature review
should convey and can knowledgeably evaluate and the theoretical framework.
the material that is presented. When you critically 3. Although a hypothesis can legitimately be
appraise clinical questions, remember that they nondirectional, it is preferable, and more
should be focused and specify the patient or common, for the researcher to indicate the
problem being addressed, the intervention, and direction of the relationship between the
the outcome for a particular patient population. variables in the hypothesis. You will find
The author should provide evidence that the clini- that when data for the literature review are
cal question guided the literature search and that unavailable (i.e., the researcher has chosen
the question suggests the design and level of evi- to study a relatively undefined area of inter-
dence to be obtained from the study findings. est), a nondirectional hypothesis may be
appropriate. Enough information simply
CRITIQUING THE HYPOTHESES may not be available for making a sound
As illustrated in the Critiquing Criteria box, judgement about the direction of the pro-
several criteria for critiquing the hypotheses posed relationship. All that can be proposed
should be used as a standard for evaluating the is that there will be a relationship between
86 PART ONE Research Overview

two variables. Essentially, you will want to should immediately be suspicious of

determine the appropriateness of the hypotheses that are not stated objectively.
researcher’s choice regarding directionality 6. You should recognize that how the pro-
of the hypothesis. posed relationship of the hypothesis is
4. The notion of testability is central to the phrased suggests the type of research design
soundness of a hypothesis. One criterion that is appropriate for the study, as well as
related to testability is that the hypothesis the level of evidence to be derived from the
should be stated in such a way that it can findings. For example, if a hypothesis pro-
be clearly supported or dismissed. Although poses that treatment 1 will have a greater
this criterion is very important to keep in effect on than treatment 2, an experimen-
mind, you should also understand that tal (level II evidence) or quasiexperimental
ultimately theories or hypotheses are never design (level III evidence) is suggested. If
proved beyond a doubt through hypothesis a hypothesis proposes that there will be a
testing. Claims that certain data have positive relationship between variables
proved the validity of their hypothesis and , a nonexperimental design (level IV
should be regarded with grave reservation. evidence) is suggested. Table 4-4 contains
At best, findings that support a hypothesis additional examples of hypotheses, the type
are considered tentative. If repeated repli- of research design, and the level of evidence
cation of a study yields the same results, that is suggested by each hypothesis. The
more confidence can be placed in the con- design and level of evidence have important
clusions advanced by the researchers. It implications for the remainder of the study
is important to remember about testability in terms of the appropriateness of sample
that although hypotheses are more likely to selection, data collection, data analysis, inter-
be accepted with increasing evidence, they pretation of findings, and—ultimately—
are ultimately never proved. the conclusions advanced by the researcher.
5. Another point about testability to consider 7. If the research report contains research
is that the hypothesis should be objectively questions rather than hypotheses, you will
stated and devoid of any value-laden words. want to evaluate whether this is appropriate
Value-laden hypotheses are not empirically for the study. One criterion for making this
testable. uantifiable phrases—such as decision, as presented earlier in this chapter,
greater than less than decrease is whether the study is of an exploratory, a
increase positively related nega- descriptive, or a qualitative nature. If it is,
tively related and related —convey the then it is appropriate to have research ques-
idea of objectivity and testability. You tions rather than hypotheses.

APPRAISING THE EVIDENCE

The Research Question and Hypotheses
The care taken by a researcher when developing the objective is to answer the research question. This task
research question or hypothesis is often representative of may be a time-consuming, sometimes frustrating endea-
the overall conceptualization and design of the study. A vour for the researcher, but in the final analysis, the
methodically formulated research question provides the outcome, as evaluated by the consumer, is most often
basis for hypothesis development. In a quantitative worth the struggle. Because this text focuses on the nurse
research study, the remainder of the study revolves as a critical consumer of research, the following sections
around testing the hypothesis or, in some cases, the pertain primarily to the evaluation of research questions
research question. In a qualitative research study, the and hypotheses in published research reports.
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 87

CRITIQUING CRITERIA
THE RESEARCH QUESTION appropriate theoretical 5. Is each of the hypotheses
1. Is the research question framework? specific to one relationship so
introduced promptly? 7. Has the significance of the that each hypothesis can be
2. Is the question stated clearly research question been either supported or not
and unambiguously in identified? supported?
declarative or question form? 8. Have pragmatic issues, such as 6. Is the hypothesis stated in such
3. Does the research question feasibility, been addressed? a way that it is testable?
express a relationship between 9. Have the purpose, aims, or 7. Is the hypothesis stated
two or more variables or at goals of the study been objectively, without value-laden
least between an independent identified? words?
variable and a dependent 8. Is the direction of the
variable, thereby implying its THE HYPOTHESES relationship in each hypothesis
empirical testability? 1. Is the hypothesis related clearly stated?
4. Does the research question directly to the research 9. Is each hypothesis consistent
specify the nature of the question? with the literature review?
population being studied? 2. Is the hypothesis stated 10. Is the theoretical rationale for
5. Has the research question been concisely in a declarative form? the hypothesis explicit?
substantiated by adequate 3. Are the independent and 11. Are research questions
experiential and scientific dependent variables identified appropriately used (i.e., for an
background material? in the statement of the exploratory, descriptive, or
6. Has the research question been hypothesis? qualitative study or in relation
placed within the context of an 4. Are the variables measurable or to ancillary data analyses)?
potentially measurable?

CRITICAL THINKING CHALLENGES KEY POINTS

■ Discuss how the wording of a research question • Formulation of the research question and stating
or hypothesis suggests the type of research the hypothesis are key preliminary steps in the
research process.
design and level of evidence that will be • The research question is refined through a process
provided. that proceeds from the identification of a general
■ Using the study about decisional involvement of idea of interest to the definition of a more specific
senior nurse leaders by Wong and associates and circumscribed topic.
(2010) in Appendix B, diagram how the • A preliminary literature review reveals related
factors that appear to be critical for the research
hypotheses flow from the theoretical framework topic of interest and helps further define the
and literature review. research questions.
■ Using Sobieraj and colleagues’ (2009) study in • The significance of the research question must be
Appendix C, describe how the significance of the identified in terms of its potential contribution to
research problem and purpose of the study are patients, nurses, the medical community in
general, and society. The applicability of the
linked to the research objectives. question for nursing practice and its theoretical
■ A nurse is caring for patients in a clinical relevance must be established. The findings should
situation that produces a clinical question that also have the potential for formulating or altering
has no ready answer. The nurse wants to nursing practices or policies.
develop and refine this clinical question by using • The feasibility of a research question must be
examined in light of pragmatic considerations: for
the PICOT approach so that it becomes the basis example, time; the availability of participants,
for an evidence-informed practice project. How money, facilities, and equipment; the nurse’s
can the nurse accomplish that objective? experience; and ethical issues.
Continued
88 PART ONE Research Overview

• The final research question consists of a statement with theory. The appropriateness of the hypothesis
about the relationship of two or more variables. in relation to the type of research design is also
The question clearly identifies the relationship examined. In addition, the appropriate use of
between the independent variables and dependent research questions is evaluated in relation to the
variables, specifies the nature of the population type of study conducted.
being studied, and implies the possibility of
empirical testing.
• Focused clinical questions arise from clinical
practice and guide the literature search for the best REFERENCES
available evidence to answer the clinical question.
Amorim, A. R., Linne, Y. M., Lourenco, P. M. C.
• A hypothesis is an attempt to answer the research
question. When the validity of the assumptions of (2007). Diet or exercise or both, for weight reduction
the theoretical framework is tested, the hypothesis in women after childbirth. Cochrane Database of
connects the theory and reality. Systematic Revie s, ( ), CD005627
• A hypothesis is a declarative statement about the Bailey, J. J., Sabbagh, M., Loiselle, C. G., Boileu, J.,
relationship between two or more variables in McVey, L. (2010). Supporting families in the IC : A
which an expected outcome is predicted. The descriptive correlational study of informational
characteristics of a hypothesis include a support, anxiety, and satisfaction with care. Intensive
relationship statement, implications regarding Critical Care Nursing, (2), 114-122.
testability, and consistency with a defined theory Boscart, V. M. (2009). A communication intervention
base.
for nursing staff in chronic care. ournal of Advanced
• Hypotheses can be formulated directionally or
nondirectionally. Hypotheses can be further Nursing, (9), 1823-1832. doi: 10.1111/j.1365-
categorized as either research or statistical (null) 2648.2009.05035.x
hypotheses. Cullum, N. (2000). ser’s guides to the nursing
• Research questions may be used instead of literature: An introduction. Evidence-Based Nursing,
hypotheses in exploratory, descriptive, or (2):71-72.
qualitative research studies. Research questions de Witt, L., Ploeg, J., Black, M. (2010). Living alone
may also be formulated in addition to hypotheses with dementia: An interpretative phenomenological
to answer questions related to ancillary data. study with older women. ournal of Advanced
• The purpose, research question, or hypothesis Nursing, (8), 1698-1707. doi: 10.1111/j.1365-
provides information about the intent of the
2648.2010.05295.x
research question and hypothesis and suggests the
level of evidence to be obtained from the study Eser, I., Korshid, L., Gunes, . Y., Demir, Y. (2007).
findings. The effect of different body positions on blood
• The critiquing criteria are a set of guidelines for pressure. ournal of Clinical Nursing, (1):137-140.
evaluating the strengths and weaknesses of the Gillies, C. L., Abrams, K. R., Lambert, P. C., Cooper,
research question and hypotheses as they appear N. J., Sutton, A. J., Hsu, R. T., Khunti, K. (2007).
in a research report. Pharmacological and lifestyle interventions to prevent
• In critiquing, the reader assesses the clarity of the or delay type 2 diabetes in people with impaired
research question and the related subquestions, glucose tolerance: Systematic review and meta-
the specificity of the population, and the analysis. British Medical ournal, (7588), 299.
implications for testability.
doi: 10.1136/bmj.39063.689375.55
• The interrelatedness of the research question, the
literature review, the theoretical framework, and Harrowing, J. N., Mill, J. (2010). Moral distress
the hypotheses should be apparent. among gandan nurses providing HIV care: A critical
• The appropriateness of the research design ethnography. International ournal of Nursing
suggested by the research question is also Studies, , 723-731.
evaluated. Ives, M., Melrose, S. (2010). Immunizing children
• The purpose of the study (i.e., why the researcher who fear and resist needles: Is it a problem for
is conducting the study) should be differentiated nurses Nursing orum, (1), 29-39.
from the research question. Keefe, M. R., Kajrlsen, K. A., Lobo, M. L., Kotzer,
• The reader evaluates the wording of the hypothesis A. M., Dudley, W. N. (2006). Reducing parenting
in terms of the clarity of the relational statement,
stress in families with irritable infants. Nursing
its implications for testability, and its congruence
Research, (3), 198-205.
 CHAPTER 4 Developing Research Questions, Hypotheses, and Clinical Questions 89

Koeman, M., van der Ven, A. J., Hak, E., Kaasjager, K., Purc-Stephenson, R. J., Thrasher, C. (2010). Tele-
de Smet, A. G. A., Dormans, T. P. J., . . . Bonton, phone triage and advice: A meta-ethnography.
M. J. M. (2006). Oral decontamination with chlorhex- ournal of Advanced Nursing, (3), 482-494. doi:
idine reduces the incidence of ventilator-associated 10.1111/j.1365-2648.2010.05275.x
pneumonia. American ournal of Respiratory and Sackett, D. L., Straus, S. E., Richardson, W. S.,
Critical Care, , 1348-1355. Rosenberg, W., Haynes, R. B. (2000). Evidence-
Look AHEAD Research Group, Pi-Sunyer, ., Black- based medicine: o to practise and teach EBM.
burn, G., et al. (2007). Reduction in weight and London: Churchill Livingstone.
cardiovascular risk factors in overweight and obese Simonson, D. C., Ahern, M. M., Hendryx, M. S.
people with type 2 diabetes: One-year results of (2007). Anesthesia staffing and anesthetic complica-
the Look AHEAD trial. Diabetes Care, , tions during cesarean delivery: A retrospective
1374-1383. analysis. Nursing Research, (1), 9-17.
Martinez, A. M., D’Artois, D., Rennick, J. E. (2007). Sobieraj, G., Bhatt, M., LeMay, S., Rennick, J.,
Does the 15-minute (or less) family interview Johnston, C. (2009). The effect of music on parental
in uence family nursing practice ournal of amily participation during pediatric laceration repair.
Nursing, (2), 157-178. Canadian ournal of Nursing Research, (4), 68-82.
Melnyk, B. M., Fineout-Overholt, E. (2011). Streubert, H. J., Carpenter, D. R. (2011). ualitative
Evidence-based practice in nursing and healthcare: research in nursing: Advancing the humanistic
A guide to best practice (2nd ed). New York: Wolters imperative (5th ed.). Philadelphia: Wolters Kluwer.
Kluwer. Thompson, C., Cullum, N., McCaughan, D., Sheldon,
Meneses, K. D., McNees, P., Loerzei, V. W., Su, ., T., Raynor, P. (2004) Nurses, information use, and
hang, Y., Hassey, L. A. (2007). Transition from clinical decision making: The real world potential for
treatment to survivorship: Effects of a psychoeduca- evidence-based decisions in nursing. Evidence-Based
tional intervention on quality of life in breast cancer Nursing, (3):68-72. doi:10.1136/ebn.7.3.68
survivors. ncology Nursing orum, (5), Thorne, S., Armstrong, E., Harris, S. R., et al. (2009).
1007-1016. Patient real-time and 12-month retrospective
Mulford, D., Taggart, H. M., Nivens, A., Payrie, C. perceptions of difficult communications in the cancer
(2008). Arch support use for improving balance and diagnostic period. ualitative ealth Research,
reducing pain in older adults. Applied Nursing (10), 1983-1394. doi: 10.1177/1049732.309348382
Research, (3), 153-158. Wong, C. A., Laschinger, H., Cummings, G. G.,
Nordgren, L., Asp, M., Fagerberg, I. (2007). Living Vincent, L., O’Connor, P. (2010). Decisional
with moderate-severe chronic heart failure as a involvement of senior nurse leaders in Canadian
middle-aged person. ualitative ealth Research, acute care hospitals. ournal of Nursing Management,
(1), 4-13. , 122-133. doi: 10.1111/j.1365-2834.2010.01053.x
O’Donnell, S., MacIntosh, J., Wuest, J. (2010). A
theoretical understanding of sickness absences among
women who have experienced workplace bullying.
FOR FURTHER STUDY
ualitative ealth Research (4), 439-452. Go to Evolve at http://evolve.elsevier.com/
Peterson, W. E., Sword, W., Charles, C., DiCenso, A. Canada/LoBiondo/Research for Audio Glossary, how-to
(2007). Adolescents’ perceptions of inpatient instructions for Writing Proposals for Funding, and
postpartum nursing care. ualitative ealth additional research articles for practice in reviewing
Research, (2), 201-212. and critiquing.
C H A PTER 5

Finding and Appraising the

Literature
Stephanie Fulton | Barbara Krainovich-Miller | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Discuss the relationship of the literature review to nursing theory, research, education, and
practice.
• Discuss the purposes of the literature review from the perspective of the research investigator and
the research consumer.
• Discuss the use of the literature review for quantitative designs and qualitative methods.
• Discuss the purpose of reviewing the literature in development of evidence-informed practice.
• Differentiate between primary and secondary sources.
• Compare the advantages and disadvantages of the most commonly used online databases for
conducting a literature review.
• Identify the characteristics of an effective electronic search of the literature.
• Critically read, appraise, and synthesize primary and secondary sources used for the development
of a literature review.
• Apply critiquing criteria to the evaluation of literature reviews in selected research studies.

KEY TERMS
Boolean operator literature review refereed (peer-reviewed) journal
citation management software online database secondary sources
concept operational definition theory
conceptual definition primary sources Web browser
controlled vocabulary print indexes

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

90
 CHAPTER 5 Finding and Appraising the Literature 91

YOU MAY WONDER WHY AN ENTIRE chapter of a The conceptual framework, or theoretical
research text is devoted to finding and appraising framework, of a research report is a structure of
the literature. The main reason is that searching concepts or theories pulled together as a map for
for, retrieving, and critically appraising the litera- the study this map provides rationale for the
ture is a key step in the research process for development of research questions or hypotheses.
researchers and also for nurses involved in This section of a research report is often a titled
evidence-informed practice. A more personal subsection of the literature review and may be
question you might ask is Will knowing more accompanied by a diagram illustrating the pro-
about how to critically appraise and gather the posed relationships between and among the con-
literature really help me as a student or later as a cepts. Alternatively, the conceptual/theoretical
practising nurse The answer is that it most cer- framework may not be separately identified it
tainly will Your ability to locate and retrieve may be embedded in the literature review section
research studies, critically appraise them, and of an article or simply not included. The links
decide that you have the best available evidence between theory, research, education, and practice
to inform your clinical decision making is a skill are intricately connected together they create the
essential for your current role as a student and knowledge base for the nursing discipline, as
your future role as a nurse who is a competent shown in Figure 5-1.
research consumer. The purpose of this chapter is to introduce you
Your critical appraisal, also called a criti ue of to the literature review as it is used in research
the literature, is an organized systematic approach and evidence-informed practice projects. It pro-
to evaluating a research study or group of research vides you with the systematic tools to (1) consider
studies. It involves the use of a set of established how the theoretical or conceptual framework
critical appraisal criteria to objectively determine guides development of a research study (2)
the strength, quality, and consistency of evidence critically appraise a research study or group
these characteristics help you determine the of research studies (3) locate, search for, and
applicability of the evidence to research, educa- retrieve research studies, systematic reviews,
tion, or practice. As a research consumer, you will documents, and statistical reports and (4) dif-
become skilled at critically appraising research ferentiate between a research article and a
studies, combining the evidence with your clini- conceptual article or book. This set of tools will
cal experience and the patient population that you help you develop your research consumer
are caring for, to make an evidence-informed
decision about the applicability of a particular
nursing intervention for your patient or for the
patient population in your practice setting.
Research
The section of a published research report
titled Literature Review generally appears near
the beginning of the report. It provides an abbre- Review
Education Practice
viated version of the complete literature review of
literature
conducted by a researcher and represents the
foundation for the study. Therefore, the literature
Theory
revie a systematic and critical appraisal of the
most important literature on a topic, is a key step
in the research process that provides the basis of FIGURE 5-1 Relationship of the literature review to theory,
a research study. research, education, and practice.
92 PART ONE Research Overview

competencies and prepare your academic papers p. 530 in Appendix B ) of this article is a con-
and evidence-informed practice projects. ceptual framework diagram that depicts the par-
ticipation in organizational decision making and
THE CONCEPTUAL OR THEORETICAL identifies the major variables of the study.
FRAMEWORK As you can see, theory, practice, and research
As discussed in Chapter 2, the conceptual frame- are interconnected: Practice enables testing of
work, or theoretical framework, of a research theory and generates research questions research
report is a structure of concepts or theories that contributes to theory building and provides sup-
provides the basis for development of research porting evidence for effective nursing inter-
questions or hypotheses. A conce t is an image or ventions in clinical practice. Therefore, what is
a symbolic representation of an abstract idea. The learned through practice, theory, and research is
researcher uses a concept, a set of concepts, or a combined to create the knowledge of the disci-
particular theory or set of theories to build the pline of nursing, which you are then taught in
theoretical framework of the study. Concepts are your nursing courses.
the major components of theory and convey the
abstract ideas within a theory. A theory is a set of REVIEW OF THE LITERATURE
interrelated concepts, definitions, and proposi-
tions that convey a systematic view of phenomena The Literature Review: The Researcher’s
for the purpose of explaining and making predic- Perspective
tions about those phenomena. A conce tual de The overall purpose of the literature review in a
nition includes the general meaning of a concept. research study is to present a strong knowledge
An o erational de nition is a description of the base for the conduct of the research study. Spe-
method used to measure the concept once the cific objectives 1 through 7 listed in Box 5-1
concept is linked to a measurement method or re ect the purposes of a literature review for the
instrument, it is regarded as a variable. conduct of quantitative research and most qualita-
For example, Wong and colleagues (2010), tive research. It is important to understand when
investigating the scope and degree of involve- you read a research article that the researcher’s
ment of senior nurse leaders in executive level main goal when developing the literature review
decisions in acute care organizations across was to develop the knowledge foundation for a
Canada (p. 122), used a theoretical framework sound study and to generate research questions
adapted from previous work on strategic decision and hypotheses.
making in health care organizations to guide the An extensive literature review is essential for
development of the study. The content related to all steps of the quantitative research process and
the theoretical framework is identified in a sepa- for some qualitative methods. From this perspec-
rately titled section of the article and follows the tive, the review is broad and systematic, as well
presentation and critical appraisal of the literature as in-depth. It is a critical collection and evalua-
on senior nurse role changes with restructuring tion of the important published literature in jour-
and in organizational decision making (see nals, monographs, books, and book chapters, as
Appendix B). well as unpublished research print and online
As described in the introduction to this chapter, materials (e.g., doctoral dissertations and masters’
the theoretical or conceptual framework of a theses), audiovisual materials (e.g., audiotapes
research study is often illustrated with a diagram. and videotapes), and sometimes personal com-
For example, in the study by Wong and col- munications (e.g., conference presentations and
leagues (2010) (see Appendix B), Figure B-1 (see one-on-one interviews).
 CHAPTER 5 Finding and Appraising the Literature 93

BOX 5-1 steps of the quantitative research process as

OVERALL PURPOSES OF A LITERATURE REVIEW follows:
Theoretical or conceptual frame or : A lit-
MAJOR GOAL
erature review reveals conceptual tradi-
To develop a strong knowledge base to carry out a
research study or an evidence-informed practice tions, concepts, theories, or conceptual
project models from nursing and other related dis-
OBJECTIVES ciplines that can be used to examine prob-
A review of the literature helps you do the following: lems. This framework presents the context
1. Determine what is known and unknown about for studying the problem and can be viewed
a subject, concept, or problem as a map for understanding the relationships
2. Determine gaps, consistencies, and
inconsistencies in the literature about a subject, between or among the variables in quantita-
concept, or problem tive studies. The literature review provides
3. Discover conceptual traditions used to examine rationale for the variables and explains con-
problems
cepts, definitions, and relationships between
4. Generate useful research questions and
hypotheses or among the independent and dependent
5. Determine an appropriate research design, variables used in the theoretical framework
methodology, and analysis for answering the of the study.
research questions or hypotheses on the basis
of an assessment of the strengths and Primary and secondary sources: The author
weaknesses of earlier works of a literature review should use mainly
6. Determine the need for replication of a study or ri ary sources that is, articles and books
refinement of a study
7. Synthesize the strengths and weaknesses and by the original author. Sometimes it is
findings of available studies on a topic or appropriate to use secondary sources
problem which are published articles or books that
8. Uncover one or more new practice
are written by persons other than the indi-
interventions or obtain supporting evidence for
revising or maintaining current interventions, vidual who developed the theory or con-
protocols, and policies ducted the research study. The studies
9. Promote evidence-informed revision and selected for the literature review should
development of new practice protocols,
policies, and projects or activities related to offer the strongest and most consistent level
nursing practice of evidence available on the topic (Table
10. Generate clinical questions that guide 5-1 lists examples).
development of evidence-informed practice
projects Research uestion and hypothesis: The lit-
erature review helps you determine what is
known and not known uncover gaps, con-
sistencies, or inconsistencies or to disclose
From a researcher’s perspective, the objectives unanswered questions in the literature about
in Box 5-1 direct the questions the researcher asks a subject, concept, theory, or problem that
while reading the literature to determine one or generate or allow for refinement of research
more useful research questions or hypotheses and questions, hypotheses, or both.
how best to design a particular study. Design and method: The literature review
The following brief overview about the use of exposes the strengths and weaknesses of
the literature review in relation to the steps of the previous studies in terms of designs and
quantitative and qualitative research process will methods and helps the researcher choose
help you to understand the researcher’s focus. A an appropriate new, replicated, or refined
critical review of relevant literature affects the design, including data-collection method,
94 PART ONE Research Overview

TABLE 5-1
PRIMARY AND SECONDARY SOURCES
PRIMARY: ESSENTIAL SECONDARY: USEFUL

Material written by the original person who Material written by one or more individuals other than the
conducted the study, developed the theory person who conducted the research study or developed a
(model), or prepared the scholarly discussion on theory; the author is someone other than the original author
a concept, topic, or issue of interest (i.e., the who writes about or presents the original author’s work. The
original author). material is usually in the form of a summary or critique (i.e.,
analysis and synthesis) of someone else’s scholarly work or
body of literature.
Primary sources can be published or unpublished. Secondary sources can be published or unpublished.
Research example: An investigator’s report of his or Secondary source examples are the following: response,
her research study (e.g., articles in Appendixes A commentary, or critique articles of a research study, a theory
through D). or model, or a professional view of an issue; review of
literature article published in a refereed scholarly journal;
abstracts of a published work written by someone other than
the original author; examples: a biography or a systematic
review.
Theoretical example: Senior nurse leaders’ Hint: Use secondary sources sparingly; however, secondary
participation in organizational decision making is sources—especially of studies that include a research
the theoretical framework used by Wong et al. critique—are a valuable learning tool for a beginning
(2010) in their study of the scope and degree of research consumer.
senior nurse leaders’ contributions to executive-
level decisions. The theoretical framework used
in this study was adapted from the work of
Ashmos, Huonker, and McDaniel (1998) and
Anderson and McDaniel (1998) and cited as such
in the research report (see Appendix B).
Other primary source examples include
autobiographies, diaries, films, letters, artifacts,
periodicals, and tapes.
Hint: Critical evaluation of mainly primary sources
is essential in a thorough and relevant review of
the literature.

sampling strategy and size, valid and reli- theoretical articles or books described
able measurement instruments, an effective earlier in the article in the literature review
data analysis method, and appropriate and uses this conceptual and research litera-
informed consent forms. Often, because of ture to interpret and explain the study’s
journal space limitations, researchers only findings. For example, in the Discussion
include abbreviated information about these section of their article, Wong and colleagues
aspects in their journal article. (2010) commented that Although there is
utcome of the analysis i e , ndings, dis- some indication in the healthcare literature
cussion, implications, and recommenda- that increased involvement in organiza-
tions : The literature review is used to help tional decision-making by physicians and
the researcher accurately interpret and registered nurses was associated with out-
discuss the results/findings of a study. In the comes such as lower costs in hospitals
discussion section of a research article, the (Ashmos, Huonker, McDaniel, 1998)
researcher refers to the research studies and and improved resident outcomes in nursing
 CHAPTER 5 Finding and Appraising the Literature 95

Theoretical framework for nurses involved in evidence-informed prac-

tice projects.
Recommendations Problem/need/
significance As a student or practising nurse, you may be
asked to generate a clinical question for an
evidence-informed practice project and search
Implications Review of Question/
literature hypothesis for, retrieve, review, and critically appraise the
literature to identify the best available evidence
that provides the answer to a clinical question and
informs your clinical decision making. A clear
Findings Design/methodology
and precise articulation of a question is crucial
FIGURE 5-2 Relationship of the review of the literature to for finding the best evidence. Evidence-informed
the steps of the quantitative research process.
questions may sound like research questions, but
they are questions used to search the existing
literature for answers. The evidence-informed
homes (Anderson McDaniel, 1999), practice process follows the PICOT format to
we found that SNL decision involvement generate clinical questions. For example, students
(timing, breadth and number of activities) in an adult health course were asked to generate
and in uence in decision-making did not a clinical question related to health care promo-
predict their perceived quality of organiza- tion for older women by using the PICOT format.
tional decisions (p. 130). Figure 5-2 relates As discussed in Chapter 4, The PICOT format is
the literature review to all aspects of the as follows:
quantitative research process. P Problem/patient population specifically
defined group
The Literature Review: The Consumers I Intervention what intervention or event
Perspective will be studied
From the perspective of the research consumer C Comparison of intervention with what the
rather than that of a researcher, you review a intervention will be compared
number of studies to answer a clinical question O Outcome the effect of the intervention
or to solve a clinical problem. Therefore, you T Time the time frame
search the literature widely and gather multiple An example of its use is given in the following
resources to answer your question by using an Practical Application box.
evidence-informed practice approach. This As a practising nurse you may be called on to
process includes (1) asking clinical questions (2) revise or continue current evidence-informed
identifying and gathering the evidence (3) criti- practice protocols, practice standards, or policies
cally appraising and synthesizing the evidence or in your health care organization or to develop
literature (4) acting to change practice by using new ones. This requires that you know how to
the best available evidence, coupled with your retrieve and critically appraise research articles,
clinical experience and patient preferences systematic reviews, and practice guidelines to
(values, setting, and resources) and (5) evaluat- determine the degree of support or lack of support
ing the use of the research evidence found to found in the literature. A critical appraisal of the
assess applicability of the research findings to the literature related to a specific clinical question
practice change. In addition to objectives 1 uncovers data that contribute evidence to support
through 7 in Box 5-1, objectives 7, 8, and 9 spe- current practice and clinical decision making, as
cifically re ect the purposes of a literature review well as for making changes in practice.
96 PART ONE Research Overview

major focus of reviewing the literature as a con-

Practical Application sumer is to uncover multiple sources of evidence
One group of students was interested in on a given topic that have been generated by
whether regular exercise prevented
researchers in their research studies that can
osteoporosis for postmenopausal women who had
osteopenia. The PICOT format for the clinical potentially be used to improve clinical practice
question that guided their search was as follows: and patient outcomes.
P: Postmenopausal women with osteopenia
I: Regular exercise program
C: No regular exercise program Helpful Hint
O: Prevention of osteoporosis Remember that the findings of one study on a
T: After 1 year topic do not usually provide sufficient evidence to
Their assignment required that the students do the support a change in practice; be cautious when a nurse
following: colleague tells you to change your practice on the basis
• Search the literature by using electronic of the results of one study.
databases (e.g., Cumulative Index to Nursing
and Allied Health Literature [CINAHL], via
EBSCO; MEDLINE; Scopus; and Cochrane
Database of Systematic Reviews) for the From a student perspective, the ability to criti-
background information that enabled them to cally appraise the literature is essential to acquir-
identify the significance of osteopenia and ing a skill set for successfully completing
osteoporosis as a health problem in women.
• Identify systematic reviews, practice guidelines, scholarly papers, presentations, debates, and
and individual research studies that provided evidence-informed practice projects. Both types
the “best available evidence” related to the of literature reviews are similar in that both
effectiveness of regular exercise programs on
prevention of osteoporosis. should be framed in the context of previous
• Critically appraise systematic reviews, practice research and theoretical literature and pertinent to
guidelines, and research studies in accordance the objectives presented in Box 5-1.
with standardized critical appraisal criteria (see
Chapter 20).
• Synthesize the overall strengths and weaknesses Evidence-Informed Practice Tip
of the evidence provided by the literature.
For a research consumer, formulating a clinical
• Establish a conclusion about the strength,
question provides a focus that guides the literature
quality, and consistency of the evidence.
review.
• Make recommendations about applicability of
the evidence to clinical nursing practice that
guides development of a health promotion
project about osteoporosis risk reduction for SEARCHING FOR EVIDENCE
postmenopausal women with osteopenia.
In your student role, when you are preparing an
academic paper, you read the required course
materials, as well as additional literature retrieved
from the library. Students often state, I know
Research Conduct and Consumer how to do research. Perhaps you have thought
of Research Purposes: Differences the same thing because you researched a topic
and Similarities for a paper in the library. In that situation,
How does the literature review differ when it is however, it would be more accurate for you to say
used for research purposes versus for a consumer that you have been searching the literature to
of research purposes The literature review in a uncover research and conceptual information
research study is used to develop a sound research to prepare an academic term paper on a certain
proposal for a research study that will generate topic. You search for primary sources, which are
knowledge. From a broader perspective, the articles, books, or other documents written by the
 CHAPTER 5 Finding and Appraising the Literature 97

person who conducted the study, developed the requires the same critical thinking and reading
theory, or prepared the scholarly discussion on a skills, a literature review for a research proposal
concept, topic, problem, or issue of interest. You is usually much more extensive and comprehen-
also search for secondary sources, which are sive, and the critiquing process is more in-depth.
materials written by persons other than the indi- From an academic standpoint, requirements for a
viduals who conducted a research study or devel- literature review for a particular assignment
oped a particular theory. Table 5-1 provides more differ, depending on the level and type of course,
extensive definitions and examples of primary as well as the specific objective of the assign-
and secondary sources. ment. These factors determine whether a student’s
Although reviewing the literature for research literature search requires a limited, selective
purposes and research consumer activities review or a major or extensive review. Regardless

TABLE 5-2
STEPS AND STRATEGIES FOR CONDUCTING A LITERATURE SEARCH
STEPS OF LITERATURE REVIEW STRATEGY

Step 1: Determine the clinical question or Keep focused on the characteristics of patients you deal with in your
research topic. work setting. You know what works and does not work in the delivery
of nursing care. In your student role, keep focused on the assignment’s
objective; use the literature to support opinions or develop a concept
under discussion.
Step 2: Identify the key variables/terms. Ask your reference librarian for help, and read research guidebooks,
which are usually found near the computers that are used for student
searches; include “research” as one of your variables.
Step 3: Conduct a computer search by using Conduct the search yourself or with the help of your librarian; it is
at least two recognized online databases. essential to use at least two health-related databases, such as CINAHL
via EBSCO, MEDLINE, PsycINFO, or ERIC.
Step 4: Review abstracts online and Scan through your search, read the abstracts provided, and make a note
disregard irrelevant articles. of only those that fit your topic; select “references,” as well as “search
history” and “full-text articles” if available, before printing, saving, or
e-mailing your search.
Step 5: Retrieve relevant sources. Organize by article type or study design and year and reread the abstracts
to determine whether the articles chosen are relevant and worth
retrieving.
Step 6: Print or download articles; if you are Save yourself time and money: Buy a library copying card ahead of time
unable to print directly from the or bring plenty of change so that you avoid wasting time midway to
database, you can order them through secure change; you can also bring a thumb drive to download PDF
interlibrary loan. versions of your articles.
Step 7: Conduct preliminary reading and Review critical reading strategies (see Chapter 3; e.g., read the abstract at
disregard irrelevant sources. the beginning of the articles, and see the example in this chapter).
Step 8: Critically read each source Use the critical appraisal strategies from Chapter 1 (e.g., use a
(summarize and critique each source). standardized critiquing tool), take time to type up each summary and
critical appraisal (no more than one page long), include the references
in APA style at the top or bottom of each abstract, and attach the
original article.
Invest time in learning a citation management software tool. This will
save you the hassle of formatting all of your citations.
Step 9: Synthesize critical summaries of Decide how you will present your synthesis of overall strengths and
each article. weaknesses of the reviewed articles (e.g., chronologically or according
to type: research or conceptual) and type up the synthesized material
and a reference list.
APA, American Psychological Association; CINAHL, Cumulative Index to Nursing and Allied Health Literature; ERIC, Education Resources Information
Center.
98 PART ONE Research Overview

of extent, discovering knowledge is the goal of highest-level information resource to facilitate

any search therefore, a consumer of research your search for the best evidence about your cli-
must know how to search the literature. Refer- nical question or problem (DiCenso, Bayley,
ence librarians can provide excellent help in Haynes, 2009). This model, as illustrated in
searching for various sources of scholarly litera- Figure 5-3, suggests that when searching the lit-
ture. If you are unfamiliar with the process of erature, consider prioritizing your search strategy
conducting a scholarly computer search, your ref- and begin by looking for the highest-level infor-
erence librarian can help. Table 5-2 provides you mation resource available. For example, indivi-
with steps and strategies for conducting a litera- dual original studies such as those found in
ture search. MEDLINE or CINAHL (e.g., a randomized
clinical trial) are at the lowest level of the
Prioritizing the Search for Evidence information resource pyramid. The next informa-
In Chapter 3, an evidence hierarchy is presented tion resource levels of the 6S pyramid are syn-
for grading the strength and quality of evidence opses of studies (e.g., a brief summary of a
provided by individual studies or resources high-quality study), then syntheses (e.g., Cochrane
located during a search. In this chapter, an Library), followed by synopses of syntheses
evidence-informed model, called the 6S (comprehensive summary of all the research
pyramid, is used to help you identify the related to a clinical question), and summaries

EXAMPLES OF RESOURCES

Systems Computerized decision-support systems

Summaries Evidence-informed clinical practice guidelines

Evidence-informed textbooks

Synopses of syntheses DARE; heath-evidence.ca

Evidence-informed abstraction journals

Syntheses
Systematic reviews (e.g., Cochrane Library)

Synopses of Studies Evidence-informed abstraction journals

Studies Original articles published in journals

FIGURE 5-3 The 6S levels of organization of evidence from health care research.
Adapted by permission from BMJ Publishing Group Limited. Alba DiCenso, Liz Bayley, R Brian
Haynes. (2009). “Accessing pre-appraised evidence: fine-tuning the 5S model into a 6S model.”
Evidence Based Nursing. Copyright © 2009, BMJ Publishing Group Ltd and the RCN Publishing
Company Ltd.
 CHAPTER 5 Finding and Appraising the Literature 99

(e.g., evidence-informed clinical practice guide- affects patient care If you take the time to learn
lines and textbooks. how to perform a sound database search, you will
The highest information resource level per- have the essential competency needed for your
tains to computerized decision support systems, career in nursing. The Critical Thinking Decision
a resource built into an electronic medical record Path illustrates a method for locating evidence to
that links your patient’s distinctive needs with support your research or clinical question.
current evidence-informed practice guidelines.
These computerized systems are under develop- TYPES OF RESOURCES
ment, but this does not mean that the information
found within the other information resource Print: Books, Journals, and Indexes
levels is not useful or appropriate. The 6S model Most college and university libraries have an
is a tool that can help guide your search for the online card catalog to find print and online books,
strongest and most relevant evidence-informed journals (titles only), videos and other media
information however, it does not replace the items, scripts, monographs, conference proceed-
importance of critically reading each piece of evi- ings, masters’ theses, dissertations, archival mate-
dence and assessing its quality and applicability rials, and more.
for current practice. Before the 1980s, a search was usually done
manually with rint inde es which were listings
of published material. This was a tedious and
Helpful Hint time-consuming process. The print indexes are
• Make an appointment with your educational useful today for finding sources that have not
institution’s reference librarian so you can take
advantage of his or her expertise in accessing been entered into electronic (online) databases.
electronic databases. Some of your professors might talk about the
• Take the time to set up your computer for Red Books in referring to print versions of what
electronic library access.
• If the full text of an article is unavailable through is now CINAHL. The print index started in 1956
your electronic search, read the abstract to but is no longer produced. Print resources are still
determine whether you want to order the article necessary if a search requires materials not
through interlibrary loan.
entered into an electronic database before a
certain year.

Performing an Online Database Search Print: Refereed Journals

A major portion of most literature reviews con-
Why Use an Online Database? sists of journal articles. In contrast to books and
Perhaps you are still not convinced that online textbooks, which take much longer to publish,
database searches are the best way to acquire journals are a ready source of up-to-date informa-
information for a review of the literature. Maybe tion on almost any subject. Therefore, journals
you have searched by using Google or Yahoo and are the preferred mode of communicating the
found relevant information. This is an under- most recent theory or results of a research study.
standable temptation, especially if your assign- For you as a beginning research consumer, refer-
ment requires you to use only five articles. Try to eed journals should be your first choice when
think about it from another perspective, and ask looking for theoretical, clinical, or research arti-
yourself the following question: Is this the most cles. A re ereed eer revie ed ournal has a
appropriate and efficient way to find out what is panel of internal and external reviewers who
the latest and strongest research on a topic that review submitted manuscripts for possible
100 PART ONE Research Overview

CRITICAL THINKING DECISION PATH

Search for Evidence Thought Flow

Ask an answerable
question with
PICOT elements

Synthesized or Summarized
Sources
Select best source
for evidence ACP PIER
BMJ Clinical Evidence
Cochrane Systematic Review
Evidence-informed nursing
journal Government
guidelines
Appraise and
summarize your Yes Find answer?
findings

No

Primary Sources
Select next source CINAHL
to locate evidence MEDLINE
PsycINFO

Appraise and
summarize your Yes Find answer?
findings

No

Consult your
librarian and faculty.
You may need to
consider research
project

Based on Kendall, S. American College of Physicians (ACP), ©2001, 2003, 2004, as found in Kendall, S. (2008). Evidence-based resources simplified.
Canadian Family Physician, 54(2), 241–243.
BMJ, British Medical Journal; CINAHL, Cumulative Index to Nursing and Allied Health Literature; PICOT, population, intervention, comparison, outcome,
time; PIER, Physicians’ Information and Education Resource.
 CHAPTER 5 Finding and Appraising the Literature 101

publication. The external reviewers are drawn

from a pool of nurse scholars, and possibly schol- Internet: Online Secondary
ars from other related disciplines, who are experts or Summary Databases
in various specialties. In most cases, the reviews Some databases contain more than just journal
are blind that is, the reviewers do not know article information. These online resources
who the authors of manuscript are. The reviewers contain either summaries or synopses of studies,
use a set of scholarly criteria to judge whether overviews of diseases or conditions, or a summary
a manuscript meets the publication standards of of the most recent evidence to support a particular
the journal. These criteria are similar to those treatment. For example, the Cochrane Library is
that you use when you are critically appraising an online resource that consists of six databases,
the strengths and weaknesses of a study. The including the Cochrane Database of Systematic
credibility of a published theoretical or research Reviews. sing at least two electronic health-
article is strengthened by the peer review process. related databases such as CINAHL and MEDLINE
Most refereed journals are available in print (see Box 5-2) is recommended.
and accessible electronically through your
library’s online resources. If you cannot access Internet: Online Search Engines
some research articles from your library, consider You are probably familiar with accessing a eb
the use of Google Scholar as a search engine to bro ser (a software program used to connect to
locate them. or search the World Wide Web e.g., Internet
Explorer, Mozilla Firefox, or Safari) to conduct
searches for music or other entities and using
Internet: Online Bibliographic and search engines such as Google to find information
Abstract Databases or articles. However, surfing the Web is not a
An online database is used to find journal sources good use of your time for scholarly literature
(periodicals) of research and conceptual articles searches. Table 5-3 lists sources of free online
on a variety of topics (e.g., doctoral disserta- information. Review the table carefully to deter-
tions), as well as the publications of professional mine whether it is a good source of primary
organizations and various governmental agen- research studies. Note that it includes government
cies. These sources contain bibliographic citation Web sites for such .S. organizations as the
information such as the author, title, journal, date, Health Resources and Services Administration
and indexed terms for each record. Some also and the National Institutes of Health, which are
include the abstract. Box 5-2 lists examples of the sources of health information, and some clinical
more commonly used online databases. guidelines based on systematic reviews of the
Your college or university probably enables literature, but most Web sites are not a primary
you to access such databases from your residence, source of research studies. Less common and less
whether it is or is not on campus. The most rel- used sources of scholarly material are audiotapes,
evant and frequently used source for nursing lit- videos, personal communications (e.g., letters or
erature remains CINAHL. Full text has been telephone or in-person interviews), unpublished
added to this database thus, in many cases you doctoral dissertations, master’s theses, and con-
can find the full article. Another premier resource ference proceedings.
is MEDLINE, which is produced by the National Most searches with electronic databases
Library of Medicine. MEDLINE focuses on the include not only citation information but also the
life sciences, and its sources dates back to the abstract of the article and options for obtaining
early 1950s. the full text. When possible, print or copy the full
102 PART ONE Research Overview

BOX 5-2
ONLINE DATABASES
AMERICAN COLLEGE OF PHYSICIANS: THE PHYSICIANS’ • References date from 1966
INFORMATION AND EDUCATION RESOURCE (PIER) • Available from the ERIC Web site and by subscription
• Produced by the American College of Physicians from EBSCO, OCLC, and Ovid Technologies
(ACP)
• Designed to be a point-of-care evidence-informed EXCERPTA MEDICA
resource for 300 types of disease • Biomedical database
• Available online from ACP (n.d.) or through StatRef • More than 24 million indexed records
• Approximately 7,500 current, mostly peer-reviewed
CLINICAL EVIDENCE FROM THE BRITISH MEDICAL journals
JOURNAL
• Produced by the British Medical Journal MEDLINE (MEDICAL LITERATURE ANALYSIS AND
• Systematic reviews that summarize the current state RETRIEVAL SYSTEM ONLINE)
of knowledge, or lack thereof, of medical conditions • Produced by the National Library of Medicine
• Provides evidence reviews for more than 250 • Premier bibliographic database for journal articles in
conditions life sciences
• Available online from the British Medical Journal and • References date from 1950, and approximately 5,200
Ovid Technologies worldwide journals are indexed
• Indexed with MeSH (Medical Subject Headings)
COCHRANE LIBRARY • MEDLINE is available for free through PubMed and
• Collection of databases that contain high-quality by subscription from EBSCO, OCLC, and Ovid
evidence Technologies
• Includes the Cochrane Database of Systematic
Reviews PROQUEST DISSERTATIONS AND THESES
• Full Cochrane Library available from Wiley Online • Produced by ProQuest
Library other databases that make up the Cochrane • Earliest records from 1637
Library available from other vendors, including Ovid • PDF downloads available for over 1 million
Technologies dissertations
• Cochrane systematic reviews are indexed and • Available from ProQuest (n.d.)
searchable in both CINAHL and MEDLINE
PSYCINFO
CUMULATIVE INDEX TO NURSING AND ALLIED HEALTH • Produced by the American Psychological Association
LITERATURE (CINAHL) (APA, n.d.)
• Initially called Cumulative Index to Nursing Literature • An abstract database of the psychosocial literature
• Produced by CINAHL beginning with citations dating back to 1800
• Electronic version available as part of the EBSCO • Covers more than 2,150 journals
online service • Of the journals covered, 98% are peer reviewed
• Over 1,800 journals indexed for inclusion in • Also includes book chapters and dissertations
database • Indexed with the Thesaurus of Psychological Index
• Citations in CINAHL are assigned index terms from a Terms
controlled vocabulary • Available through APA PsycNET, EBSCO, Ovid
Technologies, and ProQuest
EDUCATION RESOURCE INFORMATION CENTER (ERIC)
• Sponsored by the Institute of Education and the U.S. SCOPUS
Department of Education • Largest abstract and citation database of peer-
• Focuses on education research and information reviewed science literature and quality Web sources
• Currently indexes more than 600 journals and also • Provides 100% MEDLINE coverage
includes references to books, conference papers, and • Offers sophisticated tools to track, analyze, and
technical reports visualize research

OCLC, Online Computer Library Center.

 CHAPTER 5 Finding and Appraising the Literature 103

TABLE 5-3
SELECTED EXAMPLES OF WEB SITES AND OUTCOMES FOR LITERATURE SEARCHES
WEB SITE SCOPE NOTES

Virginia Henderson Access to the Registry of Nursing Research Service offered without charge; locate
International Nursing database, which contains nearly 30,000 conference abstracts and research study
Library: http:// abstracts of research studies and abstracts. This library is supported by
www.nursinglibrary.org conference papers. Sigma Theta Tau International, honour
society of nursing.
National Guideline Public resource for evidence-informed Offers a useful online feature of side-by-
Clearinghouse: http:// clinical practice guidelines. It contains side comparison of guidelines.
www.guidelines.gov more than 1,900 guidelines, including
non–U.S. publications.
National Institute of Nursing Promotes science for nursing practice, Able to link to Computer Retrieval of
Research: http:// funding for nursing and interdisciplinary Information on Scientific Projects (CRISP)
www.nih.gov/ninr research, and nurse scientist training and PubMed (search service of the
programs. National Library of Medicine), which
Provides links to many nursing organizations accesses literature via MEDLINE and
and search sites. PreMEDLINE and other related material
Excellent site for graduate students. from online journals; however, this site
has limited utility for the beginning
consumer of research for conducting
scholarly review of nursing research
literature because MEDLINE alone does
not include all nursing literature;
searching CINAHL and MEDLINE on your
own would be your first choice.
Useful site for graduate students in addition
to CINAHL and MEDLINE and as third
database related to topic.
Turning Research into Content from a wide variety of free online Provides a wide sampling of available
Practice (TRIP): http:// resources, including synopses, guidelines, evidence.
www.tripdatabase.com medical images, electronic textbooks, and
systematic reviews; accessed together by
the TRIP search engine.
Cochrane Collaboration: Provides free access to abstracts from the Abstracts of Cochrane Reviews are available
http://www.cochrane.org Cochrane Database of Systematic Reviews. without charge and can be browsed or
Full text of reviews and access to the searched; many databases are used in its
databases that are part of the Cochrane reviews, including CINAHL via EBSCO
Library—Database of Abstracts of Reviews and MEDLINE; some are primary sources
of Effectiveness, Cochrane Controlled (e.g., systematic reviews/meta-analyses);
Trials Register, Cochrane Methodology others (if commentaries of single studies)
Register, Health Technology Assessment are a secondary source.
database (HTA), and National Health Important source for clinical evidence but
Service (NHS) Economic Evaluation limited as a provider of primary
Database (EED)—are accessible through documents for literature reviews.
Wiley Online library
Statistics Canada: http:// Collects data on the Canadian population Free source of primary data essential for
www.statcan.gc.ca that are related to demographic trends, comprehensive demographic data and
labour, health, trade, and education. socioeconomic trends; updated daily.
Data on health trends are useful in
identifying populations at risk and suggest
associations among health determinants,
health status, and population
characteristics.
Research papers on a variety of topics are
also published.
104 PART ONE Research Overview

text, which of course will include the abstract and is to go back in the literature at least 3 years, but
the complete references. If the text is not avail- preferably 5 years, although some research proj-
able, choose the option complete reference, ects may warrant going back 10 years or more
which will include the abstract. Reading the until the researcher is satisfied that he or she
abstract is critical for determining whether you has found literature that accurately represents
need to retrieve the article through another mech- the body of knowledge. In some cases, seminal
anism. Both the CINAHL and MEDLINE elec- research or research that has had a huge effect in
tronic databases will facilitate all steps of critically the field should be reviewed regardless of publi-
reviewing the literature, especially the gaps. cation date. For example, conducting a literature
review on the effects of stress would not be
Evidence-Informed Practice Tip complete without reading Hans Selye’s (1955)
Reading systematic reviews, if they are available,
pioneering work on stress. Extensive literature
on your clinical question or topic will enhance your reviews on particular topics or a concept clarifica-
ability to implement evidence-informed nursing practice tion methodology study helps you limit the length
because they generally offer the strongest and most of your search.
consistent level of evidence.

Helpful Hint
Ask your instructor for guidance if you are uncer-
How Far Back Must the Search Go? tain how far back you need to conduct your search.
Students often ask questions such as the follow- If you come across a systematic review on your
specific clinical topic, scan it to see what years the
ing: How many articles do I need How much review covers; then begin your search from the last year
is enough and How far back in the literature to the present.
do I need to go When conducting a search, you
should use a rigorous focusing process other- As you scroll through and mark the citations
wise, you may end up with hundreds or thousands you wish to include in your downloaded or printed
of citations. Retrieving too many citations is search, make sure you include all relevant fields
usually a sign that there was something wrong when you save or print the publications. In addi-
with your search technique or that you may have tion to indicating which citations you want and
not sufficiently narrowed your clinical question. choosing which fields to print or save, you can
Each online database offers an explanation of indicate whether you want the search history
each feature it is worth your time to click on each included. It is always a good idea to include this
icon and explore the explanations offered because information because if your instructor suggests
this will increase your confidence. Also keep in that some citations were missed, you can replicate
mind the types of articles you are retrieving. your search and together figure out what variable
Many online resources allow you to limit your or variables you omitted so that you do not make
search to randomized controlled trials or system- the same error again. This is also your opportu-
atic reviews. In CINAHL, there is a limit for nity to indicate whether you want to e-mail the
Research that will restrict the number of cita- search to yourself. If you are writing a paper and
tions you retrieve to research articles. Figure 5-4 need to produce a bibliography, you can export
shows an outcome of using the Research limit your citations to citation anage ent so t are
to locate Wong and colleagues’ (2010) article on which is a software program that formats and
senior nurse leaders (see Appendix B). stores your citations so that they are available for
A general timeline for most academic or electronic retrieval when they must be inserted in
evidenced-informed practice papers and projects a paper you are writing. uite a few of these
 CHAPTER 5 Finding and Appraising the Literature 105

STEP 1: INITIAL TRY

# Search History Results Display
1 (depression and treatment). mp. [mp = ti, sh, ab, it, ot, rw] 18,255 Display
0 Saved Searches 0 Saved Searches History 0 Delete Searches 0 Remove Duplicates
Enter Keyword or Phrase
Perform Search
Limit to:
Abstracts Consumer Health Journals English Language
Full Text Available Review Articles Human Latest Update Research
Publication Year

⇓ ⇑

Ask A Librarian

STEP 2: SECOND TRY

# Search History Results Display
2 Limit 14 to research (limit not valid in OVID, MEDLINE; records were retained) 16,874 Display
3 Limit 15 to years 2008-2013 12,216 Display
4 Limit 16 to nursing research. mp. [mp = ti, sh, ab, it, ot, rw] 9 Display

FIGURE 5-4 Illustration of a search screen obtained when the Cumulative Index to Nursing
and Allied Health Literature (CINAHL) via EBSCO interface is used to locate Wong and col-
leagues’ (2010) article on the decisional involvement of senior nurse leaders.

programs are available some, such as otero, are you can browse the controlled vocabulary terms
free, and others, including EndNote and Ref- and search. sed to conduct searches in data-
Works, must be purchased, by either you or your bases, controlled vocabulary terms are carefully
institution. selected list of words and phrases that are applied
to similar pieces of information units. If you are
What Do I Need to Know? still having difficulty, ask your reference librarian
Each database usually has a specific search guide for help.
that provides information on the organization of Figure 5-4 is an illustration of a screenshot of
the entries and the terminology used. The follow- search results with CINAHL via EBSCO. As
ing suggestions and strategies, as listed in Box noted, you have the option of searching by using
5-3, incorporate general search strategies, as well the controlled vocabulary of CINAHL or a key
as those related to CINAHL and MEDLINE. word search. In this example, key word terms
Finding the right terms to plug in as key words ( nurse leadership ) and controlled vocabulary
for a computer search is an important aspect of ( decision making ) were used. Also note that
conducting a search. When it is possible, you these two concepts are connected with the
want to match the words that you use to describe oolean o erator which defines the relation-
your question with the terms that indexers have ships between words or groups of words in your
assigned to the articles. In many online databases, literature search. Examples of Boolean operators
106 PART ONE Research Overview

BOX 5-3 To broaden this search for information on deci-

TIPS: USING CINAHL VIA EBSCO sion making, you might expand your terms
entered to include policymaking and not focus
• Locate CINAHL from your library’s home page. It
may be located under databases, online resources, on only decision making. If the results from a
or nursing resources. broader search are to be saved for later use in a
• In the “Advanced” tab, type in your key word or research paper for school, it is recommended that
phrase (e.g., decision making, nurse executives,
leadership). Do not use complete sentences. (Ask you export the bibliographic information to a
your librarian for the manual guide for each citation management soft are many of these
database or use feature in the database.) programs have online interfaces so you can export
• Before you choose “Search,” make sure you mark
directly from CINAHL, MEDLINE, and other
“Research Articles,” to ensure that you have
retrieved only articles that are actually about databases. This software is designed to create bib-
research. See the results in Figure 5-4. liographies that conform to various styles such as
• Note that in the “Limit Your Results” section, you the American Psychological Association. If you
can limit by year, age group, clinical queries, and
other specific characteristics. have not used a package such as otero, EndNote,
• To narrow your search, use the Boolean connector or RefWorks, consult your librarian.
“and” between each of the key words you wish to
use and between additional variables. To broaden Helpful Hint
your search, use the Boolean connector “or.”
• Note in Figure 5-4 that you can set up an RSS feed Look for useful tools within the search interfaces
so that you are notified whenever new citations of online databases to make your searching more effi-
are added to the CINAHL database. cient. For example, when you search for a particular age
• Once the search results appear and you determine group, use the built-in limits of the database instead of
that they are manageable, you can decide whether relying on a key word search. Other shortcuts include
to review them online; print, save, or export them; the “Clinical Queries” in CINAHL and MEDLINE that
or e-mail them to yourself. retrieve articles about therapy or diagnosis.

CINAHL, Cumulative Index to Nursing and Allied Health Literature;

RSS, Really simple syndication (syndicates content from the How Do I Complete the Search?
Internet and sends it out [feeds] it to a subscriber)..
Now the truly important aspect of your searching
begins: your critical reading of the retrieved
are conjunctions such as and and or and materials. Critically reading scholarly material,
adverbs such as not and near. To restrict especially research articles, requires several read-
the retrieval to research articles, the Research ings and the use of critiquing criteria. Do not be
Article limit has been applied. discouraged if not all of the retrieved articles are
For further reading, a series in ournal of as useful as you first thought this happens with
Emergency Nursing outlined the steps for the most experienced reviewers of literature. If
embarking on an evidence-informed search most of the articles are not useful, be prepared to
(Bernardo, 2008 Engberg Schlenk, 2007 Klem perform another search, but discuss with your
Northcutt, 2008), which covers the stages of instructor, the reference librarian, or both, the
evidence-informed practice and basic search search terms that you will use next time you may
hints. This search was aimed at locating a known want to add a third database. In the previous
article to demonstrate some of the features of example of a psychoeducational intervention for
CINAHL. Lawrence (2007) presented a step-by- women who are survivors of breast cancer, the
step approach for conducting a search in CINAHL. third database of choice may be PsycINFO (see
She described the controlled vocabulary, Boolean Box 5-2). Remind yourself how quickly you will
operators, and tips for narrowing results in the be able to perform the search, now that you have
EBSCO interface. experience with searching.
 CHAPTER 5 Finding and Appraising the Literature 107

of decreasing distress in children. Music was

Helpful Hint chosen in those studies to provide distraction
Read the abstract carefully (review the discussion because, as Sobieraj and colleagues noted, it is
on critical reading strategies in Chapter 3) to determine
whether the article is about research. It is also a good recommended for the treatment of pain however,
idea to review the references of the articles; if any seem this recommendation was not supported in the
relevant, you can retrieve them. literature review. Sobieraj and colleagues further
reported that parents who actively sing during a
LITERATURE REVIEW FORMAT: procedure may experience less anxiety. The
WHAT TO EXPECT research question and subsequent hypothesis
Familiarity with the format of the literature were then formulated on the basis of the relation-
review helps research consumers use critiquing ship between the conceptual areas: Does music
criteria to evaluate the review. To decide which broadcast via speakers have a measurable effect
style you will use so that your review is presented on parent-led distraction during simple laceration
in a logical and organized manner, you must con- repair in children aged 1 to 5
sider the following: In contrast to the styles of previous quantita-
The research or clinical question or topic tive studies, the literature reviews of qualitative
The number of retrieved sources reviewed studies are usually handled in a different manner.
The number and type of research materials Little is often known about the topic under study.
versus conceptual materials The literature review may be conducted at the
Some reviews are written according to the vari- beginning of the study or after the data analysis
ables being studied and presented chronologically is completed. The researchers always compare
in the discussion of each variable. In others, the the literature review with their findings. In some
entire material is presented chronologically, and cases, the reviewed literature is used during the
subcategories or variables are discussed within analysis process as well. In the study by Senevi-
each time period. In still others, the variables are ratne and associates (2009) on understanding
presented and the subcategories are related to the nursing on an acute stroke unit, the background
study’s type or designs or related variables. section included a literature review that focused
An example of a literature review, although on the role of nurses in stroke rehabilitation, atti-
brief, that was presented logically according to tudes and perceptions of their role in stroke care,
the variables under study was that of Sobieraj and and observations of nurses in acute and rehabili-
colleagues (2009 see Appendix C). The research- tative care settings. After the data analysis, Sen-
ers stated that the purpose of their study was to eviratne and associates returned to the literature
test an intervention using children’s songs to to help explain and discuss the findings. For
promote parent-led distraction during simple lac- example, they compared the literature on space
eration repair in children aged 1 to 5. In the limitations and time constraints to further explore
Literature Review, the researchers indicated the challenges faced by nurses in the study (see
that there was extensive research on the under- Appendix A).
treatment of pain in children and the long-term
negative effects. Establishing the importance of
Evidence-Informed Practice Tip
the study, the authors identified age as a signifi-
Sort the research articles you retrieve according to
cant variable in the literature. The literature the model of levels of evidence in Chapter 3. Remember
review then focused on psychological interven- that articles that are systematic reviews, especially meta-
tions and coping strategies by parents. The authors analyses, generally provide the strongest and most con-
sistent evidence to include in a literature review.
concluded that distraction is an effective means
108 PART ONE Research Overview

Helpful Hint Helpful Hint

When you write up your literature review, include • Use standardized critical appraisal criteria to
enough information so that your professor or fellow evaluate your research articles.
students could re-create your search path and come up • Make a table to represent the components of your
with the same results. This means specifying the data- study, and fill in your evaluation to help you see
bases searched, the date you searched, years of cover- the “big picture” of your analysis.
age, terms used, and any limits or restrictions that you • Synthesize the results of your analysis to try to
used. determine what was similar or different among
and between these studies in relation to your topic
or clinical question, and then draw a conclusion.

APPRAISING THE EVIDENCE

Review of the Literature
Whether you are a researcher writing the literature review practice project that builds on previous research, identifies
for the research study you are planning to conduct or a a conflict or gap in the literature, or proposes to extend
nurse writing a literature review for an evidence-informed the current knowledge base?” (see Box 5-1).
practice project, you need to critically appraise individual Regardless of how the literature review is organized,
research reports by using appropriate criteria. If you are it should provide a strong knowledge base for carrying
appraising an individual research study that is to be out the research, educational, or clinical practice project.
included in a literature review, it must be evaluated in Questions related to the logical organization and presen-
terms of critical appraisal criteria that are related to each tation of the reviewed studies are somewhat more chal-
step of the research process so that the strengths and lenging for beginning research consumers. The more you
weaknesses of each study can be identified. Standardized read research studies, the more competent you become
critical appraisal tools (e.g., CASP Tools and AGREE Guide- at differentiating a well-organized literature review from
lines) available for specific types of research designs (e.g., one that has no organizing framework.
clinical trials, cohort studies, systematic reviews) can also Whenever possible, read both qualitative (meta-
be used to critically appraise an individual research study. syntheses) and quantitative (meta-analyses) systematic
Critiquing the literature review of research or concep- reviews that pertain to a clinical question and provide
tual reports is a challenging task for seasoned consumers level I evidence. Systematic reviews are considered exam-
of research, so do not be surprised if you feel a little ples of secondary sources because they represent a body
intimidated by the prospect of critiquing the published of completed research studies that have been critically
research. The important issue is to determine the overall appraised and synthesized by a team other than the origi-
value of the literature review, including both the research nal researchers; however, they often represent the best
and theoretical materials. The purposes of a literature available evidence on a particular clinical issue. The
review (see Box 5-1) and the characteristics of a well- article by Cummings and associates (2010) is an example
written literature review (Box 5-4) provide the framework of a quantitative systematic review in which the authors
for developing the evaluation criteria for a literature critically appraised and synthesized the evidence from
review. research studies related to the leadership behaviours and
The literature review should be presented in an orga- outcomes for nurses and organizations. After reviewing
nized manner. The theoretical and research literature can and synthesizing 53 studies, Cummings and associates
be presented chronologically from earliest studies to most concluded that “transformational and relational leader-
recent; sometimes the theoretical literature that provided ship are needed to enhance nurse satisfaction, recruit-
the foundation for the existing research is presented first, ment, retention and health work environments” (p. 363).
followed by the research studies that were derived from In her article “Becoming a nurse: a meta-study of early
this theoretical base. Other times, the literature can be professional socialization and career choice in nursing,”
clustered by concept, grouped according to supportive or Price (2009) gathered 10 primary qualitative studies on
nonsupportive positions, or categorized by evidence that professional socialization in nursing and nurses’ career
highlights differences in theoretical and/or research find- choice decisions. After synthesizing the studies, Price
ings. The overall question to be answered is “Does the reported “that nursing socialization is strongly associated
review of the literature develop and present a knowledge with a person’s preconceived notions and expectations of
base for a research study or an evidence-informed nursing” (p. 14).
 CHAPTER 5 Finding and Appraising the Literature 109

APPRAISING THE EVIDENCE

Review of the Literature—cont’d
The Critiquing Criteria box summarizes general critiqu- The key to a strong literature review is a careful search
ing criteria for a review of the literature. In other sets of of the published and unpublished literature. Whether you
critiquing criteria, these questions may be phrased differ- write or critically appraise a literature review written for a
ently or more broadly. For instance, questions may be the published research study, it should reflect a synthesis
following: “Does the literature search seem adequate?” or compilation of the main points or value of all of the
and “Does the report demonstrate scholarly writing?” You sources reviewed in relation to the study’s research ques-
may have difficulty answering these questions; you may tion or hypothesis (see Box 5-1). The relationship between
begin, however, by determining whether the source is and among these studies must be explained. The synthe-
a refereed journal. It is reasonable to assume that the sis of a written review of the literature usually appears at
manuscripts published in a scholarly refereed journal are the end of the review section before the section about
adequately searched, are based mainly on primary the research question or hypothesis.
sources, and are written in a scholarly manner. This does Searching the literature, like critiquing the literature,
not mean, however, that every study reported in a refer- is an acquired skill. Practising your search and critical
eed journal meets all the critiquing criteria for a literature appraisal skills on a regular basis will make a huge differ-
review and other components of the study in an equal ence. Seeking guidance from faculty is essential for devel-
manner. Because of style differences and space con- oping critical appraisal skills. Synthesizing the body of
straints, each citation summarized is often very brief, or literature you have critiqued is even more challenging.
related citations may be summarized as a group and lack Critiquing the literature will help you apply new knowl-
a critique. You still must answer the critiquing questions. edge to practice. This process is vital to the “survival and
Consultation with a faculty adviser may be necessary to growth of the nursing profession and is essential to
develop skill in answering the two questions. evidence-based practice” (Pravikoff & Donaldson, 2001).

CRITIQUING CRITERIA

1. Are all the relevant concepts 7. Does the literature review build which the overall strengths and
and variables included in the on the findings of earlier weaknesses of the reviewed
review? studies? studies are presented and a
2. Does the search strategy 8. Does the summary of each logical conclusion is
include an appropriate and reviewed study reflect the established?
adequate number of databases essential components of the 11. Is the literature review
and other resources to identify study design (e.g., type and presented in an organized
key published and unpublished size of sample, reliability and format that flows logically (e.g.,
research and theoretical validity of instruments, chronologically, clustered by
sources? consistency of data-collection concept or variables), enhancing
3. Are both theoretical literature procedures, appropriate data the reader’s ability to evaluate
and research literature analysis, identification of the need for the particular
included? limitations)? research study or evidence-
4. Does an appropriate theoretical 9. Does the critique of each informed practice project?
or conceptual framework guide reviewed study mention 12. Does the literature review
the development of the strengths, weaknesses, or follow the proposed purpose of
research study? limitations of the design; the research study or evidence-
5. Are mainly primary sources conflicts; and gaps in informed practice project?
used? information related to the area 13. Does the literature review
6. What gaps or inconsistencies in of interest? generate research questions or
knowledge does the literature 10. Does the synthesis summary hypotheses or answer a clinical
review uncover? follow a logical sequence in question?
110 PART ONE Research Overview

BOX 5-4
CHARACTERISTICS OF A WELL-WRITTEN REVIEW KEY POINTS
OF THE LITERATURE
• The review of the literature is defined as a broad,
Each reviewed source of information reflects critical comprehensive, in-depth, systematic critique and
thinking and scholarly writing and is relevant to the synthesis of scholarly publications, unpublished
study, topic, or project, and the content satisfies the scholarly print and online materials, audiovisual
following criteria: materials, and personal communications.
• The literature review is organized in a • The review of the literature is used for
systematic approach. development of research studies, as well as other
• Each research or conceptual article is activities for consumers of research, such as
summarized succinctly and with appropriate development of evidence-informed practice
references. projects.
• Established critical appraisal criteria are used for • With regard to conducting and writing a literature
specific study designs to evaluate the study for review, the main objectives for the consumer of
strengths, weaknesses, or limitations, as well as research are to acquire the abilities to accomplish
for conflicts or gaps in information that relate the following: (1) conduct an appropriate search of
directly or indirectly to the area of interest. electronic or print research on a topic; (2) efficiently
• Evidence of a synthesis of the critiques is retrieve a sufficient amount of materials for a
provided to highlight the overall strengths and literature review in relation to the topic and scope
weaknesses of the studies reviewed. of project; (3) critically appraise (i.e., critique)
• The review consists of mainly primary sources; research and theoretical material in accordance
there are a sufficient number of research with accepted critiquing criteria; (4) critically
sources. evaluate published reviews of the literature in
• The review concludes with a synthesis of the accordance with accepted standardized critiquing
reviewed material that reflects why the study criteria; (5) synthesize the findings of the critiqued
or project should be implemented. materials for relevance to the purpose of the
• Research questions and hypotheses are selected scholarly project; and (6) determine
identified, or clinical questions are answered. applicability of the findings to practice.
• Primary research and theoretical resources are
essential for literature reviews.
• The use of secondary sources, such as
commentaries on research articles from peer-
CRITICAL THINKING CHALLENGES reviewed journals, is part of a learning strategy for
developing critical critiquing skills.
■ Using the PICOT format, generate a clinical • It is more efficient to use electronic rather than
print databases for retrieving scholarly materials.
question related to health promotion for children • Strategies for efficiently retrieving scholarly nursing
in elementary school. literature include consulting the reference librarian
■ How does a research article’s theoretical or and using at least two online sources (e.g., CINAHL
conceptual framework interrelate concepts, and MEDLINE).
theories, conceptual definitions, and operational • Literature reviews are usually organized according
to variables, as well as chronologically.
definitions? • Critiquing and synthesizing a number of research
■ A general guideline for a literature search is to articles, including systematic reviews, is essential
use a timeline of 3 to 5 years. When would a for implementing evidence-informed nursing
nurse researcher need to search beyond this practice.
timeline?
■ What is the relationship of the research article’s
literature review to the theoretical or conceptual
framework?
 CHAPTER 5 Finding and Appraising the Literature 111

REFERENCES National Library of Medicine. (n.d.). Retrieved from

http://www.nlm.nih.gov/bsd/disted/
American College of Physicians. (n.d.). ACP PIER pubmedtutorial/010 050.html
overvie . Retrieved from http://pier.acponline.org/ Pravikoff, D., Donaldson, N. (2001). Special section:
index.html Online journal publication. nline ournal of Issues
American Psychological Association. (n.d.) PsycIN . in Nursing, (2). Retrieved from http://
Retrieved from http://www.apa.org/pubs/databases/ www.nursingworld.org/MainMenuCategories/
psycinfo/index.aspx ANAMarketplace/ANAPeriodicals/OJIN/
Anderson, R. A., McDaniel, R. R. (1998). Intensity TableofContents/Volume62001/No2May01/
of registered nurse participation in nursing home ArticlePreviousTopic/ClinicalInnovations.aspx
decision-making. The erontologist, (1), 90-100. Price, S. L. (2009). Becoming a nurse: A meta-study of
Anderson, R. A., McDaniel, R. R. (1999). RN early professional socialization and career choice in
participation in organizational decision-making and nursing. ournal of Advanced Nursing, (1), 11-19.
improvements in resident outcomes. ealth Care doi: 10.1111/j.1365-2648.2008.04839.x
Management Revie , (10), 7-16. Pro uest. (n.d.). Pro uest dissertations and theses
Ashmos, D. P., Huonker, J. W., McDaniel, R. R. database: ey facts. Retrieved from http://
(1998). Participation as a complicating mechanism: www.il.proquest.com/products pq/descriptions/
The effect of clinical professional and middle pqdt.shtml
manager participation on hospital performance. Selye, H. (1955). Stress and disease. Science, ,
ealth Care Management Revie , (4), 7-20. 625-631.
Bernardo, L. M. (2008). Finding the best evidence, part Seneviratne, C. C., Mather, C. M., Then, K. L.
1: nderstanding electronic databases. ournal of (2009). nderstanding nursing on an acute stroke
Emergency Nursing, (1), 59-60. unit: Perceptions of space, time and interprofessional
Cummings, G. G., MacGregor, T., Davey, M., Lee, H., practice. ournal of Advanced Nursing 9),
Wong, C., Lo, E., . . . Stafford, E. (2010). Leadership 1872-1881. doi: 10.1111/j.1365-2648.2009.05053.x
styles and outcome patterns for the nursing workforce Sobieraj, G., Bhatt, M., LeMay, S., Rennick, J.,
and work environment: A systematic review. Johnston, C. (2009). The effect of music on parental
International ournal of Nursing Studies, , participation during pediatric laceration repair.
363-385. doi: 10.1016/j.ijnurstu.2009.08.006 Canadian ournal of Nursing Research, , 68-82.
DiCenso, A., Bayley, L., & Haynes, R. B. (2009). Turning Research into Practice TRIP . (n.d.). Introduc-
Accessing pre-appraised evidence: Fine-tuning the tion and bac ground TRIP database, .
5S model into a 6S model. Evidence Based Nursing, Retrieved from http://www.tripdatabase.com/about
12(4), 99-101. doi: Wiley Interface. (n.d.). Reference or s, The
Engberg, S., Schlenk, E. A. (2007). Asking the right Cochrane ibrary , Issue . Retrieved from
question. ournal of Emergency Nursing, (6), http://www.thecochranelibrary.com/view/0/index.html
571-573. Wong, C. A., Laschinger, H., Cummings, G. G.,
Haynes, B. (2007). Of studies, syntheses, synopses, Vincent, L., O’Connor, P. (2010). Decisional
summaries, and systems: The 5S evolution of involvement of senior nurse leaders in Canadian
information services for evidence-based healthcare acute care hospitals. ournal of Nursing Management,
decisions. Evidence-Based Nursing, (1), 6-7. , 122-133. doi: 10.1111/j.1365-2834.2010.01053.x
Kendall, S. (2008). Evidence-based resources simplified.
Canadian Family Physician, 54(2), 241-243.
Klem, M. L., Northcutt, T. (2008). Finding the best
FOR FURTHER STUDY
evidence part 2: The basics of literature searches. Go to Evolve at http://evolve.elsevier.com/
ournal of Emergency Nursing, (2), 151-158. Canada/LoBiondo/Research for Audio Glossary, how-to
Lawrence, J. C. (2007). Techniques for searching instructions for Writing Proposals for Funding, and
the CINAHL database using the EBSCO interface. additional research articles for practice in reviewing
A RN ournal, (4), 779-780, 782-788, 790-791. and critiquing.
C H A PTER 6

Legal and Ethical Issues

Judith Haber | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Describe the historical background that led to the development of ethical guidelines for the use of
human participants in research.
• Identify the essential elements of an informed consent form.
• Evaluate the adequacy of an informed consent form.
• Describe the role of the research ethics board in the research review process.
• Identify populations of participants who require special legal and ethical research considerations.
• Appreciate the nurse researcher’s obligations to conduct and report research in an ethical manner.
• Describe the nurse’s role as patient advocate in research situations.
• Discuss the nurse’s role in ensuring that Health Canada guidelines for testing of medical devices
are followed.
• Discuss animal rights in research situations.
• Critique the ethical aspects of a research study.

KEY TERMS
animal rights consent research ethics board
anonymity ethics respect for persons
assent informed consent risk-benefit ratio
beneficence justice risks
benefits process consent
confidentiality product testing

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

112
 CHAPTER 6 Legal and Ethical Issues 113

NURSES ARE IN AN IDEAL POSITION to promote with acquired immune deficiency syndrome
patients’ awareness of the role played by research (AIDS), as well as animals—are discussed. The
in the advancement of science and improvement nurse’s role as patient advocate, whether func-
in patient care. In Canada, the professional code tioning as researcher, caregiver, or research con-
of ethics (Canadian Nurses Association CNA , sumer, is addressed.
2008) outlines the ethical standards for practice,
which can include research and patients’ rights ETHICAL AND LEGAL
with regard to research. Not only do the standards CONSIDERATIONS IN RESEARCH:
represent rules and regulations regarding prac- A HISTORICAL PERSPECTIVE
tice, but when research becomes the domain of
a nurse, these standards can be applied to the Past Ethical Dilemmas in Research
participation of human research participants to Ethical and legal considerations regarding med-
ensure that nursing research is conducted legally ical research first arose in the nited States and
and ethically. The code states that nurses must received focused attention after World War II.
strive to uphold human rights and call attention Lawyers defending war criminals intended to
to any violations of these rights. The Code of justify the atrocities committed by Nazi physi-
Ethics for Registered Nurses, originally published cians by claiming their actions were in the name
in 1983, was revised in 2008, and the revision of medical research. On learning of this defence,
was released that June. The revised CNA code the .S. Secretary of State and the Secretary of
includes clarification of the meanings of social War asked the American Medical Association to
justice, health, and well-being, with a stronger appoint a group to develop a code of ethics for
focus on human rights. These rights can also be research to serve as a standard for judging the
translated to patients’ or participants’ rights in medical experiments committed by physicians on
research, and nurses can be advocates to ensure concentration camp prisoners. These experiments
that ethical concepts in nursing research are included the sterilization of people considered
upheld. enemies of the state (Benedict Georges, 2006).
Researchers and caregivers of patients who are The code of ethics developed as 10 rules that
research participants must be fully committed to became known as the Nuremberg Code (Box
the tenets of informed consent and patients’ 6-1). The Nuremberg Code’s definitions of the
rights. The principle the ends justify the means terms voluntary, legal capacity, suf cient under-
must never be tolerated. Researchers and caregiv- standing, and enlightened decision have been the
ers of research participants must take every pre- subject of numerous court cases and .S. presi-
caution to protect people being studied from dential commissions involved in setting ethical
physical or mental harm or discomfort (although standards in research (Creighton, 1977). The code
it is not always clear what constitutes harm or that was developed requires informed consent in
discomfort). all cases but makes no provisions for any special
The focus of this chapter is the legal and treatment of children, older adults, or people who
ethical considerations that must be addressed are mentally incompetent. Several other interna-
before, during, and after the conduct of research tional standards have followed the most notable
to ensure that the research does not harm the is the Declaration of Helsinki, which was adopted
patient. Informed consent, research ethics boards in 1964 by the World Medical Assembly and
(REBs), and research involving vulnerable revised in 1975 (Levine, 1979).
populations—older adults, pregnant women, chil- The research heritage in the nited States and
dren, prisoners, Aboriginal people, and persons Canada is well documented and is used here to
114 PART ONE Research Overview

BOX 6-1 BOX 6-2

ARTICLES OF THE NUREMBERG CODE BASIC ETHICAL PRINCIPLES RELEVANT TO THE
1. The voluntary consent of the human subject is CONDUCT OF RESEARCH
absolutely essential. RESPECT FOR PERSONS
2. The study should be conducted so as to yield People have the right to self-determination and to
fruitful results for the good of society, treatment as autonomous agents. Thus they have the
unprocurable by other means of study, and not freedom to participate or not participate in research.
random and unnecessary in nature. Persons with diminished autonomy are entitled to
3. The experiment should be so designed and based protection.
on the results of animal experimentation and
knowledge of the natural history of the disease or BENEFICENCE
other problems under study that the anticipated Beneficence is an obligation to do no harm and
results will justify the performance of the maximize possible benefits. Persons are treated in an
experiment. ethical manner when their decisions are respected,
4. The experiment should be conducted to avoid all they are protected from harm, and efforts are made to
unnecessary physical and mental suffering and secure their well-being.
injury.
5. No experiment should be conducted where there JUSTICE
is an a priori reason to believe that death or Human subjects should be treated fairly. An injustice
disabling injury will occur. . . . occurs when benefit to which a person is entitled is
6. The degree of risk to be taken should never denied without good reason or when a burden is
exceed that determined by the humanitarian imposed unduly.
importance of the problem to be solved by the
experiment. From Elder, G. (1981). Social history & life experience. In D. H.
7. Proper preparations should be made and Eichorn, J. A. Clausen, N. Haan, & P. H. Mussen (Eds.), Present and
adequate facilities provided to protect the subject past in middle life (pp. 3-31). New York: Academic Press. Copyright
1981 Academic Press.
against even remote possibilities of injury,
disability, or death.
8. The experiment should be conducted only by
scientifically qualified persons. . . .
9. . . . The human subject should be at liberty to National Commission for the Protection of
bring the experiment to an end. . . . Human Subjects of Biomedical and Behavioral
10. During the course of the experiment the scientist
in charge must be prepared to terminate the Research was created. A major charge of the com-
experiment at any stage, if he [or she] has mission was to identify the basic principles that
probable cause to believe . . . that a continuation should underlie the conduct of biomedical and
of the experiment is likely to result in injury,
behavioural research involving human partici-
disability, or death to the experimental subject.
pants and to develop guidelines to ensure that
From United States Government Printing Office. (2008). The research is conducted in accordance with those
medical case. In Trials of war criminals before the Nuernberg
Military Tribunals under Control Council Law No. 10 (Vol. 2, principles (Levine, 1986). Three ethical princi-
pp. 181-182). Washington, DC: Author, 1949. Retrieved from http:// ples were identified as relevant to the conduct of
www.loc.gov/rr/frd/Military_Law/pdf/NT_war-criminals_Vol-II.pdf.
research involving human participants: res ect
or ersons (the idea that people have the freedom
to participate or not participate in research),
illustrate the human consequences of not adher- bene cence (obligation to do no harm and maxi-
ing to ethical standards when conducting research. mize possible benefits), and ustice (the principle
Some examples are highlighted in Table 6-1 and that human subjects should be treated fairly Box
incorporated into Table 6-2 to show the violation 6-2). These three principles have formed the basis
of human rights that occurred in these studies. of many ethical guidelines in Canada.
In the nited States, under the National In Canada, for the protection of human
Research Act of 1974 (Public Law 93-348), the participants in all types of research, Health
 TABLE 6-1
HIGHLIGHTS OF UNETHICAL RESEARCH STUDIES CONDUCTED IN THE UNITED STATES AND CANADA
RESEARCH STUDY DATE OF STUDY FOCUS OF STUDY ETHICAL PRINCIPLE VIOLATED

Tuskegee syphilis 1932–1973 For 40 years, the U.S. Public Health Many of the research participants
study, Tuskegee, Service conducted a study using two who consented to participate in
Alabama groups of poor black male share- the study were not informed
croppers. One group consisted of men about the purpose and procedures
with untreated syphilis; the other group of the research. Others were
was judged to be free of the disease. unaware that they were
Treatment was withheld from the group participants. The degree of risk
with syphilis even after penicillin became outweighed the potential benefit.
generally available and accepted as Withholding of known effective
effective treatment for syphilis in the treatment violates the participants’
1950s. Steps were even taken to prevent right to fair treatment and
the research participants from obtaining protection from harm (Levine,
penicillin. The researcher wanted to 1986).
study the untreated disease.
Sterilization 1940–1944 Sterilization experiments Basic human rights and rights to fair
experiments in and ethical treatment were violated,
Auschwitz and the research participants did
concentration not give informed consent. Nurses
camp, Germany who were prisoners were forced to
participate in the experiments,
which was against their prima
facie duty to protect (Benedict &
Georges, 2006).
Dr. Ewen Cameron’s 1950s–1960s The U.S. Central Intelligence Agency (CIA) The ethical principles of respect for
psychiatric funded psychic driving, or brain- persons and beneficence were
experiments, Allan washing, experiments on patients with severely violated. Dr. Cameron
Memorial Psychiatric psychiatric illnesses (Collins, 1988, cited used patients with diminished
Institute, Montreal, by Charron, 2000). Psychic driving is a autonomy (patients with
Quebec psychiatric procedure pioneered by psychiatric illnesses), even though,
Dr. Cameron in which electroconvulsive as a physician, he was obliged to
therapy (ECT) and psychedelic drugs, protect them. The ECT treatments
such as lysergic acid (LSD), are used in did more harm than good.
an attempt at mind control. To develop
the psychic driving, increasingly higher
levels of ECT were applied to patients
as often as three times a day. This
treatment would continue for 30 days.
Considerable damage was done to
patients after such severe treatment.
Patients were unable to walk or feed
themselves and were incontinent
(Gillmor, 1987, cited by Charron, 2000).
Hyman v. Jewish 1965 Doctors injected aged, senile patients Informed consent was not obtained,
Chronic Disease with cancer cells to study the patients’ and no indication was given that
Hospital, Jewish response to injection of the cells. the study had been reviewed and
Chronic Disease approved by an ethics committee.
study, New York City The two physicians involved
claimed that they did not wish to
evoke emotional reactions or New
York City refusals to participate by
informing the research participants
of the nature of the study
(Hershey & Miller, 1976).

Continued
 TABLE 6-1
HIGHLIGHTS OF UNETHICAL RESEARCH STUDIES CONDUCTED IN THE UNITED STATES AND CANADA—cont’d
RESEARCH STUDY DATE OF STUDY FOCUS OF STUDY ETHICAL PRINCIPLE VIOLATED
Midgeville, Georgia, 1969 Researchers administered investigational The study protocol or institutional
study drugs to mentally disabled children approval of the program was not
without first obtaining the opinion of a reviewed before implementation
psychiatrist. (Levine, 1986).
San Antonio 1969 In a study of the side effects of oral Principles of informed consent were
contraceptive study, contraceptives, 76 impoverished violated; full disclosure of the
San Antonio, Texas Mexican American women were potential risk, harm, results, and
randomly assigned to an experimental side effects was not evident in the
group receiving birth control pills or a informed consent document. The
control group receiving placebos. potential risk outweighed the
Research participants were not benefits of the study. The
informed about the placebo and the participants’ right to fair treatment
attendant risk of pregnancy. Of the and protection from harm was
participants, 11 became pregnant; 10 of violated (Levine, 1986).
these women were in the placebo
control group.
Willowbrook Hospital 1972 Children with mental incompetence The principle of voluntary consent
study, New York (N = 350) were not admitted to was violated. Parents were coerced
State Willowbrook Hospital, a residential into consenting to their children’s
treatment facility, unless parents participation for the research.
consented to their children’s being Participants or their guardians
research participants in a study of the have a right to self-determination;
natural history of infectious hepatitis in other words, they should be
and the effect of γ-globulin. The free of constraint, coercion, and
children were deliberately infected undue influence of any kind.
with the hepatitis virus under various Many participants feel pressured
conditions; some received γ-globulin, to participate in studies if they are
whereas others did not. in powerless, dependent positions
(Rothman, 1982).
Schizophrenia 1983 In a study of the effects of withdrawing Although all participants signed
medication study, psychotropic medications in 50 informed consent documents, they
University of patients receiving treatment for were not informed about how
California, Los schizophrenia, 23 research participants severe their relapses might be or
Angeles suffered severe relapses after their that they could suffer worsening
medication was stopped. The goal of symptoms with each recurrence.
the study was to determine whether Principles of informed consent
some patients with schizophrenia were violated; full disclosure of
might do better without medications the potential risk, harm, results,
that had deleterious side effects. and side effects was not evident in
the informed consent document.
The potential risk outweighed the
benefits of the study. The
participants’ right to fair treatment
and protection from harm was
violated (Hilts, 1995).
Côte d’Ivoire, Africa, 1994 In research supported by the U.S. Research participants who consented
AIDS/AZT case government and conducted in the Côte to participate and who were
d’Ivoire, Dominican Republic, and randomly assigned to the control
Thailand, some pregnant women group were denied access to a
infected with HIV were given placebo medication regimen with a known
pills rather than AZT, a drug known to benefit. This denial violate the
prevent mothers from passing the virus participants’ right to fair treatment
to their babies. Babies born to these and protection (French, 1997;
mothers were in danger of contracting Wheeler, 1997).
a fatal disease.
AIDS, acquired immune deficiency syndrome; AZT, azidothymidine; HIV, human immunodeficiency virus.
 CHAPTER 6 Legal and Ethical Issues 117

TABLE 6-2
PROTECTION OF HUMAN RIGHTS
BASIC HUMAN RIGHT DEFINITION

Right to self-determination This right is based on the ethical principle of respect for persons; people should be treated
as autonomous agents who have the freedom to choose without external controls. An
autonomous agent is one who is informed about a proposed study and is allowed to
choose to participate or not to participate (Brink, 1992). Moreover, research participants
have the right to withdraw from a study without penalty.
Right to privacy and dignity This right is based on the ethical principle of respect for persons; privacy is the freedom of a
person to determine the time, extent, and circumstances under which private information
is shared or withheld from other people.
Right to anonymity and This right is based on the ethical principle of respect for persons; anonymity exists when the
confidentiality participant’s identity cannot be discerned, even by the researcher, from his or her
individual responses (American Nurses Association, 1985).
Confidentiality means that the individual identities of participants will not be linked to the
information they provide and will not be publicly divulged.
VIOLATION OF BASIC HUMAN RIGHT EXAMPLE

A participant’s right to self-determination is violated Participants may believe that their care will be adversely
through the use of coercion, deception, and covert affected if they refuse to participate in research. The
data collection. Willowbrook Hospital Study (see Table 6-1) is an example of
• In coercion, an overt threat of harm or excessive how coercion was used to obtain the consent of parents of
reward is presented to ensure participants’ vulnerable children with mental retardation, who would not
compliance. be admitted to the institution unless they participated in a
• In deception, participants are misinformed about study in which they were deliberately injected with the
the purpose of the research. hepatitis virus.
• In covert data collection, people become research The Jewish Chronic Disease Hospital Study (see Table 6-1) is an
participants and are exposed to research example of a study in which patients and their personal phy-
treatments without knowing it. sicians did not know that cancer cells were being injected.
• The potential for violation of the right to self- In Milgram’s (1963) study, research participants were deceived
determination is greater for research participants when asked to administer electric shocks to another person,
with diminished autonomy, who have decreased who was an actor pretending to suffer from the shocks.
ability to give informed consent and are Participants administering the shocks were very distressed by
vulnerable. participating in this study, although they were not admin-
istering shocks at all. This study is an example of deception.
The U.S. Privacy Act of 1974 was instituted to protect Research participants may be asked personal questions such as
participants from privacy violations. These violations “Were you sexually abused as a child?”; “Do you use
occur most frequently during data collection, when drugs?”; and “What are your sexual preferences?” When
responses to invasive questions might result in the questions are asked in the presence of hidden microphones
loss of a job, friendships, or dignity or might create or hidden recording devices, the participants’ privacy is
embarrassment and mental distress. These violations invaded because they have no knowledge that the data are
also may occur when participants are unaware that being shared with other people. Participants’ right to control
information is being shared with other people. access of other people to their records is also violated.
Anonymity is violated when the participants’ responses Researchers who choose to identify data by using the
can be linked to their identity. participant’s name are breaching the basic human right of
anonymity. Instead, researchers should assign participants a
code number that is used for identification purposes.
Research participants’ names are never used in the reporting
of findings.
Confidentiality is breached when a researcher, by Breaches of confidentiality with regard to sexual preference,
accident or direct action, allows an unauthorized income, drug use, prejudice, or person personality variables
person to gain access to study data that contain can be harmful to research participants. Data should be
information about the participant’s identity or analyzed as group data so that participants cannot be
responses, which creates a potentially harmful identified by their responses.
situation for the participant.

Continued
118 PART ONE Research Overview

TABLE 6-2
PROTECTION OF HUMAN RIGHTS—cont’d
BASIC HUMAN RIGHT DEFINITION

Right to fair treatment This right is based on the ethical principle of justice; people
should be treated fairly and should receive what they are
due or owed.
“Fair treatment” refers to the equitable selection of research
participants and their treatment during the research study.
This treatment includes selection of participants for reasons
directly related to the problem studied, as opposed to
selection of participants because of convenience, the
compromised position of the participants, or their
vulnerability. Fair treatment also extends to the treatment of
participants during the study, including fair distribution of
risks and benefits of the research regardless of age, race, or
socioeconomic status.
Right to protection from discomfort and harm This right is based on the ethical principle of beneficence;
people must take an active role in promoting good and
preventing harm both in the world around them and in
research studies.
Discomfort and harm can be physical, psychological, social, or
economic in nature.
The five levels of harm and discomfort are as follows:
1. No anticipated effects
2. Temporary discomfort
3. Unusual level of temporary discomfort
4. Risk of permanent damage
5. Certainty of permanent damage
Participants with diminished autonomy are entitled to
protection. They are more vulnerable because of age, legal
or mental incompetence, terminal illness, or confinement to
an institution.
A justification for the use of vulnerable participants must be
provided.
VIOLATION OF BASIC HUMAN RIGHT EXAMPLE

Injustices with regard to participant selection have The Tuskegee Syphilis Study that ended in 1973 (Levine, 1986),
occurred as a result of social, cultural, racial, and the Jewish Chronic Disease Study of 1965 (Hershey & Miller,
gender biases in society. 1976), the San Antonio Contraceptive Study of 1969 (Levine,
1986), and the Willowbrook Hospital Study of 1972
(Rothman, 1982) (see Table 6-1) all are examples
of unfair participant selection and the use of vulnerable
populations.
Historically, research participants were often recruited Investigators should not be late for data collection
from groups of people who were regarded as having appointments, should terminate data collection on time,
less “social value,” such as people living in poverty, should not change agreed-upon procedures or activities
prisoners, slaves, people who are mentally without consent, and should provide agreed-upon benefits,
incompetent, and people who are dying. Participants such as a copy of the study findings or a participation fee.
were often treated carelessly, without consideration
of physical or psychological harm.
 CHAPTER 6 Legal and Ethical Issues 119

TABLE 6-2
PROTECTION OF HUMAN RIGHTS—cont’d
VIOLATION OF BASIC HUMAN RIGHT EXAMPLE
Research participants’ right to be protected is violated Temporary physical discomfort involving minimal risk includes
when discomfort or disabling injury will occur and, fatigue or headache; emotional discomfort includes the
thus, the benefits do not outweigh the risks. expense involved in travelling to and from the data
collection site.
Studies of sensitive issues (such as rape, incest, or spouse
abuse) might cause unusual levels of temporary discomfort
by increasing participants’ awareness of current or past
traumatic experiences. In these situations, researchers assess
distress levels and provide debriefing sessions, during which
the participant may express feelings and ask questions. The
researcher has the opportunity to make referrals for
professional intervention.
Studies with the potential to cause permanent damage are
more likely to be medical in nature rather than nursing in
nature, inasmuch as physiological damage may be
permanent. One clinical trial of a new drug, a recombinant
activated protein C (Zovan) for treatment of sepsis, was
halted when interim findings from the phase III clinical trials
revealed that the rate of mortality among the patients
receiving treatment was lower than that among those
receiving the placebo. Evaluation of the data led to
termination of the trial to make a known beneficial
treatment available more quickly to all patients.
In some research, such as the Tuskegee Syphilis Study or Nazi
medical experiments, participants experienced permanent
damage or died. In Dr. Cameron’s study (see Table 6-1), the
continued electroconvulsive therapy increased the damage.

Canada has adopted the ood Clinical Practice: The original statement listed eight guiding
Consolidated uidelines (Health Canada, 1997). principles, which are now subsumed under the
The collaboration of the three major funding three core principles respect for person,
agencies—the Canadian Institutes of Health concern for welfare, and justice. Of the
Research (CIHR), the Natural Sciences and Engi- others, respect for human dignity is articulated
neering Research Council of Canada (NSERC), through these three core principles respect for
and the Social Sciences and Humanities Research free and informed consent and respect for vul-
Council of Canada (SSHRC)—has led to a joint nerable persons are now re ected in the princi-
statement for the protection of human partici- ple of respect for persons, and respect for
pants. The revision of this document, the Tri- vulnerable persons is also re ected in the prin-
Council Policy Statement: Ethical Conduct for ciple of justice respect for privacy and confi-
Research Involving umans (CIHR et al., 2010), dentiality is part of concern for welfare and
offers a more inclusive approach to delineating respect for justice and inclusiveness is in the
current trends in ethical issues. This revised core principle of justice. The core principle
document, sometimes called the Tri-Council concern for welfare now also includes balanc-
Policy Statement-2, is henceforth referred to as ing harms and benefits, minimizing harm, and
TCPS2. maximizing benefit.
120 PART ONE Research Overview

(Carroll Ciaffa, 2008), which can raise ethical

Helpful Hint concerns about to the purpose of the engineering.
The qualitative researcher must be especially dili- If the purpose is to treat a disease, then the
gent in protecting the privacy and confidentiality of par-
ticipants. When the participants’ verbatim quotations research is ethically acceptable, but the existence
are used in the “Results” or “Findings” section of the of germline intervention raises more significant
research report to highlight the findings, the smallness ethical concerns, because risks will extend across
of the sample size may make it easy to identify an
individual participant. Moreover, when researchers and generations, magnifying the impact of unforeseen
REBs are engaging in naturalistic observation, the TCPS2 consequences (Carroll Ciaffa, 2008). Ethical
indicates that these boards “and researchers need to concerns also arise with regard to the privacy of
consider the methodological requirements of the pro-
posed research project and the ethical implications asso-
genetic information, mandatory testing of new-
ciated with observational approaches, such as the borns, and mandatory genetic screening.
possible infringement of privacy. They should pay close Other areas of research that engender much
attention to the ethical implications of such factors as discussion and controversy are fetal tissue
the nature of the activities to be observed, the environ-
ment in which the activities are to be observed, whether research and the use of women who are of child-
the activities are staged for the purpose of the research, bearing potential as participants in drug or thera-
the expectations of privacy that prospective participants peutic studies. The Tri-Council (i.e., the CIHR,
might have, the means of recording the observations,
whether the research records or published reports the NSERC, and the SSHRC) worked over a
involve identification of the participants, and any means period of several years to make stem cell research
by which those participants may give permission to be a reality in Canada, with the appropriate ethical
identified” (CIHR et al., 2010, p. 142).
guidelines for the use of embryos, fetuses,
gametes, and pluripotent stems cells. These
guidelines are detailed in the TCPS2 (CIHR et al.,
Current and Future Ethical Dilemmas 2010).
in Research Of note is the following information (CIHR
Ethics is the theory or discipline dealing with et al., 2010, p. 182, Article 13.2):
principles of moral values and moral conduct.
The ethical dilemmas in research for the twenty- Researchers conducting genetic research shall:
first century concern biotechnology, the use of (a) in their research proposal, develop a plan for
animals for research, and the creation of an orga- managing information that may be revealed
nizational culture that values and nurtures through their genetic research;
research ethics and the rights of people who (b) submit their plan to the REB; and
(c) advise prospective participants of the plan for
engage in research either as investigators or as managing information revealed through the
participants (Pranulis, 1996). For example, in research. (p. 182, Article 13.2)
only 12 years, the Human Genome Project, an
international research project launched by the In the past, women of childbearing potential were
nited States in 1988, provided a vast amount of denied participation in studies of a drug or poten-
data on DNA, including the molecular details tial therapy because of the unknown, potentially
about the DNA of more than 26 organisms. Since teratogenic effects of drugs and other therapies
then, the focus has been on studying (1) groups that were in various stages of testing. Guidelines
of genes rather than single genes (2) functions of related to the inclusion of pregnant women as
genes rather than structures of genes and (3) research participants have been even more strin-
proteomes, the proteins based on genes (Wheeler, gent than previous guidelines as a result, women
1999). To engage in genome research, genetic have been excluded from many important drug
engineering and genetic information must exist and research studies over the years. Similarly,
 CHAPTER 6 Legal and Ethical Issues 121

inclusion of ethnic minorities in federally funded Although Robb’s comments re ect the norms of
research studies is also a priority area for con- Victorian society, they also highlight a historical
sideration in research ethics (Julion, Gross, concern for ethical actions by nurses as health
Barclay-McLaughlin, 2000). care professionals (Robb, 1900).
The TCPS2 also has a well-articulated policy In Canada, most disciplines have developed
on the ethical guidelines for research with the their own code of ethics with guidelines for
Aboriginal population to ensure protection of the research. The CNA’s first document on ethical
rights of that community (CIHR, 2007). This principles related to nursing research, Ethical
guideline comprises 15 principles, which include uidelines for Nurses in Research Involving
the need for research when benefit is mutual, uman Participants, was released in 1983. It
incorporating the role of community elders in was revised in 1994 and 2002 and is now titled
consent and the responsibilities of the researcher Ethical Research uidelines for Registered
to understand and protect sacred knowledge. The Nurses (CNA, 2002).
inclusion of a separate chapter in TCPS2 outlin- Clearly, ignorance and na vet with regard
ing how to engage with and honour First Nations, to ethical and legal guidelines for the conduct
Inuit, and M tis communities re ects the work of research is never an excuse for a nurse’s
that was accomplished over several years to failure to be familiar with and act on behalf of
advance the need for equitable partnerships and patients, whose human rights must be safeguarded
to provide safeguards with these communities. It at all times. Nurse researchers are often among
is noted that Aboriginal entities at local, regional the most responsible and conscientious investiga-
and national levels have published and imple- tors in respecting the rights of human parti-
mented principles and codes governing research cipants. All nurses should be aware that in
practice—including ethical protections—that addition to the ethical research guidelines of
emphasize collective rights, interests and respon- the CNA, universities and hospitals may also
sibilities (CIHR, 2007, p. 106). have supplemental sets of ethical guidelines to
follow.
THE EVOLUTION OF ETHICS IN
NURSING RESEARCH PROTECTION OF HUMAN RIGHTS
The evolution of ethics in nursing research can be Human rights are the claims and demands that
traced back to 1897 and the constitution of the have been justified according to an individual or
Nurses’ Associated Alumnae Organization in the by a group of individuals. The term human rights
nited States. One of the first purposes of this is applied to the following five rights outlined in
organization was to establish a code of ethics for the CNA’s (2002) guidelines and linked to the
the nursing profession. In 1900, Isabel Hampton Tri-Council’s principles of respect for research
Robb wrote Nursing Ethics: or ospital and participants:
Private se In describing the moral laws by which 1. Right to self-determination
people must abide, she stated the following: 2. Right to privacy and dignity
3. Right to anonymity and confidentiality
Etiquette, speaking broadly, means a form of behavior 4. Right to fair treatment
or manners expressly or tacitly required on particular 5. Right to protection from discomfort and
occasions. It makes up the code of polite life and
harm
includes forms of ceremony to be observed, so that
we invariably find in societies that certain etiquette is These rights apply to everyone involved in a re-
required and observed either tacitly or by expressed search project, including research team members
agreement. involved in data collection, practising nurses
122 PART ONE Research Overview

involved in the research setting, and people par- BOX 6-3

ticipating in the study. As consumers of research ELEMENTS OF INFORMED CONSENT
read a research article, they must realize that any
1. A statement that the study involves research
issues highlighted in Table 6-2 should have been 2. An explanation of the purposes of the research,
addressed and resolved before a research study is delineating the expected duration of the subject’s
approved for implementation. participation
3. A description of the procedures to be followed
and identification of any procedures that are
Procedures for Protecting Basic experimental
4. A description of any reasonably foreseeable risks
Human Rights or discomforts to the subject
5. A description of any benefits to the subject or to
Informed Consent others that may reasonably be expected from the
Informed consent, illustrated by the ethical prin- research
ciples of respect and the related right to self- 6. A disclosure of appropriate alternative procedures
or course of treatment, if any, that might be
determination, is outlined in Box 6-3. Nurses advantageous to the subject
need to understand the elements of informed 7. A statement describing the extent to which the
consent to be knowledgeable participants when anonymity and confidentiality of the records
identifying the subject will be maintained
either obtaining informed consent from patients 8. For research involving more than minimal risk, an
or critiquing this process as it is presented in explanation as to whether any medical
research articles. treatments are available if injury occurs and, if so,
what they consist of or where further information
In or ed consent is the legal principle that
may be obtained
requires a researcher to inform individuals about 9. An explanation about whom to contact for
the potential benefits and risks of a study before answers to questions about the research and
the individuals can participate voluntarily. In researcher subjects’ rights and whom to contact in
the event of a research-related injury to the
theory, this principle governs the patient’s ability subject
to accept or reject individual medical interven- 10. A statement that participation is voluntary, that
tions designed to diagnose or treat an illness. refusal to participate will not involve any penalty
or less benefit to which the subject is otherwise
According to the TCPS2 (CIHR et al., 2010), free entitled, and that the subject may discontinue
and informed consent is at the heart of ethical participation at any time without penalty or loss
research and is a process of dialogue and informa- of otherwise entitled benefits
tion sharing to allow participants the choice to From Code of Federal Regulations: Protection of human subjects,
participate in research. Free and informed consent 45 CFR 46, OPRR Reports, revised March 8, 1983.
must be given without manipulation, undue in u-
ence, or coercion. the addition of any private information on mailed
For example, Akhtar-Danesh and associates documents.
(2010) developed a learning needs assessment No investigator may involve a human being as
questionnaire for community health nurses. a research participant until the legally effective
Approval for the use of human participants was informed consent has been obtained from either
obtained from appropriate ethics review boards the participant or a legally authorized representa-
before the data collection. The informed consent tive of the participant, and prospective partici-
form described risks and benefits of the study and pants must have time to decide whether to take
the participants’ responsibilities, and it mentioned part in a study. The researcher must not coerce
that their information would be confidential and the participant into taking part in the study, nor
anonymous. Confidentiality was ensured by the may researchers collect data on participants who
use of study identification numbers rather than have explicitly refused to take part in a study.
 CHAPTER 6 Legal and Ethical Issues 123

An ethical violation of this principle is illus- In a qualitative research study of perceptions

trated by the case of alush a v niversity of of space, time, and interprofessional practice on
Sas atche an et al (1965). In this landmark an acute stroke unit (see Appendix A), Senevi-
case, a university student volunteered for a ratne and associates (2009) obtained ethics
study testing a new anaesthetic, for which he approval from the appropriate REB. They used
would be paid 50.00. He consented to be in several information sessions and interviews to
the study based on the following information introduce the study to members in the stroke unit.
disclosed to him: the test would last a few During these sessions and interviews, the
hours, the test was safe and had been conducted researchers informed stroke unit members about
many times before, and the student had nothing client confidentiality. They obtained written
to worry about. He was informed that the pro- consent only from observation and interview par-
cedure would include placement of electrodes ticipants and verbal assent from clients when par-
on his arms, legs, and head and insertion of a ticipants were providing direct care (p. 1874).
catheter into a vein in his arm. He signed a
consent form releasing the physicians and the Helpful Hint
university from liability for any untoward effects Remember that research reports rarely provide
readers with detailed information regarding the degree
or accidents, which were explained to him as to which the researcher adhered to ethical principles,
falling down at home after the test (McLean, such as informed consent; this is because of space limi-
1996, p. 49). The test proceeded with adminis- tations in journals, which make it impossible to describe
tration of an untested anaesthetic, and the all aspects of a study. Failure to mention procedures to
safeguard participants’ rights does not necessarily mean
student suffered a cardiac arrest. He was uncon- that such precautions were not taken.
scious for four days in hospital and left with a
residual inability to concentrate. The physicians Most investigators obtain consent (agreement
and the university were found negligent for to participate in a study) through personal discus-
failing to disclose that there was risk involved sion with potential participants. This process
with the use of an anaesthetic and that this par- allows the person who is the potential participant
ticular drug had not been previously tested by to obtain immediate answers to questions.
them (McLean, 1996, p. 49). However, consent forms, written in narrative or
When composing an informed consent form, outline form, highlight elements that both inform
researchers must ensure that the language is and remind participants of the nature of the study
understandable. For example, the reading level and their participation (Dubler Post, 1998
should be no higher than grade 8 for adults, and Haggerty Hawkins, 2000). When one partici-
the use of technical research language should be pant is scheduled to participate in many inter-
avoided (Rempusheski, 1991). The elements that views, the participant must give rocess consent
need to be contained in an informed consent are (voluntary continued participation in a study,
listed in Box 6-3. Note that many institutions which can be verbal) for each data collection
require additional elements. Figure 6-1 is an point.
example of an informed consent form for a quan- Assurance of anonymity and confidentiality
titative study, whereas Figure 6-2 is an example (defined in Table 6-2), which is conveyed in
of an informed consent form for a qualitative writing, is sometimes difficult in unique research
study. Note that in each consent form, the ele- situations that capture the public’s attention.
ments of participation, risk and benefits, with- For example, when physicians at Loma Linda
drawal, confidentiality, and whom to contact for niversity Hospital in California transplanted a
further queries are clearly outlined. Te t continued on p
124 PART ONE Research Overview

CONSENT TO PARTICIPATE IN A RESEARCH STUDY

Title
VISUAL DIFFERENTIATION IN LOOK-ALIKE MEDICATION NAMES

University Health Network Principal Investigator

Tasmine Halevy, Clinical Director of Pharmacy, UHN, 416-XXX-XXXX

Study Principal Investigator

Monica Blum, Associate Professor, Reed University, 416-XXX-XXXX

Co-Investigators
Joyce Davis, Vice President, ISMP Canada
Dr. Mina D. Singh, Associate Professor, York University, Faculty of Health, School of Nursing
Ravinder Sharma, Human Factors Engineer, Red Forest Consulting
Dr. Irmgard Mirren, Psychiatrist, Child and Parent Resource Institute
Evan Ross, Chief Pharmacist, Child and Parent Resource Institute

Collaborator
Jude Hartman

Sponsors
Canadian Patient Safety Institute, ISMP Canada; Red Forest Consulting; Child and Parent Resource Institute; York
University Faculty of Graduate Studies, Department of Design and Faculty of Health

Introduction
You are being asked to take part in a research study. Please read this explanation about the study and its risks and
benefits before you decide if you would like to take part. You should take as much time as you need to make your
decision. You should ask the Principal Investigator or Research Assistants to explain anything that you do not understand
and make sure that all of your questions have been answered before signing this consent form. Before you make your
decision, feel free to talk about this study with anyone you wish. Participation in this study is voluntary.

Background and Purpose

This study will look at the visual display of look-alike medication names. It is hoped that this research will contribute to
the design of shelf labelling, packaging, and computer displays to help reduce instances of medication errors due to look-
alike medication names. The results from this research are intended to support health care workers in the safe delivery of
health care. You have been asked to take part in this research study because you have or may come into contact with
look-alike medications. A total of 130 to 135 nursing staff and 10 pharmacy staff from Princess Margaret Hospital, Toronto
General Hospital, Toronto Western Hospital will participate in this study.

Study Design
You will help us evaluate the best ways to display look-alike names for ease and accuracy in recognition and selection of
medications. If you choose to participate in this study, you will be asked to answer a short questionnaire to establish
your demographic information and to ask you your opinion of current practices related to the display of look-alike
medication names. You will participate in three experiments that emulate the selection of medications. The first two
experiments are screen based, and you will be asked to identify look-alike names on a laptop display. For the third
experiment, you will be asked to select medications from a series of baskets. The tasks will be explained thoroughly
before each experiment. Your commitment for this study will be one session lasting approximately 45 to 60 minutes.

FIGURE 6-1 Example of an informed consent form for a quantitative study. ISMP, Institute for
Safe Medication Practices; UHN, University Health Network.
 CHAPTER 6 Legal and Ethical Issues 125

CONSENT TO PARTICIPATE IN A RESEARCH STUDY—cont'd

Risks Related to Being in the Study
There are no known risks if you take part in this study, but you may refuse to answer questions or stop the experiments
at any time if there is any discomfort. Your responses to the questionnaires will not have an impact on your employment,
nor will they be shared with your supervisors or managers.

Benefits to Being in the Study

You will not receive any direct benefit from being in this study. Information learned from this study may help in the safe
delivery of health care.

Voluntary Participation
Your participation in this study is voluntary. You may decide not to be in this study, or to be in the study now and then
change your mind later. You may leave the study at any time without affecting your employment status. You may refuse
to answer any question you do not want to answer on the questionnaire by writing “pass,” or stop participating in the
experiment at any time.

Confidentiality
The information that is collected for the study will be kept in a locked and secure area at York University by the study
Principal Investigator for 10 years. Only the study team and the people or groups listed below will be allowed to look at
the data. All information collected during this study will be kept confidential and will not be shared with anyone outside
the study unless required by law. Any information about you that is collected for the study will have a code and will not
show your name or address, or any information that directly identifies you. You will not be named in any reports,
publications, or presentations that may come from this study. If you decide to leave the study, the information about you
that was collected before you left the study will still be used. No new information will be collected without your
permission. Representatives of the University Health Network Research Ethics Board may look at the study records to
check that the information collected for the study is correct and to make sure the study followed proper laws and
guidelines.

Questions about the Study

If you have any questions or concerns, or would like to speak to the study team for any reason, please call: Tasmine
Halevy, University Health Network, at 416-XXX- XXXX or Monica Blum, Reed University, 416-XXX-XXXX.

The research has been reviewed and approved by the University Health Network Research Ethics Board (REB) and the
Human Participants Review Committee (HPRC) at Reed University for compliance with senate ethics policy. If you have
any questions about your rights as a research participant or have concerns about this study, call the Chair of the UHN
(REB) or the Research Ethics office number at 416-XXX-XXXX or Manager, Research Ethics—Alicia Collins-Walker: 309
Elsevier Lanes, Reed University, 416-XXX-XXXX. The HPRC and REB are groups of people who oversee the ethical conduct
of research studies. These people are not part of the study team. Everything that you discuss will be kept confidential.

Consent
This study has been explained to me and any questions I had have been answered. I know that I may leave the study at
any time. I agree to take part in this study.

Print Study Participant’s Name Signature Date

(You will be given a signed copy of this consent form)

My signature means that I have explained the study to the participant named above. I have answered all questions

Print Name of Person Obtaining Consent Signature Date

FIGURE 6-1, cont’d Example of an informed consent form for a quantitative study.
126 PART ONE Research Overview

INFORMATION AND INFORMED CONSENT STATEMENT: INTERVIEW

Title
BETTER UNDERSTANDING HOW COUPLES COPE WITH A CHILD’S LIFE-THREATENING ILLNESS
Principal Investigator
Dr. Susan Cadell, Associate Professor; Director, Manulife Centre for Healthy Living; Lyle S. Hallman Faculty of Social Work,
Wilfrid Laurier University, 519-XXX-XXXX

Co-Investigators
Dr. Rosemary Stiles, School of Nursing, York University
Dr. Anna DeLaurentis, The Centre for Health and Coping Studies, University of British Columbia

Research Assistants
Matilde Negrini, Faculty of Social Work, Wilfrid Laurier University
Julian Millman, Faculty of Social Work, Wilfrid Laurier University
Nella Leone, Faculty of Social Work, Wilfrid Laurier University

Contact Person
Matthew Philips, Research Coordinator: 1-800-XXX-XXXX

We are inviting couples to participate in the next phase of this research study. The purpose of this study is to discover the
experience of spouses/partners who are together caring for a child with a life-limiting illness. This study is being
conducted by Dr. Susan Cadell, Associate Professor and Director of the Centre for Healthy Living at Wilfrid Laurier
University, and Co-Investigator on the Canadian Institutes for Health Research’s New Emerging Team (NET): Transitions in
Pediatric Palliative and End-of-Life Care.

Information
During the interview, you and your spouse/partner will be interviewed together. You will be asked questions about your
personal experience of caring for a child with a life-limiting condition, as well as questions about the role that each
spouse/partner plays in the coping of the other. The interview will take approximately 1.5 to 2 hours. The interview will
be conducted by a trained, sensitive interviewer and will take place at a location convenient to you. In order to make sure
that we have an accurate record of what you have shared during the interview, your interview will be recorded and
transcribed. All identifying information will be removed from the transcripts and only the investigators and research staff
will have access to them. The recordings and transcripts will be identified only by code number and stored in a locked
filing cabinet or secured information system. They will be stored for 5 years after the publication of the results from this
study. After 5 years, the recordings and the transcripts will be destroyed. The recordings will not be used for any other
purposes without your additional permission.

Phase Two of this study will involve the participation of approximately 15 to 20 couples who are together caring for their
child. Due to the nature of this study, it is possible that quotes from your interview may be used in publication. To
maintain confidentiality, all identifying information will be removed from the quotations. If a specific family or disease
characteristic is rare and could potentially be identifying, the information will be changed in the quote. Please indicate
your preference below regarding the use of your quotations.

Name: _____________________________ Name: _____________________________

® Yes – I can be quoted with no identifying information. ® Yes – I can be quoted with no identifying information.
® No – Please do not use quotes. ® No – Please do not use quotes.

FIGURE 6-2 Example of an informed consent form for a qualitative study.

 CHAPTER 6 Legal and Ethical Issues 127

INFORMATION AND INFORMED CONSENT STATEMENT: INTERVIEW—cont'd

Risk
This research project deals with a sensitive topic. The interviewer will monitor your distress level and will stop the
interviewing process if you become upset. The interviewer will then ensure that you are aware of your right not to
answer any questions asked and your right to terminate the interview at any time. If necessary, the interviewer will refer
you to appropriate services and resources to ensure your support needs are met.

Benefits
You may benefit from the ability to communicate your experiences of pediatric palliative care in a safe, nonjudgemental
setting. In addition, your participation may benefit other families, researchers, and policy makers in pediatric palliative
care by providing a better understanding of the caregiver experience.

Confidentiality
Confidentiality will be provided to the fullest extent possible by law. Your identity and the identity of all family members
will be kept strictly confidential. All identifying information will be removed from the data. All documents and record-
ings will be identified only by code number and the information will be retained in a secured information system and
locked filing cabinet. All identifying information will be kept separate from the data. All documents that are kept on a
computer will be password protected. Identifying information will not be emailed to anyone at any time. You will not
be identified by name in any reports of the completed study. Only study personnel will have access to the study data.

Compensation
For participating in this study you and your spouse/partner will each receive $30 at the beginning of the interview. If you
withdraw from the study after this point, you will still receive the full amount.

Participation
Your participation in this study is voluntary; you may decline to participate without penalty. If you decide to participate,
you may withdraw from the study at any time without penalty and without loss of benefits to which you are otherwise
entitled. If you withdraw from the study before data collection is completed, your data will be destroyed. You have the
right to omit any question(s)/procedure(s) you choose.

Feedback and publication

It is expected that the results of this study will be presented and published as a journal article. If you would like to be
notified of the results of the study, please indicate below.

Name: _____________________________ Name: _____________________________

® Yes – I would like to be notified of results. ® Yes – I would like to be notified of results.

Contact
If you have questions at any time about the study or the procedures, or if you experience adverse effects as a result of
participating in this study, you may contact the Research Coordinator, Matthew Philips, at 1-800-XXX- XXXX. This project
has been reviewed and approved by the University Research Ethics Board at Wilfrid Laurier University. If you feel you have
not been treated according to the descriptions in this form, or your rights as a participant in research have been violated
during the course of this project, you may contact Dr. Mark Billingsley, Chair, University Research Ethics Board, Wilfrid
Laurier University, 519-XXX-XXXX.

Consent
I have read and understand the above information. I have received a copy of this form. I agree to participate in this study.

Participant's signature Date

Participant's signature Date

Investigator's signature Date

FIGURE 6-2, cont’d Example of an informed consent form for a qualitative study.
128 PART ONE Research Overview

baboon’s heart into a 2-week-old infant, her iden- or amend its guidelines in other ways. The REB
tity was hidden anony ity as she was known makes the final determination regarding the most
only as Baby Fae, and con dentiality was appropriate documentation format. Research con-
ensured in that the reports could not be linked to sumers should note whether and what kind of
her and her family. Maintaining anonymity and evidence of informed consent has been provided
confidentiality is particularly important for quali- in a research article.
tative researchers because the researcher often
functions as the data collection instrument and Helpful Hint
meets the participant. The consent form must be Note that researchers often do not obtain written
signed and dated by the participant. The presence informed consent when the major means of data col-
lection is through self-administered questionnaires.
of witnesses is not always necessary but does Implied consent is usually assumed in such cases; in
constitute evidence that the participant concerned other words, the return of the completed questionnaire
actually signed the form. In cases in which the reflects the respondent’s voluntary consent to
participate.
participant is a minor or is physically or mentally
incapable of signing the consent, the signature
must be obtained from a legal guardian or repre- Research Ethics Boards
sentative. The investigator also signs the form to esearch ethics boards E s are panels that
indicate commitment to the agreement of ano- review research projects to assess whether ethical
nymity and confidentiality. standards are met in relation to the protection of
In Jack and associates’ (2005) study, the par- the rights of human participants. Such boards are
ticipants’ anonymity and confidentiality were established in agencies to review biomedical and
guaranteed, but in cases in which the researcher behavioural research involving human subjects
suspected child abuse or neglect, the participants within the agency or in programs sponsored by
were clearly informed that the researcher has a the agency. niversities, hospitals, and other
legal responsibility to report any suspicions to the health agencies applying for a grant or contract
child welfare agency. Another strategy that can be for any project or program that involves the
used to ensure confidentiality is to ask the tran- conduct of biomedical or behavioural research
scribers in a qualitative study to sign a confiden- with human participants are required by the Tri-
tiality agreement, as Heaman and colleagues Council and most funding agencies to submit
(2007) did in their study of relationship work in with their application assurances that they have
an early childhood home-visiting program. established an REB that reviews the research
In general, the signed informed consent form projects and protects the rights of the human par-
is given to the participant. The researcher should ticipants (CIHR et al., 2010). The Tri-Council
also keep a copy. Some research, such as a retro- also requires that the REB have at least five
spective chart audit, may require only institu- members, including both men and women. Mem-
tional approval, not informed consent. In some bership must include at least two professionals
cases, when minimal risk is involved, the inves- who have expertise in relevant research disci-
tigator may have to provide the participant only plines, fields, and methodologies covered by the
with an information sheet and a verbal explana- REB at least one who is knowledgeable in ethics
tion. In other cases, such as a volunteer conve- at least one who is knowledgeable in the relevant
nience sample, completion and return of research law (but that member should not be the institu-
instruments constitute evidence of consent. The tion’s legal counsel or risk manager) and at least
REB advises on exceptions to these guidelines, one who is a community member and has no
as in cases in which the REB might grant waivers affiliation with the institution but is recruited
 CHAPTER 6 Legal and Ethical Issues 129

from the community served by the institution automatically exempt the researcher from obtain-
(CIHR et al., 2010). ing informed consent.
The REB is responsible for protecting partici- Not all research requires an ethical review. To
pants from undue risk and loss of personal rights follow protocol, researchers can submit a pro-
and dignity. For a research proposal to be eligible posal to their own REB however, according to
for consideration by an REB, it must already have the TCPS2 (CIHR et al., 2010), this step is not
been approved by a departmental review group, necessary when the research relies exclusively on
such as a nursing research committee, that attests information that is legally accessible to the public
to the proposal’s scientific merit and congruence and appropriately protected by law or that is pub-
with institutional policies, procedures, and licly accessible and there is no reasonable expec-
mission. The REB reviews the study’s protocol to tation of privacy. Legally accessible information
ensure that it meets the requirements of ethical includes registries of deaths, court judgements, or
research that appear in Box 6-4. Most boards public archives and publicly available statistics
provide guidelines or instructions for researchers (e.g., Statistics Canada public use files).
that include steps to be taken to receive REB REB review is also not required when research-
approval. For example, guidelines for writing a ers use exclusively publicly available information
standard consent form or criteria for qualifying that may contain identifiable information and for
for an expedited rather than a full REB review which there is no reasonable expectation of
may be available. The REB has the authority to privacy. For example, identifiable information
approve research, require modifications, or disap- may be disseminated in the public domain through
prove a research study, on the basis of the guide- print or electronic publications film, audio, or
lines outlined in Box 6-5. A researcher must digital recordings press accounts official publi-
receive REB approval before beginning to cations of private or public institutions artistic
conduct research. REBs have the authority to installations, exhibitions, or literary events freely
suspend or terminate approval of research that is open to the public or publications accessible in
not conducted in accordance with REB require- public libraries. Research that is nonintrusive and
ments or that has been associated with unexpected does not involve direct interaction between the
serious harm to participants. REB approval was researcher and individuals through the Internet
obtained from the niversity of British Columbia also does not require REB review. Online mate-
and from the British Columbia Women’s Hospital rial such as documents, records, performances,
for a study to understand breast-feeding and online archival materials, or published third-party
infant health in Canada (Chen, 2010). interviews to which the public is given uncon-
REBs in Canada also provide for reviewing trolled access on the Internet and for which there
research in an expedited manner when the risk to is no expectation of privacy is considered to be
research participants is minimal. An expedited publicly available information.
review usually shortens the length of the review
process. Although a researcher may determine
that a project involves minimal risk, however, the Adapted from Canadian Institutes of Health Research,
research cannot be conducted until the REB Natural Sciences and Engineering Research Council of
makes the final determination. Many qualitative Canada, Social Sciences and Humanities Research
Council of Canada. (2010, December). Tri-Council policy
research projects are eligible for an expedited
statement: Ethical conduct for research involving humans
review because the research is noninvasive and Chapter 2, Scope and approach. Retrieved from http://
involves methods such as observation, interviews, www.pre.ethics.gc.ca/pdf/eng/tcps2/TCPS 2 FINAL
and questionnaires. An expedited review does not Web.pdf
BOX 6-4
CANADIAN NURSES ASSOCIATION’S ETHICAL RESEARCH GUIDELINES FOR REGISTERED NURSES
These guidelines are the primary values in the Canadian GUIDELINE 4: PRESERVING DIGNITY
Nurses Association’s (2008) Code of Ethics for Registered Implementation
Nurses, and they form the structural framework for the Nurses demonstrate equal respect for persons who
Ethical Research Guidelines for Registered Nurses choose to become research subjects and for those who
(Canadian Nurses Association [CNA], 2002). Below choose not to participate (CNA, 2002, p. 12).
each guideline are examples of principles for Nurse should respect the process by which
implementation. communities determine whether and under what
conditions research can be conducted (e.g., First Nations
GUIDELINE 1: PROMOTING SAFE, COMPASSIONATE, communities) (CNA, 2002, p. 13).
COMPETENT, AND ETHICAL CARE
Implementation GUIDELINE 5: MAINTAINING PRIVACY
AND CONFIDENTIALITY
Nurses engaged in research must comply with the Code
of Ethics for Registered Nurses (CNA, 2008) and conduct Implementation
themselves with honesty and integrity in all their Nurses caring for persons involved in research must be
interactions with research subjects and research attentive to the subject’s privacy and exercise caution in
colleagues (CNA, 2002, p. 6). use, access, collection, and disclosure of information.
Nurses involved as principal investigators, clinical They should be aware of relevant provincial legislation
research coordinators or research assistants must base about the confidentiality of health and research
their research on relevant knowledge of research information (CNA, 2002, p. 14).
methods and continue to acquire new skills and In all cases where research data must be released,
knowledge to develop and maintain their level of nurses should release only the minimum amount of
competence in research (CNA, 2002, p. 6). data required and restrict the number of people to
whom data is released (CNA, 2002, p. 14).
GUIDELINE 2: PROMOTING HEALTH AND WELL-BEING GUIDELINE 6: PROMOTING JUSTICE
Implementation Implementation
Nurses engaged in nursing practice must hold people’s Nurses must seek to ensure that all persons have access
optimum health and well-being as first and foremost in to opportunities, subject to availability, to be involved as
their interactions with all those they serve (CNA, 2002, research subjects (CNA, 2002, p. 15).
p. 7). Nurses should promote participatory research where
Nurses should recognize the importance of bringing research subjects can work in partnerships with
a nursing perspective to health research and engage researchers in design, implementation and
with other health professionals in interdisciplinary health dissemination of research (CNA, 2002, p. 15).
research promoting health and well-being (CNA, 2002,
p. 8). GUIDELINE 7: BEING ACCOUNTABLE
Implementation
GUIDELINE 3: PROMOTING AND RESPECTING INFORMED Nurses engaged in research must conduct the research
DECISION MAKING within their own level of competence. Even if others
Implementation assume that nurses know and should be able to
Nurses caring for people involved in research should do perform certain research interventions required in the
their part in ensuring that consent is free and informed. study, nurses must advise researchers and clinical staff if
The investigator, or the individual designated by the they do not feel they have the level of competence
investigator, should fully inform the subject of all needed to perform these tasks. If placed in such a
pertinent aspects of the study, including the type and position, they should seek information and help from
level of commitment required and potential benefits investigators and other knowledgeable researchers
and risks (CNA, 2002, p. 9). (CNA, 2002, p. 17).
Nurses caring for persons must be alert to any signs Nurses engaged in research and nurse educators
that these individuals feel pressured or coerced into should seek to ensure their students learn about
participating in a research study. If they suspect that research design, research ethics and the need for ethics
these individuals feel pressured or coerced, nurses must approval. They should also help students be aware of
advise the investigator and/or the agency’s REB (CNA, their level of competence in conducting research (CNA,
2002, p. 9). 2002, p. 17).
From Canadian Nurses Association. (2002b). Ethical research guidelines for registered nurses. Ottawa: Author. Canadian Nurses Association. (2008).
Code of Ethics for Registered Nurses. Ottawa: Author. Retrieved April 1, 2010, from http://www.cna-nurses.ca/CNA/documents/pdf/publications/
Code_of_Ethics_2008_e.pdf. © Canadian Nurses Association. Reprinted with permission. Further reproduction prohibited.
REB, research ethics board.
 CHAPTER 6 Legal and Ethical Issues 131

BOX 6-5 researchers should consult the REB before plan-

PARTIAL GUIDELINES FOR RESEARCH ETHICS ning research that potentially involves an over-
BOARD APPROVAL OF RESEARCH STUDIES subscribed research population, such as patients
To approve research, the REB must determine that the who have undergone organ transplantation,
following guidelines have been satisfied: patients with AIDS, or captive and convenient
1. There is an analysis of balance and distribution research populations, such as prisoners. The use
of harms and benefits.
of special populations does not preclude under-
2. There is a proportionate approach based on the
general principle that the level of review is taking research safeguards must be undertaken,
determined by the level of risk presented by the however, to protect the rights of these participants
research: the lower the level of risk, the lower (CIHR et al., 2010). Davis (1981) reminded us
the level of scrutiny (delegated review); the
higher the level of risk, the higher the level of that a society can be judged by the way it treats
scrutiny (full board review). its most vulnerable people this point is worth
3. There is a formal informed consent process. remembering in research that involves children,
From Canadian Institutes of Health Research, Natural Sciences and older adults, and other vulnerable groups.
Engineering Research Council of Canada, & Social Sciences and Pediatric research can be particularly problem-
Humanities Research Council of Canada. (2010, December).
Tri-Council policy statement: Ethical conduct for research involving atic. Mitchell (1984) discussed the .S. National
humans. Retrieved from http://www.pre.ethics.gc.ca/pdf/eng/ Commission’s concept of assent versus consent
tcps2/TCPS_2_FINAL_Web.pdf
REB, research ethics board. in regard to pediatric research. Assent—an aspect
of informed consent that pertains to protecting the
rights of children as research subjects—is com-
The CIHR Web site has a complete listing of posed of the following three fundamental
updated information concerning ethical guide- elements:
lines involving human participants. In addition, 1. A basic understanding by the child of what
all researchers should consult their agency’s the child will be expected to do and what
research offices to ensure that the application will be done to the child
being prepared for REB approval adheres to the 2. A comprehension by the child of the basic
most current requirements. Nurses who are cri- purpose of the research
tiquing published research should be familiar 3. An ability of the child to express a prefer-
with current regulations to determine whether ence regarding participation
ethical standards have been met. The Critical An example of research in a pediatric group in
Thinking Decision Path illustrates the ethical which assent was used is the study of Rennick
decision-making process an REB might use in and colleagues (2004). These researchers were
evaluating the risk-benefit ratio of a research interested in identifying children at high risk for
study. psychological problems after hospitalization in a
pediatric intensive care unit. The children were
Protecting the Basic Human Rights of categorized as being at either high or low risk for
Vulnerable Groups developing persistent psychological sequelae on
Researchers are advised to consult their agency’s the basis of illness severity and the number of
REB for the most recent guidelines when consid- invasive procedures that they underwent. The
ering research involving vulnerable groups, such children gave assent verbally or in writing, and
as older adults, children, pregnant women, unborn their parents provided written consent for their
children, persons who are emotionally or physi- children to participate.
cally disabled, prisoners, deceased persons, stu- In contrast to assent, consent requires a
dents, and persons with AIDS. In addition, relatively advanced level of cognitive ability.
132 PART ONE Research Overview

CRITICAL THINKING DECISION PATH

Evaluating the Risk-Benefit Ratio of a Research Study

Evaluate the sampling methods,

consent process, and outcomes
of the study

Assess Assess
benefits risks

Risk-benefit ratio

Risks outweigh Benefits outweigh

benefits risks

Potentially Ethical
unethical study
study

Informed consent re ects competency standards permission. When children reach maturity, usually
requiring abstract appreciation of and reasoning at 18 years of age in the case of research, they
about the information provided. The issue of may render their own consent. They may do so at
assent versus consent is an interesting one. For a younger age if they have been legally declared
example, at what age can children be expected to emancipated minors. uestions regarding assent,
make meaningful decisions about participating in consent, and the age of the individual should be
research In terms of the work by Jean Piaget addressed by the REB or research administration
about cognitive ability, children aged 6 years and office and not left to the discretion of the
older can participate in giving assent. Children researcher to answer.
age 14 years and older, although not legally autho- Special ethical considerations also exist when
rized to give sole consent unless they are emanci- research is conducted with older adults. As an
pated minors, can make such decisions as capably advocate for vulnerable older adults who are
as adults (Mitchell, 1984 Piaget, 1937/1954). increasingly dependent on other people for care
If the research involves more than minimal and whose cognitive ability is declining, the
risk and does not offer a direct benefit to the American Geriatrics Society Ethics Committee
individual child, then both parents must grant (1998) stated that older adults are precisely the
 CHAPTER 6 Legal and Ethical Issues 133

class of persons who were historically and are Researchers and patient caregivers involved in
potentially vulnerable to abuse and for whom research with vulnerable people are well advised
specific legal protection is needed. The issue of to seek advice from appropriate REBs, clinicians,
the legal competence of older adults is often lawyers, ethicists, and other professionals. In all
raised (Flaskerud Winslow, 1998), but no issue cases, the burden should be on the investigator to
exists if the potential participant can supply show the REB that it is appropriate to involve
legally effective informed consent. Competence vulnerable participants in research.
is not clearly measurable. The complexity of the
study may affect an individual’s ability to consent Helpful Hint
to participate. The capacity to give informed Keep in mind that researchers rarely mention
consent should be assessed in each individual explicitly that the study participants were vulnerable
participants or that special precautions were taken to
for each research protocol being considered appropriately safeguard the human rights of this vulner-
(American Geriatrics Society Ethics Committee, able group. Research consumers need to be attentive to
1998). For example, an older person may be able the special needs of individuals who may be unable to
act as their own advocates or are unable to adequately
to consent to participate in a simple observation assess the risk-benefit ratio of a research study.
study but not in a clinical drug trial.
The issue of the necessity of requiring the
older adult to provide consent often arises. N. N. RESEARCH INVOLVING
Dubler (personal communication, 1993) referred ABORIGINAL PEOPLE
to research regulations that indicate that some or When developing ethical guidelines, attention is
all of the elements of informed consent may be paid to the culture and traditions of Aboriginal
waived for the following reasons: people in Canada. To this end, the Tri-Council
1. The research involves no more than minimal (CIHR et al., 2010) developed the following
risk to the participants. good practices for researchers and REBs to
2. The waiver or alteration will not adversely consider when engaged in research (CIHR, 2007,
affect the rights and welfare of the partici- Section 6):
pants. To respect the culture, traditions, and
3. The research could not feasibly be carried knowledge of the Aboriginal group
out without the waiver or alteration. To conceptualize and conduct research with
No vulnerable population may be singled out the Aboriginal group as a partnership
for study merely for convenience. For example, To consult members of the group who have
neither people with mental illness nor prisoners relevant expertise
may be studied simply because they are available To involve the group in the design of the
and their presence is convenient. Prisoners may be project
studied if the study pertains to them for example, To examine how the research may be shaped
studies concerning the effects and processes to address the needs and concerns of the
of incarceration. Similarly, people with mental group
illness may participate in studies that focus on To make best efforts to ensure that the
expanding knowledge about psychiatric disorders emphasis of the research, and the ways
and treatments. Students also are often a conve- chosen to conduct it, respect the many view-
niently available group. They must not, however, points of different segments of the group in
be singled out as research participants because of question
convenience the research questions must have To provide the group with information
some bearing on their status as students. respecting the following:
134 PART ONE Research Overview

Protection of the Aboriginal group’s cul- on the basis of their own interests rather than the
tural estate and other property needs of the participants ( Ownership, control,
The availability of a preliminary report access, and possession’ 2003).
for comment
The potential employment by research- SCIENTIFIC FRAUD AND MISCONDUCT
ers of members of the community appro-
priate and without prejudice Fraud
Researchers’ willingness to cooperate Periodically, articles reporting the unethical
with community institutions actions of researchers appear in the professional
Researchers’ willingness to deposit data, and lay literature. Data may have been falsified
working papers and related materials in or fabricated, or participants may have been
an agreed-upon repository coerced into participating in a research study
To acknowledge in the publication of the (Kevles, 1996 Office of Research Integrity, .S.
research results the various viewpoints of Department of Health and Human Services, 2011
the community on the topics researched Tilden, 2000). In a climate of publish or perish
To afford the community an opportunity to in academic and scientific settings and declining
react and respond to the research findings research dollars, academics and scientists are
before the completion of the final report, in under increasing pressure to produce significant
the final report or even in all relevant research findings. Job security and professional
publications recognition are coveted, essential, and often pred-
Smith and colleagues (2006), in using a par- icated on being a productive scientist and a pro-
ticipatory action research approach, were able to lific writer. These pressures have been known to
acknowledge many of the aforementioned con- overpower some people, who then take shortcuts,
siderations by involving leaders of the commu- fabricate data, and falsify findings to advance
nity and other key informants in the research their positions (Rankin Esteves, 1997 Tilden,
design. In some cases, however, these good 2000).
practices are not fully implemented. For The risks of engaging in fraudulent research
example, Smylie and associates (2005) raised are many, including harming research partici-
concerns about ethical issues in certain studies in pants or basing clinical practice on false data. As
which community members were not consulted advocates of patient welfare and professional
during the formulation of the study and cautioned practice, nurses should be aware that sometimes
researchers that using Aboriginality as a social they might observe or suspect a researcher’s mis-
construct is not paying attention to the heteroge- conduct. In such cases, nurses must contact the
neity in the various groups. appropriate group, such as the REB, to ensure that
To understand the evidence being collected, this matter receives appropriate attention and
and to make informed choices about what First review.
Nations leaders need to improve the social and
economic conditions of their people, an ethical Misconduct
framework, Ownership, Control, Access, and Of equal importance is the issue of basing nursing
Possession (OCAP), was developed by the First practice on reports that appear in journals when
Nations Statistics Initiative. This initiative is subsequent research and reports on those partici-
important because a perceived mistrust exists pants change the scientific basis for practice. Cor-
between First Nations people and researchers, rections or further research in follow-up reports
and researchers choose participants and studies may be buried, obscure, or underreported. As
 CHAPTER 6 Legal and Ethical Issues 135

patient advocates and research consumers, nurses they initiate any form of clinical testing. Medical
must keep up to date on scientific reports related devices are classified according to the extent of
to nursing practice and must adjust their practice control necessary to ensure the safety and effec-
as directed by ever-evolving, evidence-informed tiveness of each device.
research findings. In addition, researchers have a
responsibility to keep current with federal com-
pliance regulations on prevention, detection, LEGAL AND ETHICAL ASPECTS OF
investigation, and adjudication of scientific ANIMAL EXPERIMENTATION
misconduct. The laws that have been written regarding ani al
rights—guidelines used to protect the rights of
animals in the conduct of research—in research
Unauthorized Research emanate from an interesting history of attitudes
At times, ad hoc or informal and unauthorized toward animals and the value that people place
research is conducted, including roduct testing on them. In 1963, the Medical Research Council
(the testing of medical devices). Although the of Canada (now CIHR) requested that a commit-
testing may seem harmless, it is, again, not the tee be established to investigate the care and use
purview of the investigator to make that determi- of experimental animals. The Canadian Council
nation. Nurses must carefully avoid being on Animal Care (CCAC) was formed and became
involved in unauthorized research for a number a nonprofit, autonomous, and independent body
of reasons, including the following (Raybuck, in 1982 (CCAC, 2005). It is now funded by the
1997): CIHR and the NSERC and conducts assessment
These treatments or methods of care are visits to each institution every 3 years, often
usually not monitored as closely for untow- unannounced. The CCAC requires that institu-
ard effects hence, the patient may be tions conducting animal-based research, teach-
exposed to unwarranted risk. ing, or testing establish an animal care committee
Patients’ rights to informed consent in clini- and that this committee be functionally active.
cal trials are not protected. The CCAC has a detailed guide regarding devel-
The success or failure of these unrecorded oping terms of reference for animal care commit-
trials contributes nothing to the organized tees (CCAC, 2006).
scientific knowledge of the efficacy or com- The CIHR scrutinized the proposed amend-
plications of the treatment. ments to the Cruelty to Animal Provisions of
The lack of independent quality supervision the Criminal Code of Canada, Bill-C15. The
allows deviations from the adopted experi- objective of these changes is to strengthen but
mental program that may eliminate the pro- simplify the existing penal code and to
gram’s effectiveness. enhance the effectiveness of the offence provi-
sions for clearly abusive, brutal and cruel
Product Testing treatment of animals. The CIHR supported
Nurses are often approached by manufacturers to this objective in principle and, with the
test products on patients. Often, nurses assume NSERC, prepared a joint submission to the
the role of research coordinator in clinical drug House of Commons Standing Committee on
or product trials (Raybuck, 1997). Consequently, Justice and Human Rights in the fall of 2001,
nurses should be aware of the Health Products recommending amendments to clarify certain
branch guidelines (see Health Canada, n.d.) and provisions of the bill with regard to their
regulations for testing of medical devices before application to health research.
136 PART ONE Research Overview

This section serves only as an introduction to Consent was provided by the gamete
the concept of legal and ethical issues related to donors
animal experimentation. Principles of protection Embryos exposed to manipulations not
of animal rights in research have evolved over directed specifically to their ongoing
time. Animals, unlike humans, cannot give normal development will not be trans-
informed consent, but other conditions related to ferred for continuing pregnancy
their welfare must not be ignored. Nurses who Research involving embryos will take
encounter the use of animals in research should place only during the first 14 days after
be alert to their rights. their formation by combination of the
gametes, excluding any time during
which embryonic development has been
RESEARCH INVOLVING HUMAN suspended (Article 12.8)
GAMETES, EMBRYOS, OR FETUSES Research involving a fetus or fetal tissue:
Research on the human genome and other repro- Requires the consent of the woman
ductive issues have caused much ethical debate Should not compromise the woman’s
and concern thus, the Tri-Council (CIHR et al., ability to decide whether to continue her
2010) developed pertinent guidelines, as demon- pregnancy (Article 12.9)
strated by these examples: Nurses working in labour rooms, especially
Materials related to human reproduction for those being required to assist with embryonic
research use shall not be obtained through research, should be aware of these ethical issues.
commercial transaction, including exchange To protect participant and institutional privacy,
for services (Article 12.6). the locale of the study frequently is described in
Research on in vitro embryos already general terms in the report’s subsection on the
created and intended for implantation to sample. For example, the article might state that
achieve pregnancy is acceptable if: data were collected at a 500-bed tertiary care
The research is intended to benefit the centre in Ontario, without mentioning the centre’s
embryo name. Protection of participant privacy may be
Research interventions will not compro- explicitly addressed by statements indicating that
mise the care of the woman or the sub- the anonymity or confidentiality of the data was
sequent fetus maintained or that grouped data were used in the
Researchers closely monitor the safety data analysis.
and comfort of the woman and the safety Determining whether participants were sub-
of the embryo jected to physical or emotional risk is often accom-
Consent was provided by the gamete plished indirectly by evaluating the study’s
donors (Article 12.7) Methods section. The reader evaluates the ris
Research involving embryos that have been bene t ratio that is, the extent to which the ben
created for reproductive or other purposes e ts of the study—the potential positive outcomes
permitted in Canada under the Assisted of participation in a research study—are maxi-
Human Reproduction Act (Assisted Human mized and the ris s—the potential negative out-
Reproduction Canada, 2004), but are no comes of participation in a research study—are
longer required for these purposes, may be minimized, so that participants are protected from
ethically acceptable if: harm during the study (Dubler Post, 1998
The ova and sperm from which they are Pruchino Hayden, 2000). The Practical Appli-
formed were obtained in accordance cation boxes list examples of how researchers
with Article 12.7 attempt to protect study participants from harm.
 CHAPTER 6 Legal and Ethical Issues 137

APPRAISING THE EVIDENCE

The Legal and Ethical Aspects of a Research Study
Research articles and reports often do not contain detailed who is asked to serve as a member of an REB will find
information regarding either the degree to which or all of the critiquing criteria useful in evaluating the legal and
the ways in which the investigator adhered to the legal ethical aspects of the research proposal.
and ethical principles presented in this chapter. Space Information about the legal and ethical considerations
considerations in articles preclude extensive documenta- of a study is usually presented in the “Methods” section of a
tion of all legal and ethical aspects of a research study. research report, probably in the subsection on the sample
Lack of written evidence regarding the protection of or data collection methods. The author most often indi-
human rights does not imply that appropriate steps were cates in a few sentences that informed consent was
not taken. obtained and that approval from an REB or similar commit-
The Critiquing Criteria box provides guidelines for tee was granted. A manuscript without such a discussion
evaluating the legal and ethical aspects of a research will probably not be accepted for publication; thus, it is
report. Although research consumers reading a research almost impossible for unauthorized research to be pub-
report will not see all areas explicitly addressed in the lished. Therefore, when a research article provides evidence
research article, they should be aware of them and of having been approved by an external review committee,
should determine that the researcher has addressed them the reader can feel confident that the ethical issues raised by
before gaining REB approval to conduct the study. A nurse the study have been thoroughly reviewed and resolved.

CRITIQUING CRITERIA

1. Was the study approved by an the purpose and nature of the 7. Were participants coerced or
REB or other agency committee study? unduly influenced to participate
members? 5. Were the participants or their in this study? Did they have the
2. Is there evidence that informed representatives informed about right to refuse to participate or
consent was obtained from all any potential risks that might withdraw without penalty?
participants or their result from participation in the Were vulnerable participants
representatives? How was it study? used?
obtained? 6. Was the research study 8. Were appropriate steps taken to
3. Were the participants protected designed to maximize the safeguard the privacy of
from physical or emotional benefit or benefits and to participants? How have data
harm? minimize the risks to human been kept anonymous or
4. Were the participants or their participants? confidential?
representatives informed about

Practical Application
Practical Application Morse and associates (2000) investigated the
Chalmers and colleagues (2002) reported the efficiency and effectiveness of approaches to
attitudes, beliefs, and personal behaviours of nasogastric tube insertion during trauma care. To
baccalaureate students in regard to tobacco use. The conduct this study, the procedure was videotaped.
researchers adhered to the principles of informed Approval was granted by the local REB, and all staff,
consent and confidentiality and ensured that the patients, and visiting support individuals were
team remained sensitive to the issues of power approached for consent to participate. Extra care was
differences between the students and the faculty taken for the preservation of anonymity; all
engaged in the study. In addition, the students were identifying information was removed, and any
reassured that their participation or nonparticipation mentions of names, addresses, or other identification
in the study would not affect their education. were erased from the videotapes.
138 PART ONE Research Overview

The obligation to balance the risks and benefits chairperson of the REB. In cases in which ethical
of a study is the responsibility of the researcher. considerations in a research article are in ques-
However, the research consumer reading a tion, clarification from a colleague, agency, or the
research report also should be confident that par- researcher’s REB is indicated. Nurses should
ticipants have been protected from harm. pursue their concerns until they are satisfied that
When considering the special needs of vulner- the patient’s rights and their rights as profession-
able participants, research consumers should be als are protected.
sensitive to whether the investigators have
addressed the special needs of individuals who
are unable to act on their own behalf. For example,
has the right of self-determination been addressed
CRITICAL THINKING CHALLENGES
by the informed consent protocol identified in the
research report Schell and associates (2010) ■ As part of a needs assessment for future health
conducted a study to compare upper arm and calf care delivery planning, the Ministry of Health is
automatic blood pressures in a convenience interested in determining the number of babies
sample of 221 children, aged 1 to 8 years, admit- infected with the human immunodeficiency virus
ted to a pediatric intensive care unit of a 180-bed (HIV). A province-wide study is funded that will
teaching hospital. Informed consent was obtained include the testing of all newborns for HIV, but
the mothers will not be told that the test is being
from the parent or guardian for all enrolled par-
done, nor will they be told the results. Using the
ticipants. Informed assent was obtained from
basic ethical principles found in Box 6-2, defend
children aged 7 and 8 years, if appropriate. or refute this practice.
When qualitative studies are reported, verba- ■ The REB of your health care agency does not
tim quotations from informants often are incor- include a nurse, and you think it should. You
porated into the Findings section of the article. discuss this matter with your supervisor, who
In such cases, the reader will evaluate how effec- states that including a nurse is not necessary
tively the author protected the informant’s iden- because the REB uses strict guidelines. What
tity, either by using a fictitious name or by essential arguments and explanations should
withholding information such as age, gender, your proposal address for including a nurse on
occupation, or other potentially identifying data. your institution’s REB?
Although the need for guidelines for the use of ■ A qualitative researcher intends to conduct a
phenomenological study on caring and to recruit
human and animal participants in research is
informants who are severely and persistently
evident and the principles themselves are clear,
mentally ill and attend an outpatient clinic. The
many instances arise in which nurses must use REB denies the study, indicating that informed
their best judgement, as both patient advocates consent cannot be obtained and that these
and researchers, when evaluating the ethical patients will not be able to tolerate an interview.
nature of a research project. In any research situ- What assumptions have the members of this REB
ation, the basic guiding principle of protecting the made? If you were the researcher and you were
patient’s human rights must always apply. When given the opportunity to address their concerns,
con icts arise, nurses must feel free to raise suit- what would you say? Include information from
able questions with appropriate resources and Table 6-2.
personnel. In an institution, raising questions may ■ How do you see computer electronic databases
include contacting the researcher first then, if and Web sites assisting researchers in conducting
ethical studies? Do you think that REBs can use
there is no resolution, the matter must be raised
this technology to assist them in their goals?
with the director of nursing research and the
 CHAPTER 6 Legal and Ethical Issues 139

REFERENCES
KEY POINTS Akhtar-Danesh, N., Valaitis, R. K., Schofieled, R.,
nderwood, J., Martin-Misener, R., Baumann, A.,
• Ethical and legal considerations in research first Kolotylo, C. (2010). A questionnaire for assessing
received attention after World War II, during the
community health nurses’ learning needs. estern
Nuremberg trials, which resulted in the
development of the Nuremberg code. This code ournal of Nursing Research, (8), 1055-1072.
became the standard for research guidelines American Geriatrics Society Ethics Committee. (1998).
protecting the human rights of research participants. Informed consent for research on human subjects
• The U.S. National Research Act, passed in 1974, with dementia. ournal of the American eriatric
created the National Commission for the Protection Society, , 1308-1310.
of Human Subjects of Biomedical and Behavioral American Nurses Association. (1985). Code for nurses
Research. The findings, contained in the Belmont ith interpretive statements. Kansas City, MO:
Report (National Commission for the Protection of Author.
Human Subjects of Biomedical and Behavioral Assisted Human Reproduction Canada. (2004). The
Research, 1978), are discussed with regard to the
Assisted uman Reproduction Act A R Act .
three basic ethical principles of respect for persons,
beneficence, and justice that underlie the conduct Retrieved from http://www.ahrc-pac.gc.ca/v2/aaa-app/
of research involving human participants. U.S. wwr-qnr/ahra-alpa/ahra-lspa-eng.php.
federal regulations developed in response to the Benedict, S., Georges, J. M. (2006). Nurses and the
Commission’s report provide guidelines for sterilization experiments of Auschwitz: A postmod-
informed consent and REB protocols. ernist perspective. Nursing In uiry, , 277-288.
• Protection of human rights includes the rights to Brink, P. J. (1992). Autonomy versus do no harm.
(1) self-determination, (2) privacy and dignity, estern ournal of Nursing Research, , 264-266.
(3) anonymity and confidentiality, (4) fair treatment, Canadian Council on Animal Care. (2005). About the
and (5) protection from discomfort and harm. Canadian Council on Animal Care. Retrieved from
• Procedures for protecting basic human rights
http://www.ccac.ca/en/CCAC Main.htm.
include obtaining informed consent, which
illustrates the ethical principle of respect, and Canadian Council on Animal Care. (2006). Terms of
obtaining REB approval, which illustrates the ethical reference for animal care committees. Retrieved from
principles of respect, beneficence, and justice. http://www.ccac.ca/Documents/Standards/Policies/
• Special consideration of ethics should addressed in Terms of reference for ACC.pdf.
studies involving vulnerable populations, such as Canadian Institutes of Health Research. (2007). CI R
children, older adults, prisoners, and those who guidelines for health research involving Aboriginal
are mentally or physically disabled. people. Retrieved from http://www.cihr-irsc.gc.ca/e/
• Scientific fraud or misconduct represents unethical 29134.html.
conduct, and professional responsibility must Canadian Institutes of Health Research, Natural
include monitoring for such conduct. Informal, ad
Sciences and Engineering Research Council of
hoc, or unauthorized research may expose patients
to unwarranted risk and may not protect Canada, Social Sciences and Humanities Research
participants’ rights adequately. Council of Canada. (2010, December). Tri-Council
• Nurses who are asked to be involved in product policy statement: Ethical conduct for research
testing should be aware of Health Canada involving humans. Retrieved from http://
guidelines and regulations for testing medical www.pre.ethics.gc.ca/pdf/eng/tcps2/TCPS 2 FINAL
devices before becoming involved in product Web.pdf.
testing and, perhaps, violating guidelines for ethical Canadian Nurses Association. (1983). Ethical research
research. guidelines for nurses in research involving human
• Animal rights need to be protected, and participants. Ottawa: Author.
regulations for animal research have evolved over
Canadian Nurses Association. (2002). Ethical research
time. Nurses who encounter the use of animals in
research should be alert to their rights. guidelines for registered nurses. Ottawa: Author.
• As consumers of research, nurses must be Canadian Nurses Association. (2008). Code of ethics for
knowledgeable about the legal and ethical registered nurses: centennial edition. Ottawa:
components of a research study so that they can Author. Retrieved from http://www.cna-nurses.ca/
evaluate whether a researcher has ensured CNA/documents/pdf/publications/Code of
appropriate protection of human or animal rights. Ethics 2008 e.pdf.
140 PART ONE Research Overview

Carroll, M. L., Ciaffa, J. (2008). The human genome Health Canada. (1997, revised 2004). IC uidance
project: A scienti c and ethical overvie . Retrieved E : ood clinical practice: Consolidated guideline.
from http://www.actionbioscience.org/genomic/ Ottawa: Health Products and Food Branch. Retrieved
carroll ciaffa.html. from http://www.hc-sc.gc.ca/dhp-mps/prodpharma/
Chalmers, K., Sequire, M., Brown, J. (2002). applic-demande/guide-ld/ich/efficac/e6-eng.php
Tobacco use and baccalaureate nursing students: A Heaman, M., Chalmers, K., Woodgate, R., Brown, J.
study of their attitudes, beliefs and personal behav- (2007). Relationship work in an early childhood
iours. ournal of Advanced Nursing, (1), 17-24. home visiting program. ournal of Pediatric Nursing,
Charron, M. (2000). E en Cameron and the Allan , 319-330.
Memorial Psychiatric Institute: A study in research Hershey, N., Miller, R. D. (1976). uman e perimen-
and treatment ethics. Retrieved from http:// tation and the la . Germantown, MD: Aspen.
www.sfu.ca/ wwwpsyb/issues/2000/summer/charron. Hilts, P. J. (1995, March 9). Agency faults a CLA
htm study for suffering of mental patients. Ne or
Chen, W. (2010). nderstanding the cultural context of Times, pp. A1, A11.
Chinese mothers’ perceptions of breastfeeding and Jack, S. M., DiCenso, A., Lohfeld, L. (2005). A
infant health in Canada. ournal of Clinical Nursing, theory of maternal engagement with public health
, 1021-1029. nurses and family visitors. ournal of Advanced
Code of Federal Regulations. (1983). Protection of Nursing, , 182-190.
human subjects, 45 CFR 46, OPRR Reports, revised. Julion, W., Gross, D., Barclay-McLaughlin, G.
Collins, A. (1988). In the sleep room: The story of the (2000). Recruiting families of color from the inner
CIA brain ashing e periments in Canada. Toronto: city: Insights from the recruiters. Nursing utloo ,
Lester Orpen Dennys. , 230-237.
Creighton, H. (1977). Legal concerns of nursing Kevles, D. J. (1996, July 5). An injustice to a scientist is
research. Nursing Research, , 337-340. reversed, and we learn some lessons. Chronicle of
Davis, A. (1981). Ethical issues in gerontological igher Education, pp. B1-BB2.
nursing research. eriatric Nursing, , 267-272. Levine, R. J. (1979). Clarifying the concepts of research
Dubler, N. N., Post, L. F. (1998). Truth telling and ethics. astings Center Report, (3), 21-26.
informed consent. In J. C. Holland (Ed.), Te tboo of Levine, R. J. (1986). Ethics and regulation of clinical
psycho-oncology (pp. 1085-1095). New York: Oxford research (2nd ed.). Baltimore: rban and
Press. Schwartzenberg.
Elder, G. (1981). Social history and life experience. In McLean, P. (1996, November). Biomedical research and
D. H. Eichorn, J. A. Clausen, N. Haan, P. H. the law of informed consent. The Canadian Nurse,
Mussen (Ed.), Present and past in middle life , 49-50.
(pp. 3-31). New York: Academic Press. Milgram, S. (1963). Behavioral study of obedience.
Flaskerud, J. H., Winslow, B. J. (1998). Conceptual- ournal of Abnormal and Social Psychology, ,
izing vulnerable populations’ health-related research. 371-378.
Nursing Research, , 69-78. Mitchell, K. (1984). Protecting children’s rights during
French, H. W. (1997, October 9). AIDS research in Africa: research. Pediatric Nursing, , 9-10.
Juggling risks and hopes. Ne or Times, pp. A1, A12. Morse, J. M., Penrod, J., Kassab, C., Dellasega, C.
Gillmor, D. (1987). I s ear by Apollo: Dr E en (2000). Evaluating the efficiency and effectiveness of
Cameron and the CIA-brain ashing e periments. approaches to nasogastric tube insertion during
Montreal: Eden Press. trauma care. American ournal of Critical Care, ,
Haggerty, L. A., Hawkins, J. (2000). Informed 325-333.
consent and the limits of confidentiality. estern National Commission for the Protection of Human
ournal of Nursing Research, , 508-514 Subjects of Biomedical and Behavioral Research.
alush a v niversity of Sas atche an et al., 53 (1978). Belmont report: Ethical principles and
D.L.R. (2nd) 436, 52 W.W.R. 608 (Sask. CA. 1965). guidelines for research involving human subjects
Health Canada. (n.d.). ealth Products and ood (DHEW Pub. No. OS 78-0012). Washington, DC:
Branch Inspectorate Recall Policy (POL-0016). .S. Government Printing Office.
Retrieved from http://www.hc-sc.gc.ca/dhp-mps/ Office of Research Integrity, .S. Department of Health
compli-conform/info-prod/drugs-drogues/pol 0016 and Human Services. (2011). Retrieved from http://
recall policy-politique retrait ltr-doc-eng.php a2) www.ori.dhhs.gov/misconduct/cases/.
 CHAPTER 6 Legal and Ethical Issues 141

Ownership, control, access and possession and unit: perceptions of space, time and interprofessional
research ethics. (2003). The Aboriginal Nurse, , practice. ournal of Advanced Nursing, ,
10-11. 1872-1881.
Piaget, J. (1937/1954). a construction du r el che Smith, D., Edwards, N., Varcoe, C., Martens, P. J.,
l enfant The construction of reality in the child. Davies, B. (2006). Bringing safety and responsive-
New York: Basic Books. ness into the forefront of care for pregnant and
Pranulis, M. F. (1996). Protecting rights of human parenting Aboriginal people. Advances in Nursing
subjects. estern ournal of Nursing Research, , Science, (2), E27-E44.
474-478. Smylie, J., Tonelli, M., Hemmelgarn, B., Yeates, W. M.,
Pruchino, R. A., Hayden, J. M. (2000). Interview Joffres, M. R., Tataryn, I. V., et al. (2005). The ethics
modality: Effects on costs and data quality in a of research involving Canada’s Aboriginal popula-
sample of older women. ournal of Aging and tions. Canadian Medical Association ournal, ,
ealth, (1), 3-24. 977-979.
Rankin, M., Esteves, M. D. (1997). Perceptions of Tilden, V. P. (2000). Preventing scientific misconduct—
scientific misconduct in nursing. Nursing Research, Times have changed. Nursing Research, , 243.
, 270-276. nited States Government Printing Office. (2008). The
Raybuck, J. A. (1997). The clinical nurse specialist as medical case. In Trials of ar criminals before the
research coordinator in clinical drug trials. Clinical Nuremberg Military tribunals under Control Council
Nurse Specialist, (1), 15-19. a No (Vol. 2, pp. 181-182). Washington, DC:
Rempusheski, V. F. (1991). Elements, perceptions, and Author, 1949. Retrieved from http://www.loc.gov/rr/
issues of informed consent. Applied Nursing frd/Military Law/pdf/NT war-criminals Vol-II.pdf
Research, , 201-204. Wheeler, D. L. (1997, December 12). Three medical
Rennick, J. E., Morin, I., Kim, D., Johnston, C., organizations embroiled in controversy over use of
Dougherty, G., Platt, R. (2004). Identifying placebos in AIDS studies abroad. Chronicle of igher
children at risk for psychological sequelae after Education, A15-A16.
pediatric intensive care unit hospitalisation. Pediatric Wheeler, D. L. (1999, August 13). For biologists, the
Critical Care Medicine, , 358-363. postgenomic world promises vast and thrilling new
Robb, I. H. (1900). Nursing ethics: or hospital and knowledge. Chronicle of igher Education,
private use. Milwaukee: GN Gaspar. A17-A18.
Rothman, D. J. (1982). Were Tuskegee and Willow-
brook studies in nature astings Centre Report,
(2), 5-7.
Schell, K., Brienning, E., Lebet, R., Pruden, K.,
Pawheiser, S., Jackson, B. (2010). Comparison of
FOR FURTHER STUDY
arm and calf automatic non-invasive blood pressures Go to Evolve at http://evolve.elsevier.com/
in pediatric intensive care patients. ournal of Canada/LoBiondo/Research for Audio Glossary, how-to
Pediatric Nursing, , 3-12. instructions for Writing Proposals for Funding, and
Seneviratne, C. C., Mather, C. M., Then, K. L. additional research articles for practice in reviewing
(2009). nderstanding nursing on an acute stroke and critiquing.
This page intentionally left blank
143

PA RT T WO
PART ONE Part Title

Qualitative Research

7 Introduction to Qualitative
Research
8 Qualitative Approaches to
Research
RESEARCH VIGNETTE desire to add to the knowledge
that will help practising nurses
Creating Qualitatively Derived and real patients in some mean-
ingful way.
Knowledge for a Practice Discipline I happened to enter graduate
studies just at the time when
the worldview of qualitative
Sally Thorne RN, PhD, FCAHS study of human complexity has approaches was emerging as an
Professor been active for generations. Since alternative to the more conven-
School of Nursing
nurse scholars began to adopt tional scientific methods with
University of British Columbia
Vancouver, British Columbia these methods in the 1980s, they which nurse scholars had been
have appreciated that the needs of struggling. The prevailing assump-
this applied health discipline are tion was that qualitative research-
To build knowledge for a practice quite distinct from those of the ers were entirely different thinkers
discipline with the complexity and more theoretical social sciences than their quantitative colleagues,
dynamism of nursing, ideas must thus, qualitative nursing research and the two kinds of research
be drawn from a wide variety approaches have been evolving products were completely incom-
of perspectives, disciplines, and from being primarily theoretical mensurate. Thus, my entry into a
inquiry approaches. In the practice and toward being more practically nursing research career coincided
arena, nurses universally recog- applicable. The kind of knowl- with a time of tension and transi-
nize the patient’s perspective as edge that is desirable is not tion in the methodological uni-
among the fundamentally impor- concerned simply with building verse we inhabited. As time
tant aspects that must be consid- theories it is about translating passed, I was fortunate enough to
ered when decisions are made. what is known into what nurse be part of the emerging explora-
Beyond what can be gleaned researchers can potentially use. tions in methodology for example,
from individual patient-centred For example, nurses have the systematic and rigorous inte-
approaches, general knowledge become increasingly dissatisfied gration of the unique insights
about how patient perspectives with research that merely descri- derived from both measurement
are constituted and expressed is bes something instead, they are and interpretation into robust
needed to guide that individual- seeking knowledge that both knowledge platforms. Instead
ized assessment, to help nurses describes and interprets: telling of arguing the merits of the various
understand what they are looking not only what seems to be happen- ways of studying phenomena,
for and then how to interpret and ing but also why that is important. nurse researchers have increas-
make sense of what they find. The Because they understand that their ingly joined together in solving
primacy of this patient perspective knowledge products are most the complex problems faced by
knowledge as a foundational core valuable when they contribute to health care planners, policymak-
of disciplinary practice creates the the evidentiary basis on which ers, and clinicians. By applying
intellectual climate within which health practice and policy are the best parts of all of our different
nurse researchers have led the way constructed, nurses increasingly perspectives in a thoughtful and
among the applied health disci- orient their questions and the rational manner, we are trying to
plines to develop groundbreaking methods by which they seek to build and implement systems of
qualitative, methodological inno- answer them to the most pressing care that provide the best possible
vations for understanding health and hard-to-solve clinical chal- conditions for our patients. It is an
phenomena. lenges. Nurses who engage in a exciting time to be involved in
Many techniques used in quali- qualitative study are doing so not nursing scholarship
tative research actually originated merely out of curiosity or theoreti- The kinds of studies that most
in the social sciences, in which the cal inclination, but rather with a attract my interest these days are

144
those addressing aspects of health from skilled and competent com- varied contextual understanding
care in which usual nursing prac- munication and experience harm to the trends that can be detected
tice is not yet as effective as it from miscommunications. Patients quantitatively: why certain kinds
ought to be and meaningful ought not to launch their cancer of information exchange are
improvements can be envisioned. journeys with levels of confusion, more satisfying to some kinds of
For me, professional-patient com- fear, anxiety, or emotional distress patients and not others, why par-
munication is one such complex that are caused directly by how ticular forms of communication
challenge. Although all nurses nurses interact with them, and trigger frustration and despair,
experientially know how powerful there is much to be gained from and so on.
human communication can be in studying as wide a range of human Interpretive description, which
the thoughts, feelings, and behav- experience as possible to keep is the applied qualitative research
ioral responses to critical events in refining the sense of how to com- approach that nurse scholars use,
life, it is extraordinarily difficult municate well. is derived from a clinician’s per-
to articulate and enforce evidence- In a topic such as cancer com- spective on what kinds of knowl-
informed best practices in this munication, although I orient my edge are likely to be most useful.
regard. research questions from the per- Nurse researchers read and con-
In cancer care, communication spective of how a nurse sees sider social theoretical perspec-
ought to be a priority, in view of the problem, I also recognize the tives, but they do not believe that
what nurses know of its power inherent interdisciplinarity of the one theoretical perspective ought
to nurture or de ate, to inform challenge. I therefore thrive on to ground or frame their studies.
or misinform, to discourage or working with strong interpro- Rather, they believe that a clinical
encourage hope. Because commu- fessional and interdisciplinary logic is the most appropriate intel-
nication is so nuanced, complex, research teams. In the cancer com- lectual scaffolding on which to
and various, it is not particularly munication work, about half of base design decisions, including
amenable to conventional quanti- my team is composed of nurses, all aspects of research orientation
tative inquiry methods. Although and the other half represents epi- and data collection, analysis, and
a few things may be quantifiable demiology, physical therapy, radi- interpretation. Nurse scholars try
in relation to communication, ation oncology, and social work. to design studies that will not
much of what we learn from quan- In our program of research, we only illuminate common patterns
tification is rather irrelevant to the have chosen to focus attention on but also guide them in detecting
overall subjective experience of the patient perspective, acknowl- and making sense of predictable
being in a communicative encoun- edging that what we can glean is and even infrequent variations,
ter, and the quantitative evidence not in and of itself a truth but since nursing never deals with
provides very little guidance for rather a detectable pattern of sub- standardization in the absence of
improving patient experience. jective material that—together individualization.
Therefore, communication is an with material we can obtain from What nurses aspire to is the
ideal phenomenon to be studying other angles of vision—creates an kind of research report that will
from multiple angles, including— evolving body of understanding. elucidate the relationship between
of most importance—the perspec- Although we recognize the possi- practice elements that may not
tive of those involved in the bility that patient reports could be have been well aligned previously,
communication encounter. One skewed in particular directions, so that thoughtful clinicians can
might argue that interpersonal we have found over many years in see directionality in practice
communication about difficult this field that they are remarkably more clearly and be more confi-
issues such as a cancer diagnosis authentic to the perceptions that dent in their practice improve-
is so complex that it ought not to clinicians hold about what goes on ments. Nurses also aim to help
be possible however, nurses in the practice context. Further- teachers of communication com-
know that patients truly benefit more, they provide a rich and petencies look beyond generalities

145
and into more finely tuned expert informed, guided, supported, and As you read this book and
practices, understand how to connected throughout their cancer familiarize yourself with the won-
nurture and support those prac- services. I am also convinced derful world of nursing research, I
tices, and to challenge poor that qualitative research provides hope that your imagination will be
practices when they exist. nurses with some insight that inspired with directions that you
I genuinely believe that nurses would otherwise be missing in might take your own inquiries on
have a pivotal role in shaping the the evidentiary base that allows behalf of our profession. Great
communicative environment in them to advocate on behalf of research really can be a powerful
which patients with cancer are patients. tool for practice ■

146
C H A PTER 7

Introduction to Qualitative
Research
Julie Barroso | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Differentiate between qualitative and quantitative research paradigms.
• Describe the beliefs generally held by qualitative researchers.
• Describe the components of a qualitative research report.
• Identify the links between qualitative research and evidence-informed practice.
• Identify four ways in which qualitative findings can be used in evidence-informed practice.
• Discuss significant issues that arise in conducting qualitative research.

KEY TERMS
bracketing “grand tour” question purposive sample
context dependent inclusion criteria qualitative research
data saturation inductive reflexivity
deductive metasynthesis text
exclusion criteria naturalistic setting triangulation
focus group

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

147
148 PART TWO Qualitative Research

LET’S SAY THAT YOU ARE READING an article that ualitative studies most often help researchers
reports findings that men infected with the human begin to formulate an understanding of a phe-
immunodeficiency virus (HIV) are more adherent nomenon. Although qualitative research has a
to their antiretroviral regimens than are HIV- long history in the social sciences, it is only since
infected women. You wonder why that is so: Why 1990 that it has become more accepted in nursing
would women be less adherent in taking their research. For many years, doctoral nursing stu-
medications Certainly, it is not solely because dents were dissuaded from conducting qualitative
they are women. Or you are working on a post- studies the push was for the traditional quantita-
partum unit and have just discharged a new tive approach, which was viewed by many
mother who has debilitating rheumatoid arthritis. authorities in the hard sciences as being more
You wonder what the process is by which dis- credible. Thus, as nursing gained its foothold in
abled women decide to have children: How do academics, doctoral students were urged to
they go about making that decision These, like conduct research by using the quantitative para-
so many other questions that nurses have, can be digm, or worldview (beliefs and practices, shared
best answered through research conducted with by communities of researchers), to help nursing
qualitative methods. ualitative research yields gain legitimacy in academe. However, as academe
the answers to those difficult questions. Although and research evolved along two different but par-
qualitative methods can be used at many different allel channels, qualitative research received
points in a program of research, you can most greater acceptance today’s generation of nurse
often use them to answer questions that nurses scholars are trained in qualitative methods and
have when a particular phenomenon in nursing is encourage students to use the method that best
not well understood. answers their research questions, as opposed to
In this chapter, the basic tenets of qualitative using methods that might add a veneer of scien-
research are reviewed the components of a quali- tific legitimacy to its conduct but do not answer
tative report, qualitative research, and evidence- the research question at hand.
informed practice are explored and the issues in ualitative research is discovery oriented it is
qualitative research are examined. explanatory, descriptive in nature. Words, as
opposed to numbers, are used to explain a phe-
WHAT IS QUALITATIVE RESEARCH? nomenon. The data gathered in qualitative
ualitative research is a systematic, interactive, research come from the text. The term te t used
and subjective research method used to describe in this context means that data are in textual form:
and give meaning to human experiences. This that is, narrative or words written from interviews
broad term encompasses several methodologies that were recorded and then transcribed or notes
that share many similarities in the conduct of such written from the researcher’s observations. uali-
research. According to Denzin and Lincoln tative research lets us see the world through the
(1994), qualitative researchers study things in eyes of another: the woman who struggles to take
their natural settings, attempting to make sense her antiretroviral medication or the woman with
of, or interpret, phenomena in terms of the mean- a debilitating illness who has nonetheless care-
ings people bring to them (p. 2). A naturalistic fully thought through what it might be like to
setting is one that people live in every day. There- have a baby. ualitative researchers assume that
fore, the researcher conducting qualitative nurses can understand these experiences only if
research goes wherever the participants are: in they consider the context in which the experi-
their homes, schools, communities, and, some- ences take place, and this is why most qualitative
times, in the hospital or an outpatient setting. research takes place in naturalistic settings.
 CHAPTER 7 Introduction to Qualitative Research 149

ualitative studies make the world of an indi- context dependence what an individual sees
vidual visible to the other people. ualitative depends on who that individual is and what expe-
research encompasses modes of inquiry oriented riences the individual brings to the situation.
toward how the social world is interpreted, under- ualitative researchers believe that the discovery
stood, experienced, produced, or constituted of meaning is the basis for knowledge. ualita-
(Mason, 2002, p. 3). Refer to the Critical Think- tive researchers know that they must describe the
ing Decision Path in Chapter 2 (p. 32) to see an phenomenon under study well ideally, the reader,
illustration of the different views and approaches if evenly slightly acquainted with the phenome-
within the qualitative and quantitative research non, would have an Aha moment in reading a
processes. This decision algorithm shows that dif- well-written qualitative report.
ferent beliefs lead to different questions, which You may now be saying, Wow This sounds
lead to selecting different research approaches. great ualitative research is for me Many
nurses do feel very comfortable with this approach
WHAT DO QUALITATIVE because they are educated in how to talk to people
RESEARCHERS BELIEVE? about the health issues concerning them they are
ualitative researchers believe that there are mul- used to listening and listening well. The most
tiple realities for example, they believe that the important consideration for any research study,
experience of having a baby, although some however, is whether the methodology fits the
aspects are common to all deliveries, is not the question. It must fit, or else the study will con-
same for any two women and is definitely differ- tribute little to they scientific knowledge base for
ent for a disabled mother. ualitative researchers practice. This is also the first question you should
believe that reality is socially constructed and ask yourself when you read studies and are con-
conte t de endent—that is, the meaning of an sidering them as evidence on which to base your
observation is defined by its circumstance or the practice: Does the methodology fit with the
environment. For example, even the experience research question under study
of reading this book is different for any two stu-
dents one may be completely engrossed by the Helpful Hint
content, while another is reading but is worrying All research is based on a paradigm, but the para-
about whether her financial aid application will digm is seldom specifically identified in a research
report.
be approved soon. Figure 7-1 is an illustration of

FIGURE 7-1 Shifting perspectives: seeing the world as others see it.
From GARFIELD © 1983 Paws, Inc. Reprinted with permission of UNIVERSAL UCLICK. All rights
reserved.
150 PART TWO Qualitative Research

DOES THE METHODOLOGY FIT WITH THE COMPONENTS OF A QUALITATIVE

RESEARCH QUESTION BEING ASKED? RESEARCH REPORT
As stated before, qualitative methods are often Review of the Literature
best for helping researchers determine the nature The first step has already been discussed: being
of a phenomenon. Sometimes, authors state that clear that a qualitative approach is the best way
they are using qualitative methods because little to answer the research question. Next, the author
is known about a phenomenon that alone is not presents a quick review of the relevant literature.
a good reason for conducting a study, however. This may require creativity on the author’s part
Little may be known about a phenomenon because because published research on the phenomenon
it does not matter Before researchers ask people in question may not exist. However, there are
to participate in a study, to reveal themselves and likely to be studies on similar participants, with
their lives to them, the researchers should be the same patient population, or on a closely related
asking about things that will help make a differ- concept. For example, the author may want to
ence in people’s lives, in how professionals research how women who have a disabling illness
provide nursing care. You should be able to artic- make decisions about becoming pregnant.
ulate a valid reason for conducting a study, Although no other studies in this particular area
beyond little is known about . In the first may have been conducted, there may be some on
example at the start of this chapter, researchers decision making in pregnancy when a woman
need to know why HIV-infected women are less does not have a disabling illness. These would be
adherent to their medication regimens so that the important inclusions in the review of the literature
researchers can work to change these barriers and to show readers that the author is familiar with
can anticipate these barriers when a woman is the research on this process in a nondisabled
ready to start taking these pills. In the second woman. Or there may be literature on decision
example, when researchers learn about the making in pregnancy when a woman has a differ-
decision-making processes that disabled women ent but not disabling illness, such as cancer or
use to decide whether to have a child, researchers HIV infection. Reading all of this related litera-
can better answer the questions of the next woman ture will help you discern whether the author
who is going through this process. really seems to know the field and thus the kinds
The parts of a qualitative research study are of questions they asked the participants.
discussed next. Box 7-1 outlines the steps of a Assume that the author wanted to examine
qualitative study. HIV-infected women’s adherence to regimens of
antiretroviral therapy. If there is no research
directly on this topic, the author might examine
research on adherence to therapy for other chronic
BOX 7-1 illnesses, such as diabetes or hypertension. The
STEPS IN THE QUALITATIVE RESEARCH PROCESS author might want to include studies of gender
Review of the literature differences in medication adherence or the author
Study design might want to examine the literature on adher-
Sample ence in a stigmatizing illness or to examine
Setting: recruitment and data collection
Data collection appointment adherence for women, to see what
Data analysis facilitates or acts as a barrier to attending health
Findings care appointments. The major point is that even
Conclusions
though no literature on the author’s exact subject
 CHAPTER 7 Introduction to Qualitative Research 151

may exist, the author should review the literature. they needed Lack of a pilot study is not a deficit,
In fact, it usually is more challenging to write the however.
review of the literature for a qualitative study
because the authors must be creative and think Sample
of all of the other comparisons they need to The next part of the report is the description of
make, whether it is on the study subject, relevant the sample and setting. This section contains criti-
study concepts, or similar/dissimilar patient cal information that enables you to understand
groups. how qualitative research differs from quantitative
At the conclusion of the review, the most research. In qualitative studies, the researchers
important points that you have learned should be are usually looking for a ur osive sa le a
clear, and you should be able to articulate the group consisting of particular people who can
problem to be studied and the purpose for study- elucidate the phenomenon they want to study.
ing it. As discussed in previous chapters, some Therefore, their recruitment materials must be
qualitative researchers conduct a very limited very specific so that when people read their
review because they want to be amenable to dis- recruitment yers, they know whether they satisfy
covering and learning about the phenomenon the criteria. Thus, if the researchers want to talk
under study and not be swayed or otherwise in u- to HIV-infected women about adherence, they
enced by previous findings in the field. may distribute yers recruiting for women who
are adherent and those who are not. Or they may
Study Design want to talk to women who qualify for only one
In the next part of the report, the authors should of those categories. The researchers who are
explain the study design. In qualitative research, examining decision making in pregnancy among
there may simply be a descriptive or naturalistic women with disabling conditions would clearly
design, in which the researchers adhere to the list the conditions they want to study. For example,
general tenets of qualitative research but do not they may describe wanting to talk to women with
commit to a particular method. However, there multiple sclerosis or those with rheumatoid
are different types of qualitative methods, which arthritis.
are discussed in Chapter 8. What is important, as Researchers may impose other parameters as
you read from this point forward, is that the study well, such as requiring that participants be older
design is congruent with the philosophical beliefs than 18 years, or not using illicit drugs, or decid-
that qualitative researchers hold. In other words, ing about a first pregnancy (as opposed to subse-
you would not expect to read about a random quent pregnancies). These parameters are known
sample, a battery of questionnaires administered as inclusion criteria (criteria that people must
in a hospital outpatient clinic, or a complicated satisfy to participate in a study) and e clusion
statistical analysis. sually, the researchers also criteria (criteria used to exclude people from par-
indicate that they have received ethical approval ticipating in a study). It is critical that the authors
from the appropriate research ethics board. make these criteria transparent to the reader, so
You may read about a pilot study in the opening the reader can judge the abilities of the partici-
of the design section this is work that the pants to shed light on the phenomenon in
researchers performed before undertaking the question.
main study to make sure that the logistics of Often the researchers make decisions such as
the proposed study were reasonable: Were they how to define a long-term survivor of a certain
able to recruit participants Did the questions illness. In this case, they need to tell you, the
they asked of participants yield the information reader, why and how they decided who would
152 PART TWO Qualitative Research

qualify this category. Is a long-term survivor researcher might see an oxygen tank, a glass of
someone who has had an illness for 5 years For water, medications, telephone, television, a box
10 years What is the median survival time for of tissues, and so on. This may be an indicator
people with this diagnosis The researchers’ deci- that the participant is someone for whom getting
sions should be based on sound scientific ratio- around is tremendously difficult. In any event,
nale. Then the researchers need to describe how a good qualitative researcher considers this
they found these participants. In the example of setting as additional data to help complete the
finding HIV-infected women who are having dif- complex, rich scenario that is being rendered
ficulties adhering to a medical regimen, they may in the study.
report distributing yers describing the study at
acquired immune deficiency syndrome (AIDS) Data Collection
service organizations, support groups for HIV- Data collection is another part of the process in
infected women, clinics, and other places where which the two research paradigms differ tremen-
people with HIV may seek services. Again, this dously. In a qualitative study, the data to be col-
is one of the most critical parts of the qualitative lected are usually words: The researcher may
research process, and you should read it with interview an individual, interview a group of
great care. people in what is called a ocus grou or observe
In qualitative research, there is no set sample an individual as she or he goes about a task such
size as there is in a quantitative study (see Chapter as sorting medications into a pill minder. In each
12). ualitative researchers gather participants of these cases, however, the data collected are
until data saturation occurs. ata saturation is expressed in words. The researcher asks the par-
the point in a qualitative study when the informa- ticipant about the phenomenon of interest and
tion being shared with the researcher from par- then listens. However, the researcher does not
ticipants become repetitive in other words, the have to do this without some technical assistance.
ideas shared by the participants have been shared Most qualitative researchers use audio recorders
by previous participants and no new ideas emerge. to ensure that they have captured the participant’s
exact words. This also takes some of the pressure
Setting: Recruitment and Data Collection off of researchers to write down every single
The setting section may actually describe two word, and it frees them up to listen fully. The
settings: the setting in which participants were recordings are usually transcribed verbatim, and
recruited and the setting in which data were col- then the researcher who conducted the interviews
lected. As already discussed, the settings in which listens to the recordings for accuracy.
data were collected are another area of critical The data collection section should also describe
difference between quantitative and qualitative details such as whether informed consent was
studies. In a qualitative study, data is usually col- obtained and the steps from when a participant
lected in a naturalistic setting the participants are contacted the researcher to the end of the study
not usually brought into a clinic interview room. visit. It is important to also know how long each
The setting for data collection is often the partici- interview or focus group lasted and how much
pant’s home, which can be an incredible window time overall the researcher spent in the field
into other aspects of the participant’s life. To be collecting data.
in someone else’s home is a great privilege and Another important component in this section
helps the researcher understand what that partici- is the description of when the researcher decided
pant values. For example, those who are ill may that the sample was sufficient. In qualitative
have everything they could need to get through studies, researchers generally continue to recruit
a day clustered around a favorite chair: The participants until they have reached data
 CHAPTER 7 Introduction to Qualitative Research 153

saturation: that is, when no new information is and children can be all-demanding, leaving the
emerging from the interviews. As stated earlier, mother with little to no time to take care of
the number of participants to be selected is usually herself. Therefore, a neutrally worded question
not predetermined as in quantitative studies about the in uence of children would be a
rather, the researchers keep recruiting until they prompt if the participants do not mention it
have the data they need. One important exception spontaneously.
to this is a study in which a researcher is very The sample may be described in the data col-
interested in getting different types of people in lection section or in the Findings section. In
the study for example, in the study of HIV- any event, besides the typical demographic data
infected women and medication adherence, the collected in any study, a qualitative researcher
researchers may want to interview some women should also report on key areas of difference in
who were very adherent in the beginning but then the sample: In a sample of HIV-infected women,
became less so over time, women who were not there should be information about stage of illness,
adherent in the beginning but then became adher- what kind/how many medications they must take,
ent, or women with children and those without how many children they have, and so on. This
children to determine the in uence of being a information helps you, the reader, place the data
mother on adherence. However, sample sizes tend into some context.
to be fairly small (fewer than 30 participants)
because of the enormous amounts of written text Data Analysis
that need to be analyzed by the researcher. Next in the report is the description of data analy-
Finally, you should read in this section about sis, in which the researcher describes how he or
the kinds of questions the researchers asked the she handled the raw data, which are usually tran-
participants. These are different from the research scripts of the recorded interviews in a qualitative
question or questions, which should be broad and study. Many qualitative researchers use computer-
perhaps written in fairly esoteric language. The assisted data analysis programs to help with this
interview questions should be clear, be plain, and task, which can seem overwhelming because of
elicit exactly what the researcher wants to know. the sheer quantity of data to be dealt with.
In qualitative studies, there may be a broad However, other researchers analyze the data
overview or grand tour uestion such as themselves. In either situation, the goal is to find
Tell me about taking your medications: the commonalities and differences in the interviews,
things that make it easier and the things that make and then to group these into broader, more
it harder, or Tell me what you were thinking abstract, overarching categories of meaning that
about when you decided to get pregnant. Along capture much of the data. For example, in the case
with this overview question, there are usually a regarding pregnancy for disabled women, one
series of prompts (additional questions) that were woman might talk about having discussed the
derived from the literature these are areas that need for assistance with her friends and found
the researcher believes are important to cover and that they were willing and able to help her with
that the participant will probably cover, but they the baby. Another woman might talk about how
are available to remind the researcher in case the she discussed the decision with her mother and
material is not mentioned. For example, with sisters and found them to be a ready source of aid.
regard to medication adherence, the researcher A third woman may say that she talked about this
may have read in other studies that motherhood with her church study group, and they told her
can in uence adherence in two very different that they could arrange to bring meals and help
ways: children can become a reason to live, which with housework during the pregnancy and after-
would facilitate taking antiretroviral medication, ward. On a more abstract level, these women are
154 PART TWO Qualitative Research

all talking about social support. Thus, it is pos- the themes and eshing out each theme with a
sible to find a term that is all-encompassing for thorough explanation of the role that it plays in
these descriptions. In an ideal situation, the the question under study. The author should also
authors might even describe an example such as provide quotations that support each of the
the one you just read, but the page limitations of themes. Ideally, the quotation will be staged,
most journals do not permit this level of detail. which gives you some information about the par-
Chapter 15 includes a more in-depth exploration ticipant from whom it came: Was it a woman of
on qualitative data analysis methods. colour with newly diagnosed HIV infection who
did not have children Was it a disabled woman
Evidence-Informed Practice Tip who has chosen to become pregnant but has suf-
Qualitative researchers use more flexible proce-
fered two miscarriages Staging of quotes allows
dures than do quantitative researchers. While collecting you to put the information into some social
data for a project, they consider all of the experiences context.
that may occur. In a really good report of qualitative research,
some of the quotations will give you an Aha
feeling: you will have a sense that the researcher
Findings has done an excellent job of getting to the core of
At last, we come to the results First, the authors the problem. uotations are as critical to qualita-
should discuss whether they are describing a tive reports as numbers are to a quantitative study
process (as in the decision-making example) or a you would not have a great deal of confidence in
list of circumstances that are functioning in some a quantitative report in which the author asks you
way (such as a list of barriers to and facilitators to believe some finding without giving you some
for taking medications), a set of conditions that statistical findings to back it up.
must be present for something to occur (what At the end of the report is the conclusion. The
parents state they need to care for a ventilator- researcher should summarize the results and
dependent child at home), or a description of should compare the findings with those in the
what it is like to go through some health-related existing literature. How are these findings similar
transition (what it is like to become the caretaker to and different from those in the existing litera-
of a parent with dementia). This is by no means ture The author can also describe new findings
an all-inclusive list but rather examples to help or new conceptual conclusions in this section
you know what you should be looking for. because the findings may have revealed areas that
After the description, the author presents the were not anticipated at the beginning of the study.
results, usually by breaking them down into units This is one of the great contributions of qualita-
of meaning that help the data cohere and tell a tive research: opening up new venues of discov-
story. It is very useful if the author describes the ery that were not heretofore anticipated. The
logic for breaking down the units as they are: Are researcher also makes suggestions regarding how
they discussing the themes from most prevalent to use the findings in practice and offers further
to least prevalent Are they describing a process directions for future research.
in temporal terms Are they starting with topics
that were most important to the participant and
then moving to less important topics Helpful Hint
After describing how the story will be told, Values are involved in all research. It is important,
the author should proceed with a thorough however, that they not influence the results of the
research.
description of the phenomenon, defining each of
 CHAPTER 7 Introduction to Qualitative Research 155

shared pathway or meaning, depiction of experi-

EVIDENCE-INFORMED PRACTICE ential variation, and dense explanatory descrip-
Because nursing is a practice discipline, the most tion. From these, Kearney proposed four modes
important purpose of nursing research is to use of clinical application: (1) insight or empathy, (2)
research findings to improve the care of patients. assessment of status or progress, (3) anticipatory
The best way to start to answer questions that guidance, and (4) coaching. She argued that the
have not been addressed or when a new perspec- greater the complexity and discovery within qual-
tive is needed in practice is through the use of itative findings, the stronger is the potential for
qualitative methods. The answers to questions clinical application. The evidence that qualitative
provided by qualitative data re ect important evi- studies provide is used conceptually by the nurse:
dence that may offer the first systematic insights ualitative studies let nurses gain access to the
about a phenomenon and the setting in which it experiences of patients and help nurses expand
occurs. Therefore, broadening evidence models their ability to understand, which should lead to
beyond a narrow hierarchical perspective is more helpful approaches to care (Kearney, 2001
imperative. see Table 7-1).
nfortunately, qualitative research studies do Kearney (2001) argued that findings restricted
not fare well in the typical systematic reviews on by an existing set of ideas or a priori frameworks
which evidence-informed practice recommenda- provide little or no evidence for practice no new
tions are based. Randomized clinical trials and discoveries can be made when the analyst has
other types of intervention studies traditionally used an existing theory to explain the findings.
have been the major focus of evidence-informed Descriptive categories can portray a high level of
practice, as exemplified by the systematic reviews discovery when a phenomenon is vividly por-
conducted by groups such as the Cochrane Col- trayed from a new perspective. For evidence-
laboration. Typically, the selection of studies to informed nursing practice, these findings can
be included in systematic reviews is guided by serve as maps of previously uncharted territory in
evidence hierarchies (see Chapter 3) that focus on human experience. The third category, shared
the effectiveness of interventions according to pathway or meaning, is more complex in that
their strength and consistency. Because evidence these reports are syntheses of a shared process or
models are hierarchical in nature, which perpetu- experience. The integration of concepts or themes
ates the use of intervention studies—for example, results in a logical, complex portrayal of the phe-
randomized controlled studies (RCTs) are the nomenon. The researcher’s ideas at this level
gold standard of research design—the value of reveal how discrete bits of data can be organized
qualitative studies and the evidence offered by in a meaningful whole. For nursing practice,
their results have remained unclear. ualitative these syntheses represent the bigger picture and
studies historically have been ranked lower in a what it means for the human experience (Kearney,
hierarchy of evidence, as a weaker form of 2001).
research design. Research on depiction of experiential variation
Kearney (2001) developed a useful typology is even more complex than that on shared path-
of levels and applications of qualitative research ways or meaning it describes the main essence
evidence (Table 7-1). She described five catego- of an experience but goes on to show how this
ries of qualitative findings that are distinguished experience varies, depending on the individual or
from one another in their levels of complexity context. For nursing practice, these studies reveal
and discovery: those restricted by a priori (exist- a variety of viewpoints and realizations of a
ing theory) frameworks, descriptive categories, human experience and the contextual sources of
156 PART TWO Qualitative Research

TABLE 7-1
KEARNEY’S CATEGORIES OF QUALITATIVE FINDINGS,* FROM LEAST TO MOST COMPLEX
CATEGORY DEFINITION EXAMPLE

Restricted by a priori (existing Discovery is aborted because researcher Use of the theory of “relatedness” to describe
theory) frameworks has obscured the findings with an women’s relationships without
existing theory substantiation in the data and when an
alternative explanation may describe how
women exist in relationship to others; the
data seem to point to another explanation
other than “relatedness”
Descriptive categories Phenomenon is vividly portrayed from a Children’s descriptions of pain, including
new perspective; provides a map into descriptors, attributed causes, and what
previously uncharted territory in the constitutes good care during a painful
human experience of health and episode
illness
Shared pathway or meaning Synthesis of a shared experience or Description of women’s process of recovery
process; integration of concepts that from depression; each category was fully
provides a complex picture of a described, and the conditions for
phenomenon progression were laid out; the origins of a
phase are discernible in the previous phase
Depiction of experiential Description of the main essence of an Description of how pregnant women
variation experience, but also demonstration of recovering from cocaine addiction might or
how the experience varies, depending might not move forward to create a new
on the individual or context life, depending on the amount of structure
they imposed on their behavior and their
desire to give up drugs and change their
lives
Dense explanatory Rich, situated understanding of a Unique cultural conditions and familial
description multifaceted and varied human breakdown and hopelessness led young
phenomenon in a unique situation; people to deliberately expose themselves
portrayal of the full range and depth to HIV infection in order to find meaning
of complex influences; findings are and purpose in life; description of loss of
densely interwoven in a cohesive social structure and demands of
structure adolescents caring for their diseased or
drugged parents who were unable to
function as adults
HIV, human immunodeficiency virus.
*Information from Kearney, M. H. (2001). Levels and applications of qualitative research evidence. Research in Nursing Health, 21, 45-153.

that variety. Conditional models that explain how the findings are densely interwoven in a cohesive
different variables can produce different conse- structure that provides a rich fund of clinically
quences broaden nurses’ thinking about a phe- and theoretically useful information for nursing
nomenon. Finally, dense explanatory description practice, in which the layers of detail interconnect
is the highest level of complexity and discovery, to increase understanding of human choices and
and is a rich, situated understanding of a multi- responses in particular contexts (Kearney, 2001).
faceted and varied human phenomenon in a How can nurses further use qualitative evi-
unique situation. Such studies portray the full dence The simplest mode, according to Kearney
depth and range of complex in uences that propel (2001), is to use the information to better under-
people to make decisions. Physical and social stand the experiences of patients, which in turn
context are fully accounted for. In these studies, helps nurses offer more sensitive support.
 CHAPTER 7 Introduction to Qualitative Research 157

ualitative findings can also help nurses assess TABLE 7-2

the patient’s status or progress, through descrip- KEARNEY’S MODES OF CLINICAL APPLICATION
tions of trajectories of illness or through offering FOR QUALITATIVE RESEARCH*
a different perspective on a health condition. MODE OF CLINICAL
They enable nurses to consider a range of possi- APPLICATION EXAMPLE
ble responses from patients. Nurses can then Insight or empathy: Nurse is better able to
determine the fit of a category to a particular Nurses can better understand the behaviors
understand their of a patient who is a
patient or try to locate the patient’s condition on patients and offer more woman recovering from
an illness trajectory. sensitive support depression
Anticipatory guidance includes sharing quali- Assessment of status or Nurse is able to describe the
tative findings directly with patients. Patients can progress: Descriptions trajectory of recovery from
of trajectories of illness depression and can assess
learn about others with a similar condition and how the patient is
what to anticipate in the future. This enables them progressing through this
to better obtain resources for future events or to trajectory
Anticipatory guidance: Nurse is able to explain the
look for markers of improvement. Anticipatory Sharing of qualitative phases of recovery from
guidance can also be tremendously comforting findings with the depression to the patient
because sharing research results can help patients patient and reassure her that she
is not alone, that others
realize that they are not alone, that other people have made it through a
have been through a similar experience. Finally, similar experience
Kearney (2001) argued that coaching is a way Coaching: Advising Nurse describes the six
of using qualitative findings in this instance, patients of steps they stages of recovery from
can take to reduce depression to the patient
nurses can advise patients of steps they can take distress or improve and, in ongoing contact,
to reduce distress, improve symptoms, or monitor adjustment to an points out the patient's
trajectories of illness. Table 7-2 is an overview illness, according to the progress through the
evidence in the study stages, coaching her to
of clinical applications of qualitative research. recognize signs that her
Sandelowski and Barroso (2003) also suggested condition is improving
a typology of qualitative research. Grace and *Information from Kearney, M. H. (2001). Levels and applications of
Powers (2009), building on the work of Kearney qualitative research evidence. Research in Nursing Health, 21, 45-153.
(2001) and Sandelowski and Barroso (2003), sug-
gested a method of assessing qualitative studies methods and has different goals, it is important
for evidence-informed practice that is based not to explore how and when to use the evidence
on a hierarchical model but on a pyramid. The use provided by findings of qualitative studies in
of a pyramid removes the focus from a hierarchi- practice. Remember too that when knowledge
cal design perspective to acknowledging the about a particular patient care situation is scarce,
importance of the question, not the design. the data provided by qualitative studies may
The definition of evidence-informed practice provide the best available evidence that informs
has three components: clinically relevant evi- a clinical question or decision about a patient
dence, clinical expertise, and patient preferences. population or patient care.
ualitative research does not test interventions
but does require researchers to apply their clinical QUALITATIVE APPROACH AND
expertise to the choice of the research question NURSING SCIENCE
and study design, as well as getting a solid under- ualitative research is particularly well suited to
standing of the patient’s experience. Although studying the human experience of health, a central
qualitative research entails the use of different concern of nursing science. Because qualitative
158 PART TWO Qualitative Research

methods focus on the whole of human experience

and the meaning ascribed by individuals living the Naturalistic Setting
experience, these methods extend understanding The central concern that arises when research is
of health beyond traditionally measured units to conducted in naturalistic settings focuses on the
include the complexity of the human health expe- need to obtain consent. Obtaining informed
rience as it occurs in everyday life. This closeness consent is a basic responsibility of the researcher,
to what is real and everyday hold the promise
of guidance for nursing practice it is also impor-
TABLE 7-3
tant for instrument and theory development.
In Figure 7-2, three examples are cited to empha- CHARACTERISTICS OF QUALITATIVE RESEARCH
THAT GENERATE ETHICAL CONCERNS
size the capacity of qualitative research methods
CHARACTERISTICS ETHICAL CONCERNS
to (1) guide nursing practice, (2) contribute to
instrument development, and (3) develop nursing Naturalistic setting Some researchers using methods
that rely on participant
theory. observation may believe that
consent is not always possible
or necessary.
ISSUES IN QUALITATIVE RESEARCH Emergent nature of Planning for questioning and
the design observation emerges over the
Ethics duration of the study. Thus,
Inherent in all research is the protection of human it is difficult to inform
participants. This requirement exists for both participants precisely of all
potential threats before they
quantitative and qualitative research approaches agree to participate.
and is discussed in Chapter 6. The basic tenets of Researcher-participant Relationships developed between
ethical practice hold true for the qualitative interaction the researcher and the
participant may blur the focus
approach. However, several characteristics of of the interaction.
qualitative methods, outlined in Table 7-3, gener- Researcher as In collecting data, the researcher
ate unique concerns and necessitate expanding instrument may misinterpret the
participant’s reality.
the protection of human participants.

Qualitative Research Methods

Guide nursing practice Contribute to Develop nursing theory

instrument
development

by

Using personal stories Using the “voice” of Enabling systematic

to enlighten and enrich research participants structuring of ideas
understanding of everyday to enable evaluation that emerge from
health experience of existing instruments persons who are
or creation of new ones experts through
life experience
FIGURE 7-2 Qualitative approach and nursing science.
 CHAPTER 7 Introduction to Qualitative Research 159

but it is not always easy in naturalistic settings.

For example, when research methods include Researcher-Participant Interaction
observing groups of people interacting over time, The nature of the researcher-participant interac-
the complexity of obtaining consent is apparent. tion over time introduces the possibility that the
These complexities generate controversy and research experience may become therapeutic: a
debate among qualitative researchers. The balance case of research becoming practice. Basic differ-
between respect for human participants and ences exist between the intention of the nurse
efforts to collect meaningful data must be con- when conducting research and when engaging in
tinuously negotiated. The reader should look for practice (Smith Liehr, 2003). In practice, the
evidence that the researcher has addressed this nurse has caring-healing intentions. In research,
issue of balance by recording attention to the the nurse intends to understand the perspective of
protection of human participants. the participant. Such understanding may be a
therapeutic experience for the participant. Some-
Emergent Nature of the Design times talking to a caring listener about things that
The emergent nature of the research design matter energizes healing, even though this result
emphasizes the necessity for ongoing negotiation was incidental. From an ethical perspective, the
of consent with the participant. During the qualitative researcher is promising only to listen
course of a study, situations change, and what and to encourage the participant’s story. If this
was agreeable at the beginning may become experience is therapeutic for the participant, it
intrusive. Sometimes, as data collection proceeds becomes an unplanned benefit of the research.
and new information emerges, the study shifts Several ethical dilemmas may emerge from the
direction in a way that is not acceptable to the qualitative researcher’s interaction with partici-
participant. pants. Glesne (2011) described several roles that
For example, if the researcher is present in a the qualitative researcher may assume, such as
family’s home during a time in which marital exploiter, intervener or reformer, advocate, and
discord arises, the family may choose to renegoti- friend. All researchers use participants to some
ate the consent. From another perspective, Morse extent to meet their own needs, such as status and
(1998) discussed the increasing involvement of recognition from ensuing publications, with no
participants in the research process, which some- recognition of the participants. Furthermore, the
times resulted in participants’ request to have researcher is in a position of power (researchers
their names published in the findings or be who conduct studies in collaboration with others
included as coauthors. Morse suggested that if the share the power). Often, the exploitation is
participant originally signed a consent form and rationalized by the good that may come of sharing
then chose an active identified role, the partici- the knowledge obtained from the research. Issues
pant should then sign a release for publication of reciprocity are particularly troublesome for
form. ethnographic researchers, who are immersed in
The underlying point of this discussion is that fieldwork for long periods of time (Lipson, 1994).
the emergent qualitative research process man- Other dilemmas are faced when the researcher
dates ongoing negotiating of researcher- attempts to intervene in a situation. For example,
participant relationships, including the consent if the researcher becomes aware of potentially
relationship. The opportunity to renegotiate dangerous drug abuse among a group of young
consent establishes a relationship of trust and adults, should the researcher intervene, with
respect, characteristic of the ethical conduct of the possible consequence of breaching the con-
research. fidentiality and, ultimately, the trust of the
160 PART TWO Qualitative Research

participants Finally, as trust and respect are monitor and become aware of their personal
established, researchers may find themselves in biases and feelings (Glesne, 2011). Through this
the role of confidant, which may, in some cases, process of re e ivity in qualitative research,
lead to friendship. Although some qualitative researchers constantly challenge themselves to
researchers find the role of friend acceptable if it understand how their perspective may be shaping
is based on trust, caring, and collaboration, an the method, interviews, analysis, and interpreta-
inherent danger exists that the data are given in tions. In addition, many researchers may return
the context of friendship and not for the purposes to the participants at critical interpretive points
of research (Glesne, 2011). Investigators may and ask for clarification or validation. Patton
also find it difficult to end the relationship and say (2002) advocated the stance of neutrality in other
goodbye to participants. Fournier and colleagues words, the researcher does not enter the field with
(2006) indicated that more attention needs to be predisposed notions but is open to understanding
given to psychological preparation, focused on the world as it unfolds (p. 51).
exiting the relationship. In participatory action Streubert and Carpenter (2011) recommended
research, the researcher also needs to consider that researchers identify their own thoughts, feel-
whether there are any long-term obligations to ings, and perceptions by compartmentalizing
sustain the project (Fournier et al., 2006). them in the process referred to as brac eting in
which personal biases about the phenomenon of
Helpful Hint interest are identified in order to clarify how per-
Researchers are privileged to enter the lives of sonal experience and beliefs may in uence what
other people and must treat the ensuing relationship is heard and reported. Bracketing is important in
with the utmost respect.
both the descriptive phenomenological and the
ethnographic traditions and is necessary for the
Researcher as Instrument researcher to be open and receptive to the phe-
ualitative research mandates that the researcher nomenon under study. Bracketing is based on the
become immersed in the field. nderstanding assumption that people can separate their per-
how other people think, act, and feel is paramount sonal knowledge about a specific phenomenon
(Patton, 2002). Because researchers are interpret- from their experiences and background. For this
ing what they observe and experience, their own reason, bracketing may not always be possible,
personal history, experiences, knowledge, and but, at a minimum, researchers should be aware
bias may distort the data. The responsibility to as much as possible of their own assumptions and
remain true to the data requires that the research- how those assumptions may affect their observa-
ers acknowledge any personal bias and interpret tions and interpretations and thus in uence the
findings in a way that accurately re ects the par- results of the study.
ticipant’s reality. Researchers need to become
aware of and monitor their own subjectivity to Triangulation
decrease any distortion of the data. This respon- Triangulation has become a buzzword in qualita-
sibility is a serious ethical obligation. To accom- tive research. The view of triangulation has pro-
plish this, researchers should prepare for gressed from merely a strategy for ensuring data
differences in other cultures and groups by accuracy (more than one data source presents the
reading, interacting, and seeking out experiences same findings) to an opportunity to more fully
outside of their own norms (Roper Shapira, address the complex nature of the human experi-
2000). ualitative researchers frequently write in ence. From this perspective, triangulation can be
personal journals during their research activity to defined as the expansion of research strategies in
 CHAPTER 7 Introduction to Qualitative Research 161

a single study or multiple studies to enhance other words, results from mixed-methods research
diversity, enrich understanding, and accomplish are more likely to be associated with post-
specific goals. Richardson (2000) suggested that positivism. Remember that post-positivism, as
the triangle be replaced by the crystal as a more discussed in Chapter 2, is the assumption that a
appropriate metaphor for the multimethod reality exists that can be observed, measured,
approach. and understood.
Five basic types of triangulation were described In spite of the dangers of mixing methodolo-
(Denzin, 1978 Janesick, 1994): gies and methods, serious readers of nursing
1. Data triangulation: the use of a variety of research do not take long to determine that
data sources in a study. For example, the approaches and methods are being combined to
researcher collects data at different times, in contribute to theory building, to guide practice,
different settings, and from different groups and to facilitate instrument development. Several
of people. mixed-methodology research designs have been
2. Investigator triangulation: the collaboration developed, many from the seminal work of nurse
of several different researchers or evalua- researchers such as Morgan (1998) and Morse
tors from divergent backgrounds. (1991). Note that researchers need to determine
3. Theory triangulation: the use of multiple the primary method (qualitative or quantitative).
perspectives to interpret a single set of data. For example, if the purpose of the study is to
4. Methodological triangulation: the use of describe, discover, or explore, then the theoretical
multiple methods to study a single problem. drive is inductive (generalizing from specific
5. Interdisciplinary triangulation: the use of data), with principal methods that are qualitative.
other disciplines to increase understanding Observations lead to the development of general-
of the phenomenon (e.g., nursing and izations and. in some cases. theory to explain
sociology). the phenomenon. However, if the purpose of the
Although support exists for the use of multiple research is to confirm a theory or hypothesis, the
research methods, controversies surround the underpinning of the research is deductive (con-
appropriateness of combining qualitative and cluded from data) and, subsequently, a quantita-
quantitative research approaches and of combin- tive drive will be used. Theory is tested by the
ing multiple qualitative methods in one study development of a hypothesis and the gathering of
(Barbour, 1998 Giddings Grant, 2007). Mixed data to accept or reject it. This recognition is
methods research can take two approaches: the imperative because it drives the design of the
mixing of different research methodologies study from the size of the sample to the analysis
(defined as the theoretical assumptions underly- of the data.
ing the research approach) or the mixing of dif- Morgan (1998) identified several models: for
ferent research methods (defined as the tools for example, (1) small, preliminary, qualitative data
collecting and analyzing data Giddings Grant, providing information useful in the development
2007). Mixing methodologies can be more diffi- of a larger quantitative study (2) limited use of
cult if the assumptions and values underlying the quantitative methods to guide the researcher in
research approaches (i.e., the methodologies decisions pertaining to the larger qualitative
being mixed) are from different paradigms. Gid- project (3) qualitative methods used to interpret
dings and Grant (2007) also argued that because results from a quantitative study and (4) quanti-
most mixed-methods studies favour the forms of tative methods used to confirm results from the
analysis associated with positivism, this form of qualitative study. Morse (2003) identified eight
research is a Trojan horse for positivism. In different types of multimethod designs with
162 PART TWO Qualitative Research

simultaneous or sequential use of qualitative and of multiple methods during the course of her
quantitative methods (Table 7-4). 15-year research program can be likened to exam-
Mixed-methods research provides researchers ining different facets of one crystal: in this case,
with a wider range of tools and options to study the experience of miscarrying. The crystallization
phenomena. The variety of methods provides dif- process has contributed to theory building,
ferent views and different levels of data. nursing practice, and instrument development.
Table 7-5 synthesizes three studies reporting Her practice contribution is highlighted by a case
multimethod analyses. The table notes the con- exemplar (Swanson, 1999), which synthesized
ceptual focus of the work, the study purposes, and her years of work with women living through the
whether the study suggests implications for experience of miscarriage.
theory, practice, and instrument development. Both Cameron (2003) and Covell and Ritchie
Swanson’s (1999) work is a good example of (2009) used mixed methodology the quantitative
a research program. tilizing a variety of meth- portion elicited data about a sample, and provid-
odologies and analysis (see Table 7-5), she ing the qualitative portion consisted of interview
addressed implications for practice, instrument questions. Cameron (2003) combined qualitative
development, and theory building focused on the (interviews) and quantitative (questionnaire)
issue of caring for women who have had a mis- methods to explore the lived experience of trans-
carriage. Her research program included an initial fer students in a collaborative baccalaureate
theory-building phase (studies 1 and 2), an instru- nursing program in Ontario (see Table 7-5). Data
ment development phase (studies 3, 4, and 5), and from the quantitative survey supported the find-
a phase of testing a practice intervention (study ings emerging from the qualitative methods. The
6). Swanson used the phenomenological method qualitative methods provided depth and substance
for studies 1 and 2 and quantitative methods for to the findings from the questionnaire—again,
each of her other studies. She did not use more like differing facets of a crystal. Covell and
than one method in any of her studies, but her use Ritchie (2009 see Table 7-5) studied how nurses

TABLE 7-4
TYPES OF MULTIMETHOD DESIGNS
DESIGN ORDER COMMENTS

INDUCTIVE PARADIGM
Qualitative + qualitative Simultaneous One method is dominant and forms the basis for the study; paradigm is
used when more than one perspective is required
Qualitative → qualitative Sequential One method is dominant and forms the basis for the study; the second
supplements the first
Qualitative + quantitative Simultaneous Inductive drive; paradigm is used when some portion of the phenomenon
can be measured
Qualitative → quantitative Sequential Inductive drive; paradigm can confirm earlier qualitative findings

DEDUCTIVE PARADIGM
Quantitative + quantitative Simultaneous One method is dominant and forms the basis for the study; paradigm
validates the finding of each instrument used
Quantitative → quantitative Sequential One method is dominant and forms the basis for the study; paradigm is
used to elicit further details
Quantitative + qualitative Simultaneous Deductive theoretical drive; paradigm is used when some aspect of the
phenomenon is not measurable
Quantitative → qualitative Sequential Deductive theoretical drive; paradigm is often used when the findings are
unexpected, and the qualitative method is used to find explanations
 TABLE 7-5
RESEARCH USING MULTIMETHOD APPROACHES
INSTRUMENT
CONCEPTUAL MULTIMETHOD THEORY-BUILDING PRACTICE DEVELOPMENT
AUTHOR, YEAR FOCUS APPROACH STUDY PURPOSE IMPLICATIONS IMPLICATIONS IMPLICATIONS

Swanson, 1999 Miscarriage Six studies, Study 1: To define common themes for women who had Yes Yes —
and each recently miscarried
caring involving the Study 2: To describe the human experience of Yes Yes —
use of one miscarriage and the meaning of caring
method
Study 3: To use descriptive data to create a survey — — Yes
instrument that is based on women’s experience of
miscarriage
Study 4: To evaluate the relevance of the survey items to — — Yes
create a miscarriage scale
Study 5: To assess the reliability and validity of the
miscarriage scale
Study 6: To test the effects of caring, measurement, and Yes Yes Yes
time on women’s wellbeing in the first year after
miscarriage
Cameron, Transition One study To explore the lived experience of students transferring Yes Yes Yes
2003 resilience involving the from college to university in a collaborative nursing
use of program
multiple
methods
Covell & Medication One study To determine the types of medication errors nurses Yes Yes —
Ritchie, errors involving the report, why errors are or are not reported, and the
2009 use of strategies believed to improve reporting
multiple
methods
CHAPTER 7 Introduction to Qualitative Research
163
164 PART TWO Qualitative Research

handle medication errors. The data were collected process detailed in the article, clinical implica-
concurrently with semistructured interviews tions and the need for further research in the area
(qualitative) and questionnaires (quantitative). were identified.
The interviews focused on the story of the medi- Essentially, metasynthesis provides a way for
cation error, whereas the questionnaire focused researchers to build up a critical amount of quali-
on barriers to reporting medication errors. tative research evidence that is relevant to clinical
The studies of Swanson (1999), Cameron practice. Sandelowski (2004) cautioned that the
(2003), and Covell and Ritchie (2009) constitute use of qualitative metasynthesis is laudable and
a range of approaches for combining methods in necessary but that researchers who use metasyn-
research studies (see Table 7-5). The mixed thesis methods must clearly understand qualita-
methods field continues to evolve as nurse tive methodologies, as well as the nuances of the
researchers strive to determine which research various qualitative methods. It will be interesting
combinations promise an enhanced understand- for research consumers to follow the progress of
ing of human complexity and a substantial con- researchers who seek to develop criteria for
tribution to nursing science. Consumers of nursing appraising a set of qualitative studies and to use
research are encouraged to follow this ongoing those criteria to guide the incorporation of these
discussion. studies into systematic literature reviews.

Synthesizing Qualitative Evidence: Evidence-Informed Practice Tip

Metasynthesis Qualitative research findings can be used in many
The depth and breadth of qualitative research ways, including improving ways clinicians communicate
with patients and with each other.
has grown over the years. It has become impor-
tant to qualitative researchers to synthesize criti-
cal amounts of qualitative findings. ualitative Foundation of Qualitative Research
etasynthesis is a type of systematic review A final example illustrates the differences in the
applied to qualitative research. nlike quantita- methods discussed in this chapter and provides
tive research, in which statistical approaches are you with the beginning skills of how to critique
used to aggregate or average data by means of qualitative research. The information in this
meta-analysis, metasynthesis involves integrating chapter, coupled with the information presented
qualitative research findings on a topic and is in Chapter 8, provides the underpinnings of criti-
based on comparative analysis and interpretative cal appraisal of qualitative research (see Critiqu-
synthesis of qualitative research findings, whereby ing Criteria box, Chapter 8). Consider the question
the researcher seeks to retain the essence and of nursing students learning how to conduct
unique contribution of each study (Sandelowski research. The empirical analytical approach
Barroso, 2007). (quantitative research) might be used in an exper-
Sandelowski and Barroso (2005) reported iment to see whether one teaching method led to
the results of a qualitative metasynthesis and better learning outcomes than did another. The
metasummary that integrated the findings of students’ knowledge might be tested with a
qualitative studies of expectant parents who pretest, the teaching conducted, and then a post-
received a prenatal diagnosis of a fetal abnormal- test of knowledge given. Scores on these tests
ity. sing the methods of qualitative research would be analyzed statistically to determine
synthesis (meta-analysis and metasummary (see whether the different methods produced a differ-
Sandelowski Barroso, 2007), they analytically ence in the results.
reviewed 17 qualitative studies retrieved from In contrast, a qualitative researcher may be
multiple databases. On the basis of the synthesis interested in the process of learning research. The
 CHAPTER 7 Introduction to Qualitative Research 165

researcher might attend the class to see what limitations. ualitative research methods could
occurs and then interview students to ask them to be added to understand how books are used by
describe how their learning changed over time. different groups and to understand the meaning
They might be asked to describe the experience of books to the inhabitants.
of becoming researchers or becoming more This idea provides an important point for qual-
knowledgeable about research. The goal would itative research. One research method does not
be to describe the stages or process of this learn- rank higher than another. Instead, various methods
ing. Or a qualitative researcher might consider the based on different paradigms are essential for the
class as a culture and could join to observe and development of a well-informed and comprehen-
interview students. uestions would be directed sive approach to evidence-informed nursing
at the students’ values, behaviours, and beliefs in practice.
learning research. The goal would be to under-
stand and describe the group members’ shared
meanings. Either of these examples are ways of
viewing a question with a qualitative perspective.
CRITICAL THINKING CHALLENGES
The specific qualitative methodologies are ■ Discuss how a researcher’s values could influence
described in Chapter 8. the results of a study. Include an example in
Many other research methods exist. Although your answer.
it is important to be aware of the basis of the ■ Can the metaphor “We do not always get closer
qualitative research methods used, it is most to the truth as we slice and homogenize and
important that the method chosen is the one that isolate [it]” be applied to both qualitative and
will provide the best approach to answering the quantitative methods? Justify your answer.
■ What is the value of qualitative research in
question being asked.
evidence-informed practice? Give an example.
A helpful metaphor about the need to use a ■ Using the model in Figure 7-2, discuss how you
variety of research methods was used by Seymour could apply the findings of a qualitative research
Kety, a key figure in the development of biologi- study about coping with a miscarriage.
cal research in psychiatry, who was the scientific
director to the .S. National Institute of Mental
Health (NIMH) for many years. He invited
readers to think about a civilization whose inhab-
itants, although very intelligent, had never seen a
KEY POINTS
book (Kety, 1960). On discovering a library, they • All research is based on philosophical beliefs, a
set up a scientific institute for studying books, worldview, or a paradigm.
which included anatomists, physical chemists, • Qualitative research encompasses different
methodologies.
molecular biologists, behavioural scientists, and • Qualitative researchers believe that reality is
psychoanalysts. Each discipline discovered socially constructed and is context dependent.
important facts, such as the structure of cellulose • Researchers’ values should be kept as separate as
possible from the conduct of research.
and the frequency of collections of letters of
• Qualitative research, like quantitative research,
varying length. However, the meaning of a book follows a process, but the components of the
continued to escape them. As he put it, We do process vary.
not always get closer to the truth as we slice and • Qualitative research contributes to evidence-
informed practice.
homogenize and isolate. He argued that a truer • Ethical issues in qualitative research involve issues
picture of a topic under study would emerge only related to the naturalistic setting, the emergent
from research by a variety of disciplines and tech- nature of the design, researcher-participant
interaction, and the researcher as instrument.
niques, each with its own virtues and particular
166 PART TWO Qualitative Research

REFERENCES health research. ualitative ealth Research, ,

362-377.
Barbour, R. S. (1998). Mixing qualitative methods: Morse, J. M. (1991). Approaches to qualitative-
uality assurance or qualitative quagmire ualita- quantitative research methodological triangulation.
tive ealth Research, , 352-361. Nursing Research, , 120-123.
Cameron, C. (2003). The lived e perience of transfer Morse, J. M. (1998). The contracted relationship:
students in a collaborative baccalaureate nursing Ensuring protection of anonymity and confidentiality.
program. npublished doctoral dissertation, niver- ualitative ealth Research, (3), 301-303.
sity of Toronto. Morse, J. M. (2003). Principles of mixed methods and
Covell, C. Ritchie, J. A. (2009). Nurses’ responses to multimethod research design. In Tashakkori, A.,
medication errors. ournal of Nursing Care uality, Teddlie, C. (Eds.), andboo of mi ed methods in
(4), 287-292. social behavioural research (pp. 189-208).
Denzin, N. K. (1978). The research act: A theoretical Thousand Oaks, CA: Sage.
introduction to sociological methods. New York: Patton, M. (2002). ualitative research evaluation
McGraw-Hill. methods (3rd ed.). Thousand Oaks, CA: Sage.
Denzin, N. K., Lincoln, Y. (Eds.). (1994). andboo Richardson, L. (2000) Writing: A method of inquiry. In
of ualitative research. Thousand Oaks, CA: N. K. Denzin Y. S. Lincoln (Eds). andboo of
Sage. ualitative research (2nd ed., pp. 923-948).
Fournier, B., Mill, J., Kipp, W., Walusimbi, M. Thousand Oaks, CA: Sage.
(2006). Discovering voice: A participatory action Roper, J., Shapira, J. (2000). Ethnography in nursing
research study with nurses in ganda. International research. Thousand Oaks, CA: Sage.
ournal of ualitative Methods, (2). Retrieved from Sandelowski, M. (2004). sing qualitative research.
http://www.ualberta.ca/ iiqm/backissues/6 2/fournier. ualitative ealth Research, (10), 1366-1386.
pdf Sandelowski, M., Barroso, J. (2003) Classifying the
Giddings, L. S., Grant, B. M. (2007). A Trojan findings in qualitative studies. ualitative ealth
horse of positivism A critique of mixed methods Research, , 905-923.
research. Advances in Nursing Science, , Sandelowski, M., Barroso, J. (2005). The travesty of
52-60. choosing after positive prenatal diagnosis. ournal of
Glesne, C. (2011). Becoming ualitative researchers: An bstetrical, ynecologic Neonatal Nursing, (4),
introduction (4th ed.). Toronto: Pearson. 307-318.
Grace, G. T., Powers, B. A. (2009). Claiming our Sandelowski, M., Barroso, J. (2007). andboo for
core: Appraising qualitative evidence for nursing synthesi ing ualitative research. Philadelphia:
questions about human response and meaning. Springer.
Nursing utloo , , 27-34. Smith, M. J., Liehr, P. (2003). The theory of
Janesick, V. J. (1994) The dance of qualitative research attentively embracing story. In Smith, M. J., Liehr,
design. In Denzin, N. K., Lincoln, Y. S. (Eds.), P. (Eds.), Middle range theory for nursing.
andboo of ualitative research (pp. 209-219). (pp 167-187. New York: Springer.
Thousand Oaks, CA: Sage. Streubert, H. J., Carpenter, D. (2011). ualitative
Kearney, M. H. (2001). Levels and applications of research in nursing: Advancing the humanistic
qualitative research evidence. Research in Nursing imperative (5th ed.). New York: Wolters Kluwer.
ealth, , 45-153. Swanson, K. M. (1999). Research-based practice with
Kety, S. (1960). A biologist examines mind and women who have had miscarriages. Image ournal
behavior. Science, , 1861-1870, 1960. of Nursing Scholarship, (4), 339-345.
Lipson, G. L. (1994). Ethical issues in ethnography.
In Morse, J. (Ed.), Critical issues in ualitative
research methods (pp. 333-355). Thousand Oaks,
FOR FURTHER STUDY
CA: Sage. Go to Evolve at http://evolve.elsevier.com/
Mason, J. (2002). ualitative researching (2nd ed.). Canada/LoBiondo/Research for Audio Glossary, how-to
Thousand Oaks, CA: Sage. instructions for Writing Proposals for Funding, and
Morgan, D. (1998). Practical strategies for combining additional research articles for practice in reviewing
qualitative and quantitative methods: Applications to and critiquing.
C H A PTER 8

Qualitative Approaches
to Research
Julie Barroso | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Identify the processes of phenomenological, grounded theory, ethnographic, and case study
methods.
• Recognize the appropriate use of historical methods.
• Recognize the appropriate use of community-based participatory research methods.
• Apply critiquing criteria to evaluate a report of qualitative research.

KEY TERMS
behavioural/materialist emic perspective intrinsic case study
perspective ethnographic method key informants
case study method ethnography lived experience
cognitive perspective etic perspective narrative inquiry
community-based participatory external criticism orientational qualitative inquiry
research grounded theory method participatory action research
constant comparative method hermeneutics phenomenological method
context historical research method phenomenology
culture instrumental case study propositions
data saturation internal criticism snowball sampling
domains intersubjectivity theoretical sampling

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

167
168 PART TWO Qualitative Research

CHAPTER 3 DESCRIBES HOW RESEARCH EVIDENCE is particular phenomenon, patient population,

categorized, according to traditional hierarchies or clinical situation.
of research evidence, from strongest to weakest, This chapter describes a variety of qualitative
with an emphasis on support for the effective- research methods, including phenomenological,
ness of interventions. This perspective does not grounded theory, ethnographic, participatory
take into account the ways that qualitative action research, historical, and case study
research can support practice, as discussed in methods. You are encouraged to use the research-
Chapter 7. There is no doubt about the merit of er’s standpoint as each method is introduced: to
qualitative studies the problem is that no one imagine how it would be to study an issue of
has developed a satisfactory method for includ- interest from the perspective of each of these
ing them in the evidence hierarchies. In addition, methods. No matter which method a researcher
qualitative studies can answer the critical why uses, there is a demand to embrace the wholeness
questions that result from many evidence- of humans, focusing on the human experience in
informed practice summaries a research ques- natural settings.
tion may have been answered, but how the The researcher using these methods believes
answer operates in the caring for people is not that each unique human being attributes meaning
explained. You as a research consumer should to his or her experience, and experience evolves
know that qualitative methods are the best way from his or her social and historical context.
to start to answer clinical and research questions Thus, one person’s experience of pain is distinct
about what little is known or when a new per- from another’s and can be known by the indi-
spective is needed in practice. The very fact that vidual’s subjective description of it. For example,
the number of qualitative research studies has for the adolescent with rheumatoid arthritis, the
increased exponentially in nursing and other researcher interested in studying the child’s lived
social sciences re ects the urgent need of clini- experience of pain spends time in the child’s
cians to better understand the experience of natural settings, such as the home and school.
illness. Thousands of reports of well-conducted Efforts are directed at uncovering the meaning of
qualitative studies exist on topics such as the pain as it extends beyond the number of medica-
following (Sandelowski, 2004 Sandelowski tions taken or a rating on a pain scale. ualitative
Barroso, 2007): methods are grounded in the belief that objective
Personal and cultural constructions of data do not capture the whole of the human expe-
disease, prevention, treatment, and risk. rience rather, the meaning of the adolescent’s
Living with disease and managing the phys- pain emerges within the context of personal
ical, psychological, and social effects of history, current relationships, and future plans as
multiple diseases and their treatment. the adolescent lives daily life in dynamic interac-
Decision-making experiences with techno- tion with the environment.
logical interventions involving the begin- The researcher using qualitative methods
ning and end of life, as well as assistive and begins collecting bits of information and piecing
life-extending care. them together, building a mosaic or a picture of
Contextual factors favouring and mitigating the human experience being studied. As with a
against quality care, health promotion, pre- mosaic, when one steps away from the work, the
vention of disease, and reduction of health whole picture emerges. This whole picture tran-
disparities. The answers provided by quali- scends the bits and pieces and cannot be known
tative data re ect important evidence from any one bit or piece. In presenting study
that provides valuable insights about a findings, the researcher strives to capture the
 CHAPTER 8 Qualitative Approaches to Research 169

human experience and present it so that other review pp. 26 to 31 to refresh your memory.
people can understand it. Make sure that you can differentiate between
post-positivism and constructivism. In addition,
several other qualitative research methods are
QUALITATIVE RESEARCH METHODS brie y described here.
Thus far, the overview of the qualitative research
approach (see Chapter 7) has focused on the Phenomenological Method
importance of evidence offered by qualitative Pheno enology is a science whose purpose
research for nursing science. This overview has is to describe particular phenomena, or the
highlighted how choice of a qualitative approach appearance of things, as lived experience
is re ective of a researcher’s worldview and the (Streubert Carpenter, 2011, p. 73). Phenome-
research question. These topics provide a founda- nological research is used to answer questions of
tion for examining the qualitative methods dis- personal meaning. This method is most useful
cussed in this chapter. The Critical Thinking when the task is to understand an experience in
Decision Path introduces you to a process for rec- the way that people having the experience under-
ognizing differing qualitative methods by distin- stand it and is well suited to the study of phenom-
guishing areas of interest for each method and ena important to nursing. For example, what is
noting how the research question might be intro- the experience of men facing prostate surgery
duced for each distinct method. The phenomeno- What is the meaning of pain for people with
logical, grounded theory, ethnographic, case study, chronic arthritis Phenomenological research is
and participatory action research methods are an important method with which to begin study-
described in detail. Each of these research tradi- ing a new topic or a topic that has been studied
tions is discussed as shown in Figure 8-1 and are but needs a fresh perspective.
based on views along a continuum from post- Phenomenological research is based on phe-
positivist to constructivist to social critical theory. nomenological philosophy, which has changed
The constructivist paradigm (multiple realities) is over time and with different philosophers. Various
the basis of most qualitative research, and the posi- phenomenological methods exist, including the
tivist or contemporary empiricist paradigm (single following:
reality) is the basis of most empirical analytical or 1. Descriptive phenomenology, which focuses
quantitative research. The philosophical founda- on rich detailed descriptions of the lived
tion and assumptions of qualitative research are world and is based on Edmund Husserl’s
discussed in Chapter 2. You may wish to quickly philosophy.

Paradigm:
Post-positivism Constructivism Social critical theory
Research tradition:
Quantitative Grounded theory/Historical/Case study/Ethnographic/Phenomenological
(Empirical analytical)
Approach to research:
Falsify hypotheses Generate theory Describe Describe and interpret
FIGURE 8-1 Continuum of philosophical foundations and qualitative research methods.
 170

CRITICAL THINKING DECISION PATH

Selecting a Qualitative Research Method

If you are interested in

PART TWO Qualitative Research

Working with
Understanding Uncovering social Learning cultural Capturing Uncovering
people to improve
human experience processes patterns unique stories the past
the present

Then you will choose the

Phenomenological Grounded Ethnographic Case study Historical Participatory

method theory method method method method action reseach

And you may pose a question that begins…

What is How does this How does this What are the How did this What community
the human social group cultural group details and historical event practices can improve
experience of . . .? interact to . . .? express their complexities of influence nursing the health of this
pattern of . . .? the story of . . .? practice today? group of people?
 CHAPTER 8 Qualitative Approaches to Research 171

2. Heideggerian phenomenology, which ex- interviews with people who have directly experienced
pands description to understanding achieved the phenomenon of interest; that is, they have “lived
through searching for the relationships and experience” as opposed to second-hand experience.
(p. 104)
meanings of phenomena.
3. Hermeneutic philosophy, which focuses on The five important concepts or values in phe-
interpretation of phenomena. nomenological research (Cohen, 1987) are as
Derived from the Greek word hermeneuein, the follows:
term her eneutics refers to a theoretical frame- 1. The heno enological ethod is a process
work in which to understand or interpret human of learning and constructing the meaning
phenomena. Hermeneutic researchers believe that of human experience through intensive
interpretation cannot be absolutely correct or true dialogue with persons who are living the
but must be viewed from the perspective of the experience. This method was developed to
historical or cultural context and the original understand meanings. The goal was to
purpose of the text. Researchers use qualitative develop a rigorous science in the service of
methods to establish context and meaning for humanity. The goal of this science is to
what people do, and hermeneutists are much uncover the roots or foundations of a topic
clearer about the fact that they are constructing in order to clarify the basic concepts and
the reality on the basis of their interpretations what their meanings.
of data with the help of the participants who 2. Phenomenology was based on a critique of
provided data in the study (Patton, 2002, positivism, or the positivist view, which
p. 115). Hermeneutic researchers clearly outline was seen as inappropriate in the study of
their own perspectives and how they may in u- some human concerns.
ence the interpretation and analysis of the data. 3. The object of study is the life world eb-
In many nursing studies, the hermeneutic ens elt , or lived experience, not contrived
approach is used to understand a particular phe- situations. In other words, as the philoso-
nomenon and scientifically interpret phenomena pher Husserl said, researchers are concerned
from text or the written word (Streubert with the appearance of things phenomena
Carpenter, 2011). rather than the things themselves noumena
Patton (2002) described many of the different For example, think about a desk in a class-
phenomenological approaches in his text, uali- room. The desk is a real physical object, the
tative Research Evaluation Methods Although noumena, which people can see. If it were
he acknowledged the complexity and differing not visible, people would bump into the
traditions of these approaches, he also stated their desk every time they passed it. In addition,
similarities: your view of that desk, the phenomenon,
changes as you move in the room. If you sit
What these various phenomenological and at the desk, you see only the top of it.
phenomenographic approaches share in common is a
focus on exploring how human beings make sense of
However, as you move away, you can see
experience and transform experience into the desk’s legs, and so on. Nurses are often
consciousness, both as individuals and as shared interested in various aspects of people’s
meaning. This requires methodologically, carefully, experiences or views of health, illness, and
and thoroughly capturing and describing how people treatment.
experience some phenomenon—how they perceive it,
4. Intersub ectivity—a person’s belief that
describe it, feel about it, judge it, remember it, make
sense of it, and talk about it with others. To gather other people share a common world
such data, one must take undertake in-depth with him or her—is an important tenet in
172 PART TWO Qualitative Research

phenomenology. Although phenomena RESEARCH QUESTION. The question that guides

differ, they also share similarities that are phenomenological research always concerns
based on the similarities in people. The some human experience. It guides the researcher
most fundamental of those similarities is to ask the participant about some past or present
that every person has a body in space and experience. The research question is not exactly
time. In other words, the physical body and the same as the question used to initiate dialogue
historical sense lead to similarities in how with the participant, but often the research
people experience phenomena. The basic question and the question used to begin dialogue
elements, or an essence, of the shared expe- are very similar. de Witt and colleagues (2010)
rience are common among people or were interested in understanding the meaning
members of a specific society. of living alone from the perspective of older
5. The phenomenological reduction, also people with Alzheimer’s disease or a related
called brac eting, is controversial and more dementia.
important in some phenomenological
approaches than in others (e.g., descriptive RESEARCHER’S PERSPECTIVE. When using the phe-
methods). Phenomenological reduction nomenological method, the researcher’s per-
means that researchers must be aware of spective is bracketed that is, the researcher
and examine their prejudices or values. identifies personal biases about the phenomenon
Whatever the form of phenomenological research, of interest to clarify how personal experience
you will find the researcher asking a question and beliefs may in uence what is heard and
about the lived experience. reported. The researcher is expected to set aside
personal biases—to bracket them—when engag-
Identifying the Phenomenon ing with the participants. By becoming aware
Because the focus of the phenomenological of personal biases, the researcher is more likely
method is the lived e erience—the undergoing to be able to pursue issues of importance intro-
of events and circumstances, as opposed to think- duced by the participant, rather than leading
ing about these events and circumstances—the the participant to issues the researcher deems
researcher is likely to choose this method when important.
studying some dimension of day-to-day existence
for a particular group of people. The example Helpful Hint
used in this chapter is an article by de Witt and Managing personal bias is an expectation of
colleagues (2010) about the meaning of living researchers who use all the methods discussed in this
alone from the perspective of older people with chapter.
dementia.
The use of phenomenological methods always
Structuring the Study entails some strategy to identify personal biases
For the purpose of describing structuring, the fol- and hold them in abeyance while the participant
lowing topics are addressed: the research ques- is interviewed. The reader may find it difficult to
tion, the researcher’s perspective, and sample identify bracketing strategies because they are
selection. The issue of human participants’ protec- seldom explicitly identified in a research manu-
tion has been suggested as a dimension of struc- script. Sometimes, the researcher’s worldview or
turing (Parse, Coyne, Smith, 1985) this issue is assumptions provide insight into biases that have
discussed generally with ethics in Chapter 7. been considered and bracketed. Bracketing is not
 CHAPTER 8 Qualitative Approaches to Research 173

mentioned in de Witt and colleagues’ (2010) interview the participant and record the interac-
study again, this is not unusual and does not tion. In either case, the researcher may return to
detract from the quality of the report. sually, ask for clarification of written or recorded tran-
you will find some mention about bracketing if scripts. To some extent, the particular data-
such an issue exists, but not if there are no brack- collection procedure is guided by the choice of a
eting issues. specific analysis technique. Various analysis tech-
niques require different numbers of interviews.
Helpful Hint In general, open-ended questions—such as What
Although the research question may not always comes to mind when you think of . . . —guide
be explicitly reported, you may identify it by evaluating the participants to describe their lived experience.
the study’s purpose or the question/statement posed to
the participants. During the interview, the researcher attempts to
gather more information by asking clarifying
questions. Data saturation usually guides deci-
SAMPLE SELECTION. As you read a report of a phe- sions regarding how many interviews are enough.
nomenological study, you will find that the As described in Chapter 7, data saturation is
selected participant is either living the experience the point in a qualitative study when the informa-
the researcher is querying about or has lived the tion from the participants becomes repetitive, so
experience in their past. Because phenomenolo- that in interviews of additional participants, no
gists believe that each individual’s history is a new data emerge.
dimension of the present, a past experience exists de Witt and colleagues (2010) conducted 14
in the present moment. Even when a participant interviews in which the participants were asked
is describing an experience occurring in the to share their thoughts on (1) what it was like to
present, remembered information is being gath- live alone with memory loss, (2) continuing to
ered. The participants in the study by de Witt and live alone, (3) safety and living alone with
colleagues (2010) were eight women aged 58 to memory loss, (4) what it was like to need and ask
87 years old. All of the participants recruited for help, and (5) the future. The researcher was
through purposive sampling were white women able to conduct a second interview, 8 to 10 weeks
of European descent and fit the following inclu- later, with six of the participants of the others,
sion criteria: they (1) had mild to moderate two were unable to participate further because of
dementia, (2) had discussed their diagnosis with worsening of the disease process. In addition to
a physician, (3) lived alone, (4) spent the night the interview data, other sources of data included
alone, (5) were English speaking, and (6) were 55 a socioeconomic questionnaire, notes written
years of age or older. during and after the interviews by the researcher,
de Witt and colleagues’ re exive journal, and a
Helpful Hint letter received from one of the participants.
Qualitative studies often involve the use of pur-
posive sampling (see Chapter 7). Data Analysis
As data are collected, data analysis begins.
Data Gathering Several techniques are available for data analysis
Written or oral data may be collected when the when the phenomenological method is used.
phenomenological method is used. The researcher Detailed information about specific techniques is
may pose the query in writing and ask for a available in the original sources (Colaizzi, 1978
written response or may schedule a time to Giorgi, Fischer, Murray, 1975 Spiegelberg,
174 PART TWO Qualitative Research

1976 van Kaam, 1969). Although the techniques codes were grouped together into subthemes,
are slightly different from each other, there is a and finally they were grouped together as overall
general pattern of moving from the participant’s themes.
description to the researcher’s synthesis of all Giving verbatim transcripts to participants can
participants’ descriptions. The steps generally have unanticipated consequences. It is not unusual
include the following: for people to deny that they said something in a
1. Thorough and sensitive reading of presence certain way or that they said it at all. Even when
with the entire transcription of the partici- the actual recording is played for them, they may
pant’s description have difficulty believing it. This is one of the
2. Identification of shifts in participant thought, more challenging aspects of any qualitative
resulting in division of the transcription into method: Every time a story is told, it changes for
thought segments the participant. The participant may sincerely
3. Specification of significant phrases in each believe that the story-as-recorded is not the story
thought segment, in the participant’s own as it is now after being described.
words
4. Distillation of each significant phrase to Describing the Findings
express the central meaning of the segment When using the phenomenological method, the
in the researcher’s words nurse researcher constructs a path of information
5. Grouping together of segments that contain leading from the research question through
similar central meanings for each parti- samples of participants’ words and the research-
cipant er’s interpretation to the final synthesis that elabo-
6. Preliminary synthesis of grouped segments rates the lived experience. When reading the
for each participant with a focus on the report of a phenomenological study, you should
essence of the phenomenon being studied find that detailed descriptive language is used to
7. Final synthesis of the essences that have convey the complex meaning of the lived experi-
surfaced in all participants’ descriptions, ence. de Witt and colleagues (2010) provided
resulting in an exhaustive description of the numerous quotations from participants to support
lived experience their findings. They identified a main theme of
de Witt and colleagues (2010) utilized three holding back time and four subthemes of (1)
techniques developed by van Manen (1997) to stored time, (2) dreaded time, (3) holding on to
isolate themes during their data analysis: (1) the the present, and (4) limited time. These themes
wholistic approach of reading each transcript in described the emotions, vulnerability, challenges,
its entirety to understand the overall meaning, (2) and issues experienced by these women. Direct
the selective approach to identify meaningful quotations from participants enable you to evalu-
portions of transcript text, and (3) the line-by- ate the connection between what the participant
line approach to discover what each line reveals said and how the researcher labelled it. For
about the participant’s experience. The research- example, the subtheme of dreaded time
ers then developed a summary paragraph from describes how the participants chose to hold on
each of the first interviews and shared this with to the present, because looking ahead to their own
the participants during the second interview, in future of worsening dementia was framed by their
which the participants had the opportunity to past experiences of observing other people with
make any changes. These summary paragraphs dementia. Becoming worse is the dreaded time.
also guided the second interview. Each of the The following quotation expresses a partici-
interview transcripts was then coded, individual pant’s dread of the future.
 CHAPTER 8 Qualitative Approaches to Research 175

Well, you saw other people (in the adult day program) about experience by including the personal and
in the worst stages and then you think ‘I’m gonna be social over time in particular places is personal,
like that . . . it’s almost the way you feel as if you’ve re ective and relational work that is autobio-
gone to the lowest end when you can’t even go to the
bathroom by yourself.
graphically meaningful and socially significant
(p. 62).
Evidence-Informed Practice Tip
Orientational Qualitative Inquiry
Phenomenological research is an important
approach for accumulating evidence when researchers In orientational ualitative in uiry an ideology
study a new topic about which little is known. or orientation is used to direct the investigation,
including the research question, methodology,
fieldwork, and analysis of the findings. Ideologies
Narrative Analysis include feminist, queer, and critical theories
When narrative in uiry is used as a form of (Patton, 2002). For example, a feminist researcher
qualitative research, stories of people are col- presumes that gender in uences all relationships
lected and examined as the primary source of and societal processes. The researcher will attend
data (Duffy, 2011). The hermeneutic tradition is to women’s ways of knowing and include the
extended to include in-depth interview tran- participants throughout the research. ueer
scripts, memoirs, stories, and creative nonfiction. theory, which emerged from feminist theory,
This discipline also draws from the phenomeno- focuses on sexual orientation and activities. For
logical tradition in its interest in the lived experi- example, Holmes and associates (2006, 2007)
ence and perceptions of experience. On the basis explored why gay men continue to have unpro-
of the stories of people, at times including those tected sex despite the associated health risks.
of the researcher, researchers using narrative Critical theory focuses on issues of power and
analysis attempt to interpret and understand expe- justice and how injustice and subjugation shape
riences in terms of cultural and social meanings people’s experiences and their view of the world.
(Patton, 2002 see Practical Application box). As Rather than studying to understand, the critical
Lindsay (2006) noted, Thinking narratively theorist attempts to critique society, name injus-
tices, and change society. Nurses are particularly
Practical Application interested in addressing and changing oppressive
practices that in uence health and health care
Lapum and colleagues (2010) used narrative
inquiry to examine patients’ experiential (Browne, 2000). Smith and colleagues (2006)
accounts of technology in open-heart surgery and took a critical postcolonial stance in their study
recovery. Data were collected from two interviews in on pregnancy and parenting experiences among
which the participants were encouraged to tell their
story through prompts and questions, such as “Tell Aboriginal peoples. In the postcolonial stance,
me about waking up from surgery.” The participants issues of power are viewed in terms of the legacy
also documented their experiences in journals for of the colonialization of Aboriginal peoples and
several weeks after surgery. The researchers listened,
heard, and felt through the stories what was the neocolonial present.
happening to the participants. Although focus was
placed on the content of the stories, emphasis was Grounded Theory Method
also placed on how the stories were put together:
In the grounded theory ethod a systematic
“We attended to facets of temporality, contextuality,
plot, scene, and characters in order to understand set of procedures is used to explore the social
processes and activities involved in narrative processes that guide human interaction and
emplotment” (Lapum, Angus, Peter, & Watt-Watson, to inductively develop a theory on the basis of
2010, p. 756).
those observations. The philosophical spectrum
176 PART TWO Qualitative Research

of grounded theory ranges from the post- The purpose of grounded theory, as the name
positivist view to the constructivist view (see implies, is to generate a theory from data.
Figure 8-1). The grounded theory method is based Grounded theory has contributed substantively to
on the sociological tradition of the Chicago the body of knowledge in the field of nursing.
School of Symbolic Interactionism, a tradition Often, the theories generated from grounded
that re ects on issues related to human behaviour. research are then tested empirically. ualitative
Glaser and Strauss (1967) developed the method data are gathered through interviews and observa-
of grounded theory and published the classic first tion. Through analysis of the data, substantive
text describing the methodology: The Discovery codes are generated and then are clustered into
of rounded Theory. According to Strauss and categories. Pro ositions link the concepts to
Corbin (1990), grounded theory create a foundation that guides further data col-
lection. Additional data that are thought likely to
is one that is inductively derived from the study of the answer generated hypotheses are collected until
phenomenon it represents. That is, it is discovered,
all categories are saturated that is, no new
developed, and provisionally verified through
systematic data collection and analysis of data information is generated. The goal of generating
pertaining to that phenomenon. Therefore, data a theory implies that laws drive at least some
collection, analysis, and theory stand in reciprocal portion of reality. The truth is sought from rele-
relationship with each other. One does not begin with vant groups: for example, patients who are dying.
a theory, then prove it. Rather one begins with an area The context is very important, as was shown
of study and what is relevant to that area is allowed
to emerge. (p. 23)
in a classic work by Glaser and Strauss (1965).
They noted that, at the time of their work, patients
In many qualitative research traditions, explan- were unwilling to talk openly about the process
atory models and theories are described and of their own dying, physicians were unwilling to
developed in relation to a human phenomenon disclose the imminence of death to patients, and
under study grounded theory is distinctive from nurses were expected not to make these disclo-
the other traditional qualitative research methods sures. This lack of communication led Glaser and
because its primary focus is on generating theory Strauss to their study of the problem of awareness
about dominant social processes. The three major of dying. They described various types of aware-
premises that continue to underlie grounded ness contexts, problems of awareness, and practi-
theory research are outlined in Box 8-1. cal uses of awareness theory. Their early fieldwork
led to hypotheses and the gathering of additional
BOX 8-1 data, and the framework was refined with further
MAJOR PREMISES OF GROUNDED THEORY analysis until they formed a systematic substan-
1. Humans act toward objects on the basis of the tive theory.
meaning that those objects have for them. Meaning
is embedded in context and, therefore, it cannot be Identifying the Phenomenon
separated from the context or from the
consequences of the meanings in a particular
Researchers typically use the grounded theory
setting. method when they are interested either in social
2. Social meanings arise from social interactions with processes from the perspective of human interac-
other people over time and are embedded socially, tions or in patterns of action and interaction
historically, culturally, and contextually. Therefore,
the focus of grounded theory is on social between and among various types of social units
interactions. (Denzin Lincoln, 1998). The basic social
3. People use interpretive processes to handle and process is sometimes expressed as a gerund, indi-
change meanings in dealing with their situations.
cating change across time as social reality is
 CHAPTER 8 Qualitative Approaches to Research 177

negotiated. Schrieber and MacDonald (2010) participants who can further clarify the emerging
explored the role and practice of nurse anaesthe- concepts. Schrieber and MacDonald (2010)
tists, with a focus on how the practice is part of recruited participants at a general meeting for
the nursing domain. nurse anaesthetists through purposive and, sec-
ondarily, through snowball sampling. Sno ball
Structuring the Study sa ling occurs when a participants recom-
RESEARCH QUESTION. Research questions appropri- mends other participants from their contacts. As
ate for the grounded theory method are those that more and more participants bring on new recruits,
address basic social processes that shape human the sample appears to grow like a snowball.
behaviour. In a grounded theory study, the Theoretical sampling occurs as key informants
research question can be a statement or a broad are sought to provide clarification on issues such
question that permits in-depth explanation of the as regulatory and legal issues. ey in or ants
phenomenon. For example, in Schrieber and are individuals who have special knowledge,
MacDonald’s (2010) study, the aim of the study status, or communication skills and who are
was to explore and develop a theory of nurse willing to teach the researcher about the phenom-
anaesthetist practice. The researcher does not enon (Creswell, 1998).
always need to identify a problem or research
question but chooses an area of interest. Data Gathering
In the grounded theory method, data are collected
RESEARCHER’S PERSPECTIVE. In a grounded theory through interviews and through skilled observa-
study, the researcher brings some knowledge of tions of individuals interacting in a social setting.
the literature to the study, but an exhaustive Interviews are audio-recorded and then tran-
literature review is not performed (Streubert scribed, and observations are recorded as field
Carpenter, 2011). Therefore, theory emerges notes. Open-ended questions are used initially to
directly from data and re ects the contextual identify concepts for further focus. Schrieber and
values that are integral to the social processes MacDonald (2010) interviewed 18 nurse anaes-
being studied. Thus, the theory product that thetic practitioners, leaders, and students. They
emerges is grounded in the data. This type of gathered documentation and observed the confer-
study was exemplified in Schrieber and MacDon- ence sessions, convention rituals, and mentoring
ald’s (2010) article, in which a background or relationships. In addition, they interviewed key
literature section was not included. informants familiar with the regulatory and legal
issues, scope of practice, and financial aspects.
SAMPLE SELECTION. Sample selection involves (1) Field notes were also part of the data collection
choosing participants for a purposive sample who they enabled the researchers to document and
are experiencing the circumstance and (2) select- have a fuller understanding of their observations
ing events and incidents that are related to the and conversations.
social process under investigation and are judged
to have good knowledge of the study domain Data Analysis
(Wuest, 2011, p. 235). As problems begin to A major feature of the grounded theory method
emerge, the researchers may conduct theoretical is that data collection and analysis occur simulta-
sa ling a sampling method used to select neously. The process requires systematic, detailed
experiences that helps the researchers test ideas record keeping through the use of field notes and
and gather complete information about develop- transcribed interview recordings. Hunches about
ing concepts. In this method, researchers seek emerging patterns in the data are noted in memos,
178 PART TWO Qualitative Research

and the researcher directs activities in the field by

Evidence-Informed Practice Tip
pursuing these hunches. This technique of theo-
When you think about the evidence generated by
retical sampling is used to select participants the grounded theory method, consider whether the
whose experiences that will help the researcher theory is useful in explaining, interpreting, or predicting
test ideas and gather complete information about the study phenomenon of interest.
developing concepts. The researcher begins by
noting indicators or actual events, actions, or for patients, becoming patient surrogates and
words in the data. Concepts, or abstractions, are taking over vital body functions (p. 554). They
developed from the indicators (Charmaz, 2000 keep this vigil by using four interconnected strat-
Strauss, 1987). egies: engaging with patient, finessing the human-
The initial analytical process is called open technology interface, massaging the message,
coding (Strauss, 1987). Data are examined care- and foregrounding nursing. Figure 8-2 visually
fully line by line, categorized into discrete parts, presents the interconnections among the strate-
and compared for similarities and differences gies. The researchers also concluded that the
(Corbin Strauss, 2008). Data are compared nursing was clearly evident in nurse anaesthesia
with other data continuously as they are acquired practice.
during research. This is a process called the con
stant co arative ethod Codes in the data are
clustered to form categories. The categories are Helpful Hint
In a report of research in which the grounded
expanded and developed, or they are collapsed theory method was used, you can expect to find a
into one another. Theory is constructed through diagrammed model of a theory in which the research-
this systematic process. As a result, data collec- er’s findings are synthesized in a systematic way.
tion, analysis, and theory generation have a direct
reciprocal relationship (Charmaz, 2000 Strauss
Corbin, 1990). Schrieber and MacDonald Ethnographic Method
(2010) used the constant comparative method Ethnographic research has a long history in the
described earlier. qualitative research tradition and is considered by
some authorities to be the oldest of the traditions
Describing the Findings (Roberts, 2009). Anthropologists developed the
Grounded theory studies are reported in sufficient ethnogra hic ethod defined as a method of
detail to provide the reader with the steps in the scientifically describing cultural groups. Scotch
process, the logic of the method, and the theory (1963) first described medical anthropology as
that has emerged. In reports of grounded theory the field of research that focuses on health and
studies, descriptive language and diagrams of the illness within a cultural system. Nurses also
process are used as evidence to ensure that the conduct medical ethnographic studies that focus
theory reported in the findings remains connected on health and illness within a cultural system
to the data. Instead of providing a description of (Roper Shapira, 2000). Although early work
people’s experiences, the focus is to provide theo- addressed the cultural patterns of village life,
retical statements about the relationships between often in distant locations, nurses now often
the concepts (Wuest, 2011). Schrieber and Mac- conduct focused ethnographic research: the study
Donald (2010) described the key concept of of distinct problems within a specific context
keeping patients safe throughout surgical proce- among a small group of people, or the study of
dures. The nurse anaesthetists see themselves as a group’s social construction and understanding
having a sacred trust to ensure the best outcomes of a health or illness experience (Roper
 CHAPTER 8 Qualitative Approaches to Research 179

ing
blish ity Po
Esta credibil Foregrounding nursing
vig litica
lic ilan l
pub ce

Massaging the
Finessing the message
Engaging with human
the patient techn0logy
interface

Tending the flock

FIGURE 8-2 Keeping vigil over the patient.

From Schrieber, R., & MacDonald, M. (2010). Keeping vigil over the patient: A grounded theory
of nurse anaethesia practice. Journal of Advanced Nursing, 66, 552–561. doi:10.1111/j.1365-
2648.2009.05207.x. © 2010 Blackwell Publishing Ltd.

Shapira, 2000). Leininger (1985) developed an group’s patterns of behaviour and customs,
ethnographic research method called ethnonurs- its way of life, and what it produces. The
ing, which has since been redefined as a cognitive ers ective is the view that
rigourous, systematic, and in-depth method for culture consists of the beliefs, knowledge,
studying multiple cultures and care factors within and ideas people use as they live. Culture
familiar environments of people and to focus on refers to the structures of meaning through
the interrelationships of care and culture to arrive which people shape experiences.
at the goal of culturally congruent care services nderstanding culture requires a holistic
(Leininger, 2006, p. 20). perspective that captures the breadth of the
Ethnogra hy (ethnographic research) is the beliefs, knowledge, and activities of the
study of cognitive models or patterns of behav- group being studied.
iour of people within a culture. Ethnographers Conte t—the personal, social, and political
seek to understand another way of life from the environment in which a phenomenon of
perspective of the people experiencing it. The interest (time, place, cultural beliefs, values,
following values underlie ethnography: and practices) occurs—is important for an
Culture is fundamental to ethnographic understanding of a culture. nderstanding
studies. Culture includes behavioural/ this context requires intensive face-to-face
materialist and cognitive perspectives. contact over an extended period of time.
Through the behavioural aterialist er People are studied where they live, in their
s ective culture is observed through a natural settings, or where an experience
180 PART TWO Qualitative Research

occurs, such as in a hospital or community their work to women. They also consider and
setting. analyze the effects of race, class, culture, ethnic-
The aim of ethnographic research is to ity, sexual preference, and other identities as
combine the e ic ers ective (the insider’s forces that cause and sustain oppression (Mac-
view of the world) with the etic ers ective quire, 1996). For example, McDonald (2006)
(the view of the researcher outsider ) to recruited 15 gay women from a university in
develop a scientific generalization about dif- Western Canada and interviewed them to under-
ferent societies. In other words, generaliza- stand the experience of lesbians who disclose
tions are drawn from special examples or their sexual orientation.
details from participant observation. Ethnogeriatrics, as the name implies, focuses
An example of ethnographic work that has been on examining the health and aging issues in the
useful to nurses is the notion of explanatory context of cultural beliefs, values, and practices
models. This idea was developed most by cogni- among racial and ethnic minority elders (Fitz-
tive anthropologists, especially Kleinman (1980). patrick Wallace, 2006, p. 179). Ethnogeriatric
Explanatory models use an interactive approach, researchers are interested in the disparities facing
emphasizing variations between patients’ and older adults from racial and ethnic minorities.
practitioners’ models of illness. They offer expla- The hope is to develop nursing knowledge and
nations of sickness and treatment, guide choices culturally appropriate interventions to guide
among available therapies and therapists, and health care systems to be more inclusive of this
give social meaning to the experience of sickness. patient population.
These cognitive models vary over time and in
response to a particular episode of illness. Identifying the Phenomenon
Several ethnographic schools of thought exist, The phenomenon under investigation in an ethno-
three of which are of particular interest to nurse graphic study varies in scope from a long-term
researchers: critical, feminist, and ethnogeriatric. study of a very complex culture, such as that of the
Critical ethnography does not entail the use of Aborigines (Mead, 1949), to a shorter-term study
different methods instead, it focuses on beliefs of a phenomenon within subunits of cultures.
and practices that limit human freedom, justice, Kleinman (1992) notes the clinical utility of eth-
and democracy ( sher, 1996). Critical ethno- nography in describing the local world of groups
graphic researchers make their values explicit. In of patients who are experiencing a particular phe-
other words, they document tacit rules that govern nomenon, such as suffering. The local worlds of
human interaction and behaviour. They also patients have cultural, political, economical, insti-
explore how dominant social groups oppress tutional, and social-relational dimensions in much
those in the minority or those without power. For the same way as larger complex societies do. Sen-
example, Pesut and Reimer-Kirkham (2010) con- eviratne and associates’ (2009) study on nurses’
ducted a critical ethnographic study to describe perceptions of the contexts of caring for acute
how religious and spiritual values and beliefs are stroke survivors (see Appendix A) provides an
negotiated in encounters between health care pro- introduction to ethnography.
fessionals and health care recipients. Finally,
many critical ethnographers consider the study Structuring the Study
participants to be co-investigators and explore RESEARCH QUESTION. When you review a report of
problems and possible solutions with them. ethnographic research, notice that questions are
Feminist researchers, like critical ethnogra- asked about lifeways, or particular patterns of
phers, focus on oppression and power but apply behaviour within the social context of a culture
 CHAPTER 8 Qualitative Approaches to Research 181

or subculture. Culture is viewed as the system of interviews of informants, and interpretation by

knowledge and linguistic expressions used by the researcher of cultural patterns (Crabtree
social groups that allows the researcher to inter- Miller, 1992). According to Boyle (1991), ethno-
pret or make sense of the world of those groups graphic research in nursing, as in other disci-
(Aamodt, 1991). Ethnographic nursing studies plines, involves interviewing in the natural setting
address questions that concern how cultural as the major data-collection method. Thus, field-
knowledge, norms, values, and other contextual work is a major focus of the method. Other tech-
variables in uence a person’s health care experi- niques may include obtaining life histories and
ence. Ethnographers have a broader definition of collecting material items re ective of the culture.
culture, whereby a particular social context is Photographs and films of the informants in their
conceptualized as a culture. In this case, in Sen- world can be used as data sources. Spradley
eviratne and associates’ (2009) study on under- (1979) identified three categories of questions for
standing nursing on an acute stroke unit, the acute ethnographic inquiry: descriptive, or broad, open-
stroke unit is seen as a culture appropriate for ended questions structural, or in-depth, questions
ethnographic study (see Appendix A). that expand and verify the unit of analysis and
contrast questions, which further clarify and
RESEARCHER’S PERSPECTIVE. When a researcher uses provide criteria for exclusion.
the ethnographic method, the researcher’s per- In Seneviratne and associates’ (2009) study,
spective is that of an interpreter entering an alien data were collected from observations obtained
world and attempting to make sense of that world by monitoring the unit staff (20 individuals) and
from the insider’s point of view (Agar, 1986). from interviews with nine of those staff. The
Like phenomenologists and grounded theory fieldwork took place over a period of 8 months
researchers, ethnographers make their own beliefs for an average of 2 to 3 hours three times a week.
explicit and bracket, or set aside, their personal Initially, the fieldworker began with 3 months of
biases as they interpret the findings and seek to general observation while located at the charting
understand the worldview of other people. desk, located at the nursing station, and walking
in the hall. After the field notes from each day
SAMPLE SELECTION. The ethnographer selects a cul- were reviewed, a more focused approach was
tural group that is experiencing the phenomenon taken to fill in the gaps. To further explore work
under investigation. The researcher gathers infor- practices, participants were interviewed with
mation from general informants and from key grand tour questions such as Can you walk me
informants. Seneviratne and associates’ (2009) through your typical day Of importance is that
research took place in an 18-bed acute stroke unit although nurses were the focus of the study, four
that provided specialized interventions, manage- other health professionals were interviewed in
ment, and investigative care, located in a large order to contextualize interprofessional perspec-
tertiary medical centre in Canada. The staff tives (see Appendix A).
included nurses licensed practical nurses patient
care attendants a nurse practitioner physicians Data Analysis
and physical, occupational, and speech therapists As with other qualitative methods, data are col-
(see Appendix A). lected and analyzed simultaneously. Data analy-
sis proceeds through several levels as the
Data Gathering researcher looks for the meaning of cultural
Ethnographic data gathering involves partici- symbols in the informant’s language. Analysis
pant observation or immersion in the setting, begins with a search for do ains or symbolic
182 PART TWO Qualitative Research

categories that include smaller categories. Lan- nursing on an acute care unit: perceptions of
guage is analyzed for semantic relationships, and space, time, and interprofessional practice. Each
structural questions are formulated to expand and of the three domains was further divided into
verify data. Analysis proceeds through increasing theme units. For example, in the domain of space,
levels of complexity until the data, grounded nurses described space as a challenge to patient
in the informant’s reality and synthesized by care with three themes: nursing in a submarine,
the researcher, lead to hypothetical propositions nursing too close, and nursing in a state of
about the cultural phenomenon under investiga- code burgundy a code burgundy referred to a
tion. Creswell (1998) provided a detailed descrip- lack of beds, so that nurses had to care for patients
tion of the ethnographic analysis process. in the hallway, which resulted in increased work-
Seneviratne and associates (2009) described a load and the ethical challenges associated with
three-step data-analysis plan. Initially, they iden- hallway care (see Appendix A).
tified three main themes and then further identi-
fied theme components. They then categorized Evidence-Informed Practice Tip
the work activities into types on the basis of rela- Evidence generated by ethnographic studies
tionships between nurses and then between nurses answers questions about how cultural knowledge,
and the other health care professionals. They norms, values, and other contextual variables influence
the health experience of a particular patient population
cross-checked their findings by reading the field in a specific setting.
notes and interview transcripts and then returning
to the field to make further observations (see
Appendix A). Case Study Method
Case study as a research method involves an
Describing the Findings in-depth description of the essential dimensions
In ethnographic studies, field notes of observa- and processes of the phenomenon being studied.
tions, interview transcriptions, and sometimes Case study research, rooted in sociology, is
other artifacts such as photographs yield large described slightly differently by major thinkers
quantities of data. Charmaz (2000) provided who write about this method, such as Yin,
guidelines for ethnographic writing that you can Stake, Merriam, and Creswell (Aita McIlvain,
use when you wish to critique descriptions of 1999). For the purpose of introducing you to this
ethnographic studies. The five techniques recom- research method, Stake’s view is emphasized.
mended in Charmaz’s guidelines are pulling the The case study ethod is about studying the
reader in, re-creating experiential mood, adding peculiarities and the commonalities of a specific
surprising observations, reconstructing ethno- case over time to provide an in-depth description
graphic experience, and creating closure for the of the essential dimensions and processes of the
study. When you critique, be aware that the report phenomenon—familiar ground for practising
of findings usually provides examples from data, nurses. Stake (2003) noted that case study is not
thorough descriptions of the analytical process, a methodological choice but rather a choice of
and statements of the hypothetical propositions what to study. Case study can include quantitative
and their relationship to the ethnographer’s frame or qualitative data, or both, but it is defined by its
of reference. focus on uncovering an individual case. Stake
Evidence provided by complete ethnographies (2003) distinguished intrinsic from instrumental
may be published as monographs. Seneviratne case study. Intrinsic case study is research that
and associates (2009) described three local is undertaken to have a better understanding of
domains that frame how nurses understand the case—nothing more or nothing less. The
 CHAPTER 8 Qualitative Approaches to Research 183

researcher at least temporarily subordinates other Stake’s (1995, 2003) defining criterion of atten-
curiosities so that the stories of those living the tion to the single case broadens the scope of phe-
case’ will be teased out (Stake, 2003, p. 122). nomenon for study. By using a single case, Stake
Instru ental case study is defined as research designated a focus on an individual, a family, a
that is performed when the researcher is pursuing community, an organization: some complex phe-
insight into an issue or wants to challenge some nomenon that mandates close scrutiny for under-
generalization. standing. Maddalena and colleagues (2010)
Case studies can be used for a variety of pur- wanted to examine the experiences of caregivers
poses, including to present data gathered with of patients of African Canadian descent who died
another method, as a teaching device, or as a of cancer. They chose three case studies of fami-
research method (Yin, 1994). Case studies have lies of African Canadian descent in Nova Scotia.
been used in various disciplines, including
nursing, political science, sociology, business, Structuring the Study
social work, economics, and psychology. Nurses RESEARCH QUESTION. The research question for a
have a long and continuing tradition of using case case study is one that provokes the curiosity of
studies for teaching and learning about patients the researcher. Stake (2003) suggested that
(e.g., Parsons, 1911). Nightingale (1858/1969) research questions be developed around issues
stressed the importance of coming to know that serve as a foundation to uncover complexity
patients and of basing practice on experience. She and pursue understanding. Although researchers
noted that knowing how to provide care requires pose questions to begin discussion, the initial
the nurse to learn about the patient’s life. In case questions are never all-inclusive. Rather, the
studies, these details are described, and the researcher uses an iterative process of growing
lessons that can be learned from the particular questions in the field that is, as data are col-
patient are made clear. Persons who have had a lected to address these questions, other questions
particular experience can provide insights that are emerge to guide the researcher in the process of
both valuable and unavailable to those who have untangling the complex story. Therefore, research
not had the experience. Obtaining these descrip- questions evolve over time and re-create them-
tions through the use of case studies can serve a selves in case study research. Maddalena and col-
variety of functions: making practitioners and leagues (2010) were initially interested in the
researchers aware of patients’ experiences clari- experiences and recollections of the caregivers
fying the concepts included in an experience or of African Canadian patients with cancer at
general label policy decision making and theory the end of life. Later, they also became interested
building by identifying hypotheses for testing in the use of complementary and alternative
with further research (Cohen Saunders, 1996). therapies and home remedies used at the end of
Maddalena and colleagues (2010) examined the life.
experiences and recollections of primary caregiv-
ers of African Canadians who died from cancer. RESEARCHER’S PERSPECTIVE. When the researcher
The researchers were also interested in the use of begins with questions developed around sus-
complementary and alternative medicine and pected issues of importance, the perspective of
home remedies at the end of life. the researcher is re ected in the questions this is
sometimes referred to as an etic perspective As
Identifying the Phenomenon the researcher begins engaging the phenomenon
Although some definitions of case study demand of interest, the story unfolds and leads the way,
that the focus of research be contemporary, shifting from an etic (researcher-based) to an
184 PART TWO Qualitative Research

emic (story-based) perspective (Stake, 2003). The

reader may recognize a shift from etic to emic Data Gathering
perspective when stories spin off of the original Data are gathered through the use of interview,
questions posed by the researcher. Maddalena observation, document review, and any other
and colleagues (2010), with the exception of the methods by which researchers accumulate evi-
first author, self-identified as African Canadian, dence that enables understanding of the complex-
which provided, at least initially, a strong etic ity of the case. The researcher will do what is
perspective. needed to get a sense of the environment and the
relationships that provide the context for the case.
SAMPLE SELECTION. Sample selection is one of the Stake (1995) advocates development of a data-
areas of which scholars in the field present dif- gathering plan to guide the progress of the study
fering views, ranging from choosing only the from definition of the case through decisions
most common cases to choosing only the most regarding reporting. You may find little explicit
unusual cases (Aita McIlvain, 1999). Stake information about data gathering in the report of
(2003) advocated selecting cases that may offer research. The primary source of data in Mad-
the best opportunities for learning. For example, dalena and colleagues’ (2010) study was indi-
if several heart transplant recipients are available vidual in-depth interviews with the primary
for the researcher to study, practical factors will caregivers. Secondary caregivers were also inter-
in uence the opinion of which patient offers the viewed to provide more depth and thus facilitated
best opportunity for learning. It would be more triangulation. Follow-up interviews were also
practical to study patients who live in the area conducted to seek clarification of issues raised in
and can be easily visited at home or in the the first interview.
medical centre than to study someone living in
another country. The researcher may want to
choose someone who has an actively participat- Data Analysis and Describing Findings
ing family, because most transplant recipients Data analysis is closely tied to data gathering and
reside in a family setting. No choice is perfect description of findings as the case study story is
when a case is selected. There is much to learn generated: ualitative case study is character-
about any one individual, situation, or organiza- ized by researchers spending extended time, on
tion during case study research, regardless of the site, personally in contact with activities and
contextual factors in uencing the unit of analy- operations of the case, re ecting, revising
sis. sing purposive sampling, Maddalena and descriptions and meanings of what is going on
colleagues (2010) followed specific inclusion (Stake, 2003, p. 450). Re ecting and revising
criteria: (1) that the caregivers were African meanings are the work of the case study
Canadian and (2) that they provided care for researcher, who has recorded data, searched for
someone who had died in the previous 3 years. patterns, linked data from multiple sources, and
Three case studies were examined: one rural devised preliminary thoughts regarding the
family, one urban family, and one recently meaning of collected data. This re ective
arrived immigrant family. From these families, a dynamic evolution is the iterative process of cre-
total of seven participants—three primary and ating the case study story that can be thought of
four secondary caregivers—were identified. The as the evidence. You may have difficulty deter-
different experiences of the primary and second- mining how data analysis was conducted because
ary caregivers provided a richer understanding the research report generally does not list research
of the experience and validated the findings. activities. Findings are embedded in (1) a
 CHAPTER 8 Qualitative Approaches to Research 185

chronological development of the case (2) the collection, organization, and critical appraisal of
researcher’s story of coming to know the case facts. One of the goals in historical methodology
(3) the descriptions of individual case dimen- is to shed light on the past so that it can guide the
sions and (4) vignettes that highlight case quali- present and the future: Through historical
ties (Stake, 1995). In Maddalena and colleagues’ research, we can better understand how nurses in
(2010) study, the verbatim-transcribed interviews the present can assume control of their practice,
were coded manually individually by four education, and roles in the contemporary health-
members of the research team. The team met to care system (Lundy, 2011, p. 383). The attention
compare coding and engage in the analysis of the in nursing to historical methodology was initiated
data. Once the team reached consensus, a the- by Teresa E. Christy, who elaborated the method
matic and discourse analysis was used to further (Christy, 1975) and the need (Christy, 1981) for
analyze the data. As prevalent themes emerged, historical research long before most nurse schol-
each was explored in the context of how culture ars accepted it as a legitimate research method.
in uenced their experiences during the time from More recently, Lusk (1997) summarized impor-
initial diagnosis through interactions with the tant information for the nurse interested in under-
health system to death and bereavement. With the standing historical research. She provided
researchers’ etic view of the black community guidance for choosing a topic, acquiring data,
and culture, the team made sense of the data addressing ethical issues, analyzing data, and
through the shared historical and cultural experi- reporting findings.
ences of the black community. sing a discourse When you appraise a study in which the his-
analysis approach, the researchers examined each torical method was used, expect to find the
of the themes in terms of their social, political, research question embedded in the phenomenon
and historical contexts. The researchers found to be studied. The question is stated implicitly
that the end of life for African Canadians was rather than explicitly.
characterized by end-of-life care provided by The three theoretical frameworks that guide
family in the home setting, community involve- historical research are as follows (Streubert
ment, a focus on spirituality, and a preference for Carpenter, 2011):
home care over institutionalized care. In the 1. Biographical history: an exploration of the
home setting, the caregivers were faced with a life of an individual to understand the
myriad of challenges. Common among the three effects of the time and culture on the per-
case studies was the use of complementary and son’s life
alternative methods and home remedies. In each 2. Social history: an exploration of the prevail-
of the three case studies, the use of prayer was ing values and beliefs in a particular period
considered a complementary method. by examining everyday events
3. Intellectual history: an exploration of the
Evidence-Informed Practice Tip ideas of a particular individual or a group
Case studies are a way of providing in-depth, of people
evidence-informed discussion of clinical topics that can An example of biographical history is the book
be used to guide practice. about Gertrude Richard Ladner (i.e., her family
and nursing life), in which the authors presented
new ideas about nursing and family life in the late
Historical Research Method nineteenth and early twentieth centuries in
The historical research ethod is a systematic Western Canada ( err, ilm, Grant, 2006).
approach for understanding the past through Although the authors believed that Ladner
186 PART TWO Qualitative Research

represented the ordinary life of women and nurses Are the ink, paper, and wax seal on the
in that time, they discovered that she was more envelope representative of Nightingale’s
than a handmaiden for the physicians with whom time
she worked. Many of the details included in the Is the wax seal one that Nightingale used in
book were gathered from her personal journal, other authentic data sources
providing a record of the work of nurses, their Is the writing truly Nightingale’s
day-to-day experiences, and nursing knowledge Only if the data source passes the test of external
that informed their practice. criticism does the researcher begin internal criti-
Reeves and associates (2010) applied a socio- cism. Internal criticis is the process of judging
historical analysis of historical documents to the reliability or consistency of information
understand how modern health care professions within the historical document (Christy, 1975).
emerged from sixteenth century craft guilds. To judge reliability, the researcher must become
Their analysis provides an understanding of how familiar with the time in which the data emerged.
the historical practice of protecting and promot- A sense of the context and language of the time
ing one’s own members of the guild is one of the is essential to understanding a document. The
roots of today’s barriers to effective collaboration meaning of a word in one era may not be equiva-
and interprofessional teamwork. lent to the meaning in another era. Knowing the
Data sources provide the sample for historical language, customs, and habits of the historical
research. The more clearly a researcher delineates period is critical for judging reliability. The
the historical event being studied, the more spe- researcher assumes that a primary source pro-
cifically data sources can be identified. Data may vides a more reliable account than does a second-
include written or video documents, interviews ary source (Christy, 1975). The further a source
with persons who witnessed the event, photo- is from an eyewitness account, the more question-
graphs, and other materials that shed light on the able is its reliability. The researcher using histori-
subject. Sometimes pivotal information cannot be cal methods attempts to establish fact, probability,
retrieved and must be eliminated from the list of or possibility (Box 8-2).
possible sources. To determine which data sources During the analytic stage, the researcher begins
were used when you review a published study, the process of interpretation of meaning. Often
look at the reference list. Sources of data may be working with incomplete records, the historian
primary or secondary. Primary sources are eye- researcher is reaching beyond the evidence to
witness accounts provided by varying sorts of make inferences. The report usually contains
communication appropriate to the time. Second- extensive samples of the data, along with evi-
ary sources provide a view of the phenomenon dence of reliability and validity: A critical
from another person’s perspective rather than a description of historical evidence, an evaluation
first-hand account. of its historical significance to contemporary
Validity of documents is established by exter- society, and creative narratives are provided in
nal criticism reliability is established by internal the written research report, including the derived
criticism. In e ternal criticis the authenticity interferences (Lundy, 2011, p. 391).
of the data source is judged. The researcher seeks
to ensure that the data source is indeed what it Helpful Hint
seems to be. For instance, if the researcher When you critique a study based on the historical
is reviewing a handwritten letter of Florence method, expect not to find a report of data analysis but
Nightingale, some of the validity issues are simply a description of findings synthesized into a con-
tinuous narrative.
the following:
 CHAPTER 8 Qualitative Approaches to Research 187

BOX 8-2 PAR requires careful collaboration with the

ESTABLISHING FACT, PROBABILITY, AND research participants and focuses on practical
POSSIBILITY WITH THE HISTORICAL METHOD problems that are particular to a practice setting
FACT or community (Streubert Carpenter, 2011).
Two independent primary sources that agree with Stringer and Genat (2004) defined action research
each other as
or
One independent primary source that receives critical a systematic, participatory approach to inquiry that
evaluation and one independent secondary source enables people to extend their understanding of
that is in agreement and receives critical evaluation problems or issues and to formulate actions directed
and no substantive conflicting data towards the resolution of those problems or issues . . .
action research seeks local understandings that are
PROBABILITY
specifically relevant to the particular context of a
One primary source that receives critical evaluation study. (p. 4)
and no substantive conflicting data
or Co unity based artici atory research
Two primary sources that disagree about particular
points (CBPR) is a method by which the voice of a com-
munity is systematically accessed in order to plan
POSSIBILITY
context-appropriate action. CBPR provides an
One primary source that provides information but is
not adequate to receive critical evaluation alternative to traditional research approaches that
or assume a phenomenon may be separated from its
Only secondary or tertiary sources context for purposes of study. . . . CBPR recog-
Adapted from Christy, T. E. (1975). The methodology of historical nizes the importance of involving members of a
research: A brief introduction. Image, 24(3), 189-192. study population as active and equal participants,
in all phases of the research project, if the research
process is to be a means of facilitating change
(Holkup, Tripp-Reimer, Salois, Weinert, 2004,
Evidence-Informed Practice Tip p. 162). Change or action is the intended outcome
The presentation of a historical study should be of CBPR, and action research is a term related to
logical, consistent, and easy to follow.
CBPR. Some scholars would consider CBPR a
sort of action research and would group both
action research and CBPR within the tradition of
Participatory Action Research critical science (Fontana, 2004).
Based on orientational qualitative inquiry, ar
tici atory action research (PAR) is also a Evidence-Informed Practice Tip
method in which a goal is to change society. Although qualitative in its approach to research,
According to the tenets of PAR, all forms of CBPR leads to an action component in which a nursing
knowledge, including indigenous knowledge, are intervention is implemented and evaluated for its effec-
tiveness in a specific patient population.
of value and can be applied to practical problems.
The researcher studies a particular setting to iden-
tify areas in which improvements in practice are In his book Action Research, Stringer (1999)
needed (Glesne, 2011). After possible solutions distilled the research process into three phases:
are identified, action is taken to implement look, think, and act. Stringer defined the
changes in partnership with the stakeholders. look as building the picture by getting to
Careful attention is given to evaluating the process know stakeholders so that the problem is defined
to ensure that the changes have the desired effect. on their terms and the problem definition is
188 PART TWO Qualitative Research

re ective of the community context. The think the look phase, the researcher explores the
phase addresses interpretation and analysis of problem by asking who is involved, what is hap-
what was learned in the look phase the pening, and how, where and when events and
researcher is charged with connecting the ideas activities occur (p. 36). Re ecting on their
of the stakeholders so that they provide evidence observations, researchers, in collaboration with
that is understandable to the community group the stakeholders, can fine-tune the final research
(Stringer, 1999). Finally, in the act phase, question, which serves as a guide to the study. As
Stringer advocated for planning, implementing, mentioned earlier, Fournier and colleagues (2007)
and evaluating, on the basis of information col- explored the experiences of nurses from ganda
lected and interpreted in the other phases of who cared for people with human immunodefi-
research. ciency virus (HIV) infection and AIDS. As part
of the research project, they worked with the par-
Identifying the Phenomenon ticipants to define their issues, suggest solutions,
PAR evolved from the work of Lewin (1948), and act on them and then to re ect on the process
who viewed action research as a means for solving and outcomes.
practical social problems and for enacting change
for the improvement of communities. PAR is RESEARCHER’S PERSPECTIVE. When using PAR
heavily used as a research methodology in educa- methods, the researcher is no longer the expert
tion, and in the health professions, PAR methods but acts more as a consultant. In their case study
are used to improve health care services in com- of pregnant Aboriginal women and parenting
munities. PAR has been applied to health and Aboriginal families, in which participatory meth-
wellness programs, program evaluation, care odology was used, Smith and colleagues (2006)
plans, community nursing, and health care deliv- stated, participatory research views all forms of
ery and policy. Studies have ranged from issues knowledge as valuable (p. E21). The partici-
and conditions stemming from chronic illness, pants are co-researchers and are engaged in the
pregnancy and childbirth, pain management, and research process as it emerges. This involvement
incontinence to rehabilitation (Stringer Genat, requires processes that are democratic, participa-
2004). For example, Fournier, Mill, Kipp, and tory, empowering, and life-enhancing (Stringer
Walusimbi (2007) were interested in the experi- Genat, 2004). PAR investigators, like ethnogra-
ences of gandan nurses and their role in caring phers, immerse themselves in the field for deep
for people with acquired immune deficiency syn- understanding and to build trust and credibility.
drome (AIDS). Their study is used in the follow- For example, Fournier, a faculty lecturer from the
ing section to illustrate PAR. niversity of Alberta, spent 5 months during two
trips on site in ganda (Fournier et al., 2007).
Structuring the Study
RESEARCH QUESTION. The first step in structuring SAMPLE SELECTION. Because it is not possible to
the PAR study, as in other qualitative methods, is include everyone who may have a stake or
to frame the research question and to identify who interest in the research question, researchers pur-
is affected by or has an effect on the problem. posively select a sample of participants who rep-
Because of the emergent nature of PAR, research- resent varied perspectives, experiences, and
ers can begin with a tentative problem and ques- backgrounds. Participants may be people who
tions and then refine or reframe them as they enter have the widest range of differences in their expe-
the field. Recall the look-think-act cycle of riences, particularly interesting backgrounds or
Stringer and Genat (2004) described earlier. In experiences those who are typical and those
 CHAPTER 8 Qualitative Approaches to Research 189

with particular knowledge of the phenomenon of data analysis is to distill and reduce the volumes
under study. For example, Fournier and col- of information into a manageable and organized
leagues (2007) selected nurses who worked in a set of concepts or ideas. The process in PAR must
variety of health care units and regularly cared for directly capture the experiences of the partici-
people with AIDS. pants and be distilled in such a way that it makes
sense to them all (Stringer Genat, 2004).
Data Gathering Stringer and Genat (2004) identified two
In the look phase, data are gathered from a approaches to analysis. The first, based on epiph-
variety of sources interviews are the principal anic moments (Denzin, 1998), focuses on the
means for understanding the experiences of the significant experiences as the primary units of
participants. PAR also includes observation in the analysis, giving voice to the participants’ experi-
field, gathering and reviewing of relevant docu- ences. A second process involves the categoriza-
ments, and the examination of relevant materials tion and coding of data to reveal patterns and
and equipment. A literature review may add themes. Regardless of the process used, PAR
information to enhance the understanding of the allows the participants to make sense of their
data emerging from the interviews and other experience and then to use the new understanding
sources. Fournier and colleagues (2007) used the to make a positive change.
photovoice method in their work with nurses in In Fournier and colleagues’ (2007) study, the
ganda. Photovoice, developed by Caroline nurses participated fully in the analysis. Each
Wang and Maryann Burris is nurse reviewed the stories and developed themes
a participatory health promotion strategy in which
independently. A leader was selected to moderate
people use cameras to document their health and a group meeting to develop broader themes from
work realities. As participants engage in a group the input of each nurse. Fournier, the principal
process of critical reflection, they may advocate for investigator, spent an additional 2 months further
change in their communities by using the power of analyzing the data with the next two authors of
their images and stories to communicate with policy
the study.
makers. In public health initiatives from China to
California, community people have used photovoice
to carry out participatory needs assessment, conduct
Describing the Findings
participatory evaluation, and reach policy makers to In accordance with Stringer and Genat’s (2004)
improve community health. (Wang & Redwood-Jones, look-think-act framework, the next step is to
2001) present the outcomes to the participants and other
The participants took pictures of what they nonparticipant stakeholders so that they under-
considered to be nurses’ work. During photovoice stand what is happening. Several dissemination
meetings, the nurses shared stories about their mechanisms may be used because formal aca-
pictures. Viewing their work from behind the demic writing is not accessible for most lay par-
camera and sharing stories allowed the nurses to ticipants. The results may be shared in written
critically assess their work. They discovered that reports, oral presentations, or performances.
they were not working at their fullest potential Written narrative accounts and storytelling are
because the hospital administration did not value often used to describe the findings. The next and
their input or opinions. most important step is to apply the findings to
solve the research problem or issue that instigated
Data Analysis the study. This action portion of PAR parallels the
In the think phase, the researchers think about nursing process of identification of goals and
and re ect on all of the data gathered. The purpose objectives, intervention, and, finally, evaluation.
190 PART TWO Qualitative Research

The action plans should include the following disciplines, such as sociology, anthropology, and
(Stringer Genat, 2004): philosophy. In the discipline of nursing, these
hy: A statement of the overall purpose methodologies are used to conduct research.
hat: A set of objectives to be obtained However, as the discipline matures, methodology
o : A sequence of tasks and steps for each based on nursing ontology (belief system)
objective emerges. Madeleine Leininger (1996), Rosemarie
ho: The people responsible for each task and Rizzo Parse (1997), and Margaret Newman
activity (1997) are nurse theorists who have created
here: The place where the tasks will be done research methods specific to their theories. In
hen: The time for initiation and completion Table 8-1, the methodologies of these theorists
The researcher should also arrange for ongoing are compared. Each method was developed over
evaluation of the process. As with the exploratory years and tested by other researchers. Each
phase, stakeholders are intimately involved in researcher has attempted to advance nursing
each step of the action plan, from identifying the knowledge through inquiry that is congruent with
plan to implementing it. The participants in the specific nursing theory.
Fournier and colleagues’ (2007) study indicated In this section of the chapter, we have explored
that one of their issues was inadequate knowledge several different traditions and methods of quali-
about antiviral medications. This finding resulted tative research however, many researchers are
in an unexpected outcome: the researchers deter- now combining a number of related or different
mined that they would develop a research pro- methods to frame their naturalistic study. Remem-
posal. These nurses, who formed supportive and ber that qualitative inquiry is evolving and chang-
strong bonds, began to take leadership roles in the ing. As Glesne (2011) noted, The open, emergent
hospital, particularly in addressing ethical issues. nature of qualitative inquiry means a lack of stan-
dardization there are no clear criteria to package
QUALITATIVE APPROACH: into neat research steps (p. 25). As you read
NURSING METHODOLOGY more qualitative research studies, you will note
The qualitative methodologies elaborated many interesting designs, all amenable to explor-
throughout this chapter are derived from other ing the complexity of the human experience.

TABLE 8-1
NURSING RESEARCH METHODOLOGIES
ASPECT LEININGER (1996) PARSE (1997) NEWMAN (1997)

Theory Culture care Human becoming Health as expanding

consciousness
Research methodology Ethnonursing is centred on Parse’s research methodology Newman’s method focuses
learning from people is the study of universal on pattern recognition and
about their beliefs, health experiences through uses multiple interviews,
experiences, and culture true presence both with involving collaboration, to
care information participants sharing life arrive at recognized life
stories and with transcribed patterns
data to uncover meaning
Research example Use of culture care theory The lived experience of Pattern of expanding
with Anglo-American and serenity: using Parse’s consciousness in women
African American older research method (Kruse, in creative movement and
persons in a long-term care 1999) narrative as modes of
setting (McFarland, 1997) expression (Picard, 2000)
 CHAPTER 8 Qualitative Approaches to Research 191

Smith and colleagues’ (2006) study on pregnant In summary, the term ualitative research is
Aboriginal women and parenting Aboriginal fam- an overriding description of multiple methods
ilies, for example, combines a postcolonial stand- with distinct origins and procedures. In spite of
point, participatory research principles, and a distinctions, each method shares a common nature
case study design. Ford-Gilboe and associates that guides data collection from the perspective
(2005) applied a feminist perspective to their of the participants to create a story that synthe-
grounded theory on the basic social processes of sizes disparate pieces of data into a comprehen-
health promotion among single-parent families sible whole that provides evidence and promises
recovering from intimate family violence. direction for building nursing knowledge.

APPRAISING THE EVIDENCE

Qualitative Research
Although general criteria for critiquing qualitative research qualitative research is a useful activity for learning the
are proposed in the following Critiquing Criteria box, each nuances of this research approach. You are encouraged
qualitative method has unique characteristics that influ- to identify a qualitative study of interest and apply the
ence what you may expect in the published research criteria for critiquing. Keep in mind that qualitative
report, and journals often have page restrictions that methods are the best way to start answering clinical and
penalize qualitative researchers because it is difficult at research questions that previously have not been
times to fully explain all of the steps in Chapter 4 in a few addressed in research studies or that do not lend them-
pages. The criteria for critiquing are formatted to evaluate selves to a quantitative approach. The answers provided
the selection of the phenomenon, the structure of the by qualitative data reflect important evidence that may
study, data gathering, data analysis, and description of provide the first insights into a patient population or clini-
the findings. Each question of the criteria focuses on cal phenomenon.
factors discussed throughout the chapter. Appraising

CRITIQUING CRITERIA

Qualitative Approaches Researcher’s Perspective DATA ANALYSIS

6. Are the biases of the 11. Can the dimensions of data
IDENTIFYING THE PHENOMENON researchers reported? analysis be identified and
1. Is the phenomenon focused on 7. Do the researchers provide a logically followed?
human experience within a structure of ideas that reflect 12. Is the participant’s reality clearly
natural setting? their beliefs? described?
2. Is the phenomenon relevant to 13. Is there evidence that the
nursing or health, or both? Sample Selection researcher’s interpretation
8. Is it clear that the selected captured the participant’s
STRUCTURING THE STUDY sample is experiencing meaning?
Research Question the phenomenon of 14. Have other professionals
3. Does the question specify interest? confirmed the researcher’s
a distinct process to be interpretation?
studied? DATA GATHERING
4. Does the question identify the 9. Are data sources and DESCRIBING THE FINDINGS
context (participant group/ methods for gathering data 15. Are examples provided to guide
place) of the process that will specified? the reader from the raw data to
be studied? 10. Is there evidence that the researcher’s synthesis?
5. Does the choice of a specific participant consent is an 16. Does the researcher link the
qualitative method fit with the integral part of the data- findings to existing theory or
research question? gathering process? literature, or is a new theory
generated?
192 PART TWO Qualitative Research

REFERENCES
CRITICAL THINKING CHALLENGES
Aamodt, A. A. (1991). Ethnography and epistemology:
■ How does the researcher select a specific type of Generating nursing knowledge. In J. M. Morse (Ed.),
qualitative research method to answer the ualitative nursing research: A contemporary
research question? dialogue (pp. 29-40). Newbury Park, CA: Sage.
■ Do findings from qualitative research studies Agar, M. H. (1986). Spea ing of ethnography. Beverly
Hills, CA: Sage.
need to be validated in subsequent studies?
Aita, V. A., McIlvain, H. E. (1999). An armchair
■ How can a nurse researcher select a qualitative adventure in case study research. In B. Crabtree
research method when he or she is attempting W. L. Miller (Eds.), Doing ualitative research (2nd
to accumulate evidence regarding a new topic ed., pp. 253-268). Thousand Oaks, CA: Sage.
about which little is known? Boyle, J. S. (1991). Field research: A collaborative
■ How can the case study approach to research be model for practice and research. In J. M. Morse (Ed.),
applied to evidence-informed practice? ualitative nursing research: A contemporary
dialogue. Newbury Park, CA: Sage.
Browne, A. J. (2000). The potential contribution of
critical social theory to nursing science. Canadian
ournal of Nursing Research, (2), 35-55.
Charmaz, K. (2000). Grounded theory: Objectivist and
KEY POINTS constructivist methods. In N. K. Denzin Y. S.
Lincoln (Eds.), andboo of ualitative research
• Qualitative research is the investigation of human
(2nd ed., pp. 509-535). Thousand Oaks, CA: Sage.
experiences in naturalistic settings, pursuing
meanings that inform theory, practice, instrument Christy, T. E. (1975). The methodology of historical
development, and further research. research: A brief introduction. Nursing Research,
• Qualitative research studies are guided by research (3), 189-192.
questions. Christy, T. E. (1981). The need for historical research in
• Data saturation occurs when the information being nursing. Research in Nursing ealth, (2),
shared with the researcher becomes repetitive. 227-228.
• Qualitative research methods include five basic Cohen, M. . (1987). A historical overview of the
elements: identifying the phenomenon, structuring phenomenological movement. Imae, (1), 31-34.
the study, gathering the data, analyzing the data, Cohen, M. ., Saunders, J. (1996). sing qualitative
and describing the findings.
research in advanced practice. Advanced Practice
• The phenomenological method is a process of
learning and constructing the meaning of human Nursing uarterly, (3), 8-13.
experience through intensive dialogue with persons Colaizzi, P. (1978). Psychological research as a
who are living the experience. phenomenologist views it. In R. S. Valle M. King
• The grounded theory method is an inductive (Eds.), E istential phenomenological alternatives for
approach that implements a systematic set of psychology (pp. 48-71). New York: Oxford niversity
procedures to arrive at theory about basic social Press.
processes. Corbin, J., Strauss, A. (2008). Basics of ualitative
• The ethnographic method focuses on scientific research. Los Angeles: Sage.
descriptions of cultural groups. Crabtree, B. F., Miller, W. L. (1992). Doing ualita-
• The case study method focuses on a selected
tive research. Newbury Park, CA: Sage.
phenomenon over a short or long time to provide
an in-depth description of its essential dimensions Creswell, J. W. (1998). ualitative in uiry and research
and processes. design: Choosing among ve traditions. Thousand
• The historical research method is the systematic Oaks, CA: Sage.
compilation of data and the critical presentation, Denzin, N. K. (1998). The practices and politics of
appraisal, and interpretation of facts regarding interpretation. In N. K. Denzin Y. S. Lincoln
people, events, and occurrences of the past. (Eds.), Collecting and interpreting ualitative
• CBPR is a method that systematically accesses the materials (pp. 458-498). Thousand Oaks, CA: Sage.
voice of a community to plan context-appropriate Denzin, N. K., Lincoln, Y. S. (1998). The landscape
action. of ualitative research. Thousand Oaks, CA: Sage.
 CHAPTER 8 Qualitative Approaches to Research 193

de Witt, L., Ploeg, J., Black, M. (2010). Living along Kruse, B.G. (1999). The lived experience of serenity:
with dementia: An interpretative phenomenological sing Parse’s research method. Nursing Science
study with older women. ournal of Advanced uarterly, , 143-150.
Nursing, (8), 1698-1707. doi: Lapum, J., Angus, J. E., Peter, E., Watt-Watson, J.
10.1111/j.1365-2648.2010.05295.x. (2010). Patients’ narrative accounts of open-heart
Duffy, M. (2011). Narrative inquiry: The method. In surgery and recovery: Authorial voice of technology.
P. Munhall (Ed.), Nursing research: A ualitative Social Science Medicine, , 754-762. doi:
perspective (5th ed., pp. 421-440). Mississauga, ON: 10.1016/j.socsimed.2009.11.021
Jones Bartlett Learning. Leininger, M. (1985). Life-health-care history:
Fitzpatrick, J. J., Wallace, M. (Eds.). (2006). Purposes, methods and techniques. In M. M.
Encyclopedia of nursing research (2nd ed.). New Leininger (Ed.), ualitative research methods
York: Springer. in nursing (pp. 119-132). New York: Grune
Fontana, J. S. (2004). A methodology for critical science Stratton.
in nursing. Advances in Nursing Science, , 93-101. Leininger, M. M. (1996). Culture care theory. Nursing
Ford-Gilboe, M., Wuest, J., Merritt-Gray, M. (2005). Science uarterly, , 71-78.
Strengthening capacity to limit intrusion: Theorizing Leininger, M. M. (2006). Culture care, diversity and
family health promotion in the aftermath of woman universality: A orld ide nursing theory (2nd ed).
abuse. ualitative ealth Research, , 477-601. Mississauga, ON: Jones Bartlett.
Fournier, B., Mill, J., Kipp, W., Walusimbi, M. Lewin, K. (1948). Resolving social con icts. New York:
(2007). Discovering voice: A participatory action Harper.
research study with nurses in ganda. International Lindsay, G. (2006). Constructing a nursing identity:
ournal of ualitative Methods, (2). Retrieved from Re ecting on and reconstructing experience.
http://www.ualberta.ca/ iiqm/backissues/6 2/fournier. Re ective Practice, , 59-72.
htm. Lundy, K. S. (2011). Historical research. In P. Munhall
Giorgi, A., Fischer, C. L., Murray, E. L. (Eds.). (Ed.), Nursing research: A ualitative perspective
(1975). Du uesne studies in phenomenological (5th ed., pp. 381-397). Mississauga, ON: Jones
psychology. Pittsburgh: Duquesne niversity Press. Bartlett Learning.
Glaser, B., Strauss, A. (1965). A areness of dying. Lusk, B. (1997). Historical methodology for nursing
Chicago: Aldine de Gruyter. research. Image, , 355-359.
Glaser, B. G., Strauss, A. L. (1967). The discovery of Macquire, P. (1996). Considering more feminine
grounded theory: Strategies for ualitative research. participatory research: What’s congruency got to do
Chicago: Aldine. with it ualitative In uiry, , 106-118.
Glesne, C. (2011). Becoming ualitative researchers: An Maddalena, V. J., Bernard, W. T., Etowa, J., Murdoch,
introduction (4th ed.). Toronto: Pearson. S. D., Smith, D., Jarvis, P. M. (2010). Cancer care
Holkup, P. A., Tripp-Reimer, T., Salois, E. M., experiences and the use of complementary and
Weinert, C. (2004). Community-based participatory alternative medicine at the end of life in Nova
research: An approach to intervention research with a Scotia’s black communities. ournal of Transcultural
Native American community. Advances in Nursing Nursing, (2), 114-122. doi: 10.1177/
Science, , 162-175. 1043659609357634
Holmes, D., O’Byrne, P., Gastaldo, D. (2006). Raw McDonald, C. (2006) Lesbian disclosure: Disrupting the
pleasure as limit experience: A Foucauldian analysis taken for granted. Canadian ournal of Nursing
of unsafe anal sex between men. Social Theory and Research, , 42-57.
ealth, , 319-333. McFarland, M.R. (1997). se of culture care theory
Holmes, D., O’Byrne, P., Gastaldo, D. (2007). Setting with Anglo and African American elders in long-term
the space for sex: Architecture, desire and health care settings. Nursing Science uarterly, ,
issues in gay bathhouses. International ournal of 186-192.
Nursing Studies, (2), 273-284. Mead, M. (1949). Coming of age in Samoa. New York:
Kleinman, A. (1980). Patients and healers in the conte t New American Library/Mentor Books. (Original work
of culture. Berkeley: niversity of California Press. published 1928).
Kleinman, A. (1992). Local worlds of suffering: An Newman, M. A. (1997). Evolution of the theory of
interpersonal focus for ethnographies of illness health as expanding consciousness. Nursing Science
experience. ualitative ealth Research, , 127-134. uarterly, , 22-25.
194 PART TWO Qualitative Research

Nightingale, F. (1969). Notes on nursing: hat it is and parenting Aboriginal people. Advances in Nursing
hat it is not. New York: Dover. (Original work Science, (2), E27-E44.
published 1858). Spiegelberg, H. (1976). The phenomenological move-
Parse, R. R. (1997). Transforming research and practice ment (Vols. I and II). The Hague: Martinus Nijhoff.
with the human becoming theory. Nursing Science Spradley, J. P. (1979). The ethnographic intervie . New
uarterly, , 171-174. York: Holt, Rinehart, Winston.
Parse, R. R., Coyne, A. B., Smith, M. J. (1985). Stake, R. E. (1995). The art of case study research.
Nursing research: ualitative and uantitative Thousand Oaks, CA: Sage.
methods. Bowie, MD: Brady. Stake, R. E. (2003). Case studies. In N. K. Denzin
Parsons, S. (1911). The case method of teaching Y. S. Lincoln (Eds.), Strategies of ualitative in uiry
nursing. American ournal of Nursing, , (2nd ed., pp. 134-164). Thousand Oaks, CA: Sage.
1009-1011. Strauss, A., Corbin, J. (1990). Basics of ualitative
Patton, M. (2002). ualitative research evaluation research: rounded theory procedures and tech-
methods (3rd ed.). Thousand Oaks, CA: Sage. ni ues. Newbury Park, CA: Sage.
Pesut, B., Reimer-Kirkham, S. (2010). Situated Strauss, A. L. (1987). ualitative analysis for social
clinical encounters in the negotiations of religious scientists. New York: Cambridge niversity Press.
and spiritual plurality: A critical ethnography. Streubert, H. J., Carpenter, D. R. (2011). ualitative
International ournal of Nursing Studies, , research in nursing: Advancing the humanistic
815-825. doi: 10.1016/j.ijnurstu.2009.11.014. imperative (5th ed.). Philadelphia: Wolters Kluwer.
Picard, C. (2000). Patterns of expanding consciousness Stringer, E. T. (1999). Action research (2nd ed.).
in midlife women: Creative movement and narrative Thousand Oaks, CA: Sage.
as modes of expression. Nursing Science uarterly, Stringer, E. T., Genat, W. J. (2004). Action research
, 150-157. in health. Thousand Oaks, CA: Sage.
Reeves, S., Macmillan, K., van Soeren, M. (2010). sher, P. (1996). Feminist approaches to research. In
Leadership of interprofessional health and social care D. Scott R. sher (Eds.), nderstanding educa-
teams: A socio-historical analysis. ournal of Nursing tional research (pp. 120-143). New York: Routledge.
Management, , 258-264. doi: 10.1111/j.1365- van Kaam, A. (1969). E istential foundations in
2834.2010.01077.x psychology. New York: Doubleday.
Roberts, T. (2009). nderstanding ethnography. British van Manen, M. (1997). Researching lived e perience:
ournal of Mid ifery, (5), 291-294. uman science for an action sensitive pedagogy (2nd
Roper, J., Shapira, J. (2000). Ethnography in nursing ed.). London, ON: Althouse Press.
research. Thousand Oaks, CA: Sage. Wang, C. Redwood-Jones, Y. (2001). Photovoice
Sandelowski, M. (2004). sing qualitative research. Ethics: Perspective from Flint Photovoice. Retrieved
ualitative ealth Research, , 1366-1386. from http://www.photovoice.org. ealth Education
Sandelowski, M., Barroso, J. (2007). andboo for Behaviour, (5), 560-572 doi: 10.1177/
synthesi ing ualitative research. Philadelphia: 109019810102800504
Springer. Wuest, J. (2011). Grounded theory: The method. In P.
Schrieber, R., MacDonald, M. (2010). Keeping vigil Munhall (Ed.), Nursing research: A ualitative
over the patient: A grounded theory of nurse perspective (5th ed., pp. 225-256). Mississauga, ON:
anaethesia practice. ournal of Advanced Nursing, Jones Bartlett Learning.
(3), 552-561. doi: 10.1111/j.1365-2648.2009. Yin, R. (1994). Case study research: Design and
05207.x methods (2nd ed.). Thousand Oaks, CA: Sage.
Scotch, N. (1963). Medical anthropology. Biennial err, S. J., ilm, G., Grant, V. (2006). abor of love
Revie of Anthropology, , 30-68. A memoir of ertrude Richards adner, to
Seneviratne, C. C., Mather, C. M., Then, K. L. . Delta, BC: G .
(2009). nderstanding nursing on an acute stroke
unit: Perceptions of space, time and interprofessional
practice. ournal of Advanced Nursing, (9),
FOR FURTHER STUDY
1872-1881. doi: 10.1111/j.1365-2648.2009. Go to Evolve at http://evolve.elsevier.com/
05053.x Canada/LoBiondo/Research for Audio Glossary, how-to
Smith, D., Edwards, N., Varcoe, C., Martens, P. J., instructions for Writing Proposals for Funding, and
Davies, B. (2006). Bringing safety and responsive- additional research articles for practice in reviewing
ness into the forefront of care for pregnant and and critiquing.
195

PA RT T HR E E
PART ONE Part Title

Quantitative Research

9 Introduction to
Quantitative Research
10 Experimental and
Quasiexperimental
Designs
11 Nonexperimental Designs
RESEARCH VIGNETTE Several qualitative and quanti-
tative studies ensued. Older adults
Tackling the Prevention of Falls described the embarrassment of
using grab bars and other assistive
Among Older Adults devices. A survey of older adults
in apartment buildings with and
without universally installed
Nancy C. Edwards, RN, BScN, MSc, PhD contribution of self-care versus grab bars identified barriers to
Full Professor collective action initiatives to access of grab bars. A comparison
Department of Epidemiology and
promote health. However, we of low-income and public apart-
Community Medicine
School of Nursing decided to meld these two inter- ment buildings with private apart-
University of Ottawa ests, and a randomized controlled ment buildings indicated that older
Ottawa, Ontario trial was born adults living in low-income
With funding from the Ontario buildings were much more likely
In the late 1980s, while working Ministry of Health and Long-Term to have access to grab bars than
as a clinical nurse consultant at Care, we compared the effect of were older adults living in private
the Ottawa Public Health Depart- fall prevention clinics and a com- buildings.
ment, I was asked to assist with a munity action strategy on the inci- Concerns about bathroom grab
needs assessment of older adults dence of falls among older adults bars and safe stairs became a
living in low-income apartment living in 48 apartment buildings. I focus for our regional fall preven-
buildings in Ottawa. The survey worked closely with the fall pre- tion coalition. We began to con-
covered a wide range of topics, vention team as we piloted and sider what research was necessary
and at the last minute, the public implemented the two interven- to inform changes to building
health nurses decided to add a tions. Although we did not detect a codes in Canada. It became appar-
few questions about older adults’ change in the rate of falls, we did ent that, to complement our com-
experiences with falls. The find- note improvements in some behav- munity-based studies, we needed
ings were startling: They revealed ioural outcomes for fall prevention the expertise of researchers who
a high incidence of falls and inju- among older adults in the commu- were able to design and conduct
ries. Published epidemiological nity action buildings in compari- laboratory studies to determine
studies indicated that our findings son with the control buildings. what grab bar configurations are
were not spurious. Indeed, one Many important issues sur- optimal to assist transfers into and
third of all older adults fall annu- faced during this trial and have out of a bathtub.
ally, and approximately 25 of become the basis of nearly a From both the laboratory and
falls result in some sort of injury. decade’s worth of research. In par- the community studies, we were
Soon after this survey was con- ticular, we became interested in able to identify the configuration
ducted, our then medical officer of the role of the built environment of grab bars that older adults find
health, Dr. Steve Corber, spoke to in falls. In some of our study easiest to use. We determined that
me about the possibility of setting buildings, older adults who had older adults are 2.8 times more
up and evaluating a fall preven- difficulties getting into and out of likely to use universally installed
tion initiative. Dr. Corber sug- the bathtub described their failed grab bars consistently, in compari-
gested that preventing falls should efforts to persuade landlords to son with grab bars they had
become a program focus within have grab bars installed. Land- installed themselves, after other
the health department. I was work- lords were refusing them permis- factors were adjusted. We also
ing with a senior nursing manager, sion because the installation of documented both the high propor-
Maureen Murphy, at the time, grab bars was thought to lower tion of falls in bathrooms and on
and we were interested in a more property values. In essence, aes- stairs and the high rates of injury
theoretical question: the relative thetics trumped safety. that result from these falls, in

196
relation to falls occurring in other building codes for the universal practitioners and managers, poli-
locations. installation of grab bars in showers cymakers, and colleagues from
In a subsequent community and bathtubs in all residential the volunteer sector to address
study funded by the Canadian In- homes. Included with the sub- common issues. Participants con-
stitutes of Health Research, we mission was an estimate of the sider how evidence might inform
identified stair hazards prevalent costs of such a requirement and intersectoral practices and policies
in private homes and public build- the cost savings that would be and how activities in the field
ings. Older adults were asked incurred. These estimates are an and in policy arenas might inform
to identify the most common important consideration for com- research. We held a series of fire-
locations of hazardous stairs. mittee members who review code side chats on fall prevention and
Churches and community centres change requests. I have also been related changes to the building
were among the locations most a member of several National codes. An outcome of these chats
frequently identified. Independent Research Council task groups was a series of resolutions on
raters also assessed indoor and and subgroups reviewing potential evidence-informed changes to
outdoor stairs, both those identi- code changes for stairs, handrails, building codes that pertain to
fied by older adults as hazardous ramps, and guards. Our research bathtub grab bars and safe stairs.
in the community and the stairs on stair falls is being reviewed as These resolutions were submitted
that the raters used in their own part of the evidence for changes to for the consideration of various
homes or apartment buildings. the building codes in the current professional associations.
The most common hazards were cycle of building code changes. Our work on preventing falls
the lack of contrast marking on Bridging the interface between has reinforced the importance
the edge of the stairs, inadequate research and policy is a critical of the role of interdisciplinary
tread length, risers that were too knowledge translation strategy. research in the design of com-
high, and nonuniform risers. Al- Informing and in uencing policy munity health interventions. Sub-
though older adults were able to change require the engagement of stantial health care changes in
identify hazardous stairs they had an informed public, which led us populations such as older adults
difficulty navigating, fewer than to establish a networking infra- will come about only through
25 of them had specific sugges- structure called CHNET-Works long-term research and policy
tions for ways to improve stair (http://www.chnet-works.ca). change efforts. uick fixes are
safety. sing teleconference lines, con- rare. As a researcher, I need to be
Our research findings on bath- ferencing software (Bridgit), and prepared to work actively with
room and stair falls are now being the support of a network anima- colleagues at the local, provincial,
used to inform policy change. For teur, we host weekly fireside and national levels and to use a
example, I submitted a request to chats for 9 months of the year. wide repertoire of knowledge
the National Research Council for Our aim is to bring together translation strategies. ■
an additional requirement to the academic researchers, front-line

197
C H A PTER 9

Introduction to Quantitative
Research
Geri LoBiondo-Wood | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Define research design.
• Identify the purpose of the research design.
• Define control as it affects the research design.
• Compare and contrast the elements that affect control.
• Begin to evaluate the degree of control that should be exercised in the design.
• Define internal validity.
• Identify the threats to internal validity.
• Define external validity.
• Identify the conditions that affect external validity.
• Identify the links between study design and evidence-informed practice.
• Evaluate the design by using the critiquing questions.

KEY TERMS
accuracy feasibility mortality
attrition Hawthorne effect objectivity
bias history threat pilot study
constancy homogeneity randomization
control instrumentation threats reactivity
control group internal validity selection bias
experimental group maturation selection effects
external validity measurement effects testing effect
extraneous variable

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

198
 CHAPTER 9 Introduction to Quantitative Research 199

THE WORD DESIGN IMPLIES THE ORGANIZATION of by the findings and their potential applicability to
elements into a masterful work of art. In the world practice.
of art and fashion, the word conjures up images of Nursing practice is concerned with a variety of
processes and techniques that are used to express activities that require varying degrees of process
a total concept. When an individual creates, and form, such as the provision of quality care,
process and form are employed. The form, process, cost-effective patient care, responses of patients
and degree of adherence to structure depend on the to disease, and factors that affect caregivers.
aims of the creator. When nurses administer patient care, they draw
The same can be said of the research process. on the nursing process. Previous chapters stressed
The research process does not need to be a sterile the importance of theory and knowledge of
procedure, but it should be one in which the subject matter to research. How a researcher
researcher develops a masterful work within the structures, implements, or designs a study affects
limits of a problem and the related theoretical the results of a research project.
basis. The framework that the researcher creates To grasp the implications and the use of re-
is the design. When reading a study, the research search, you need to understand the central issues
consumer should be able to recognize that the in the design of a research project. This chapter
research problem, purpose, literature review, the- provides an overview of the meaning, purpose,
oretical framework, and hypothesis all interrelate and issues related to quantitative research design.
with, complement, and assist in the operational- Chapters 10 and 11 discuss specific types of
ization of the design (Figure 9-1). The degree to quantitative designs.
which a fit exists between these design elements
determines the strength of the study and of the
consumer’s confidence in the evidence provided PURPOSE OF THE RESEARCH DESIGN
The purpose of the research design is to provide
the plan for answering research questions. These
questions can result in research driven by a
researcher’s curiosity or interest in a theoretical
Hypothesis question. This process is called basic research,
and its motivation is to expand nursing knowl-
edge. In contrast, applied research is designed to
solve clinical problems rather than to acquire
Design Theoretical knowledge for knowledge’s sake thus, the goal
framework
is to improve the patient’s health care condition.
The design in quantitative research then
becomes the vehicle for hypothesis testing or
answering research questions, whether they are
basic or applied. The design involves a plan, a
Problem Literature structure, and a strategy These three design con-
statement review cepts guide a researcher in writing the hypothesis
or research questions, conducting the project, and
analyzing and evaluating the data. The overall
purpose of the research design is twofold: to aid
FIGURE 9-1 Interrelationships of design, problem state-
ment, literature review, theoretical framework, and in the solution of research problems and to main-
hypothesis. tain control (see Practical Application box). All
200 PART THREE Quantitative Research

research is an attempt to answer questions. The The type of design used in a study also affects
design, coupled with the methods and analysis, is its application to practice. Chapters 10 and 11
the mechanism for finding solutions to research present a number of experimental, quasiexperi-
questions. Control is defined as the measures that mental, and nonexperimental designs. The type of
the researcher uses to hold the conditions of the design used in a study is linked to the level of
study uniform and avoid possible impingement of evidence, and, in turn, the contribution of a
bias (distortion of the results) on the dependent study’s findings is linked to evidence-informed
variable or outcome. practice. As discussed in Chapter 1, the term
evidence-informed practice is currently being
used instead of evidence-based practice because
Practical Application it is more inclusive in that it encompasses many
A research example that demonstrates how forms of evidence such as clinical experience and
the design can aid in answering a research judgement with research utilization. As you criti-
question and maintain control is the study by Ireland
and colleagues (2010). The main purpose of their cally appraise the design, take into account other
study was to examine the extent to which selected aspects of a study’s design, which are reviewed
demographic, social-psychological, physiological, in this chapter.
and adherence characteristics were predictive of
achievement of blood pressure and glucose targets in
a group of patients referred to a stroke prevention
OBJECTIVITY IN THE
clinic with either confirmed transient ischemic attack CONCEPTUALIZATION OF THE PROBLEM
(TIA) or confirmed stroke and with hypertension or In the conceptualization of the problem, objectiv-
diabetes or both. To maintain control, the researchers
had strict sample characteristics. Inclusion criteria ity is derived from a review of the literature and
were as follows: (1) TIA or stroke plus a documented development of a theoretical framework (see
history of hypertension, diabetes, or both; (2) age Figure 9-1). sing the literature, the researcher
older than 18 years; (3) ability to speak English; and
(4) ability to independently provide admission assesses the depth and breadth of available knowl-
information to a clinic registration clerk. Exclusion edge about the problem. The literature review
criteria were as follows: severe hearing or visual and theoretical framework should show that the
impairment, severe aphasia, and confusion. By
establishing the specific sample criteria and
researcher reviewed the literature critically and
participant eligibility, the researchers were able to objectively (see Chapters 2 and 5) because this
maintain control over the study’s conditions and conceptualization of the problem affects the type
suggest an extension of the study’s outcome with of design chosen. For example, for a question
further research.
about the relationship of the length of a breast-
feeding education program, either an experimen-
Various considerations, including the type of tal or a correlational design may be recommended
design, affect the accomplishment of the study. (see Chapters 10 and 11), whereas for a question
These considerations include ob ectivity—the regarding the physical changes in a woman’s
use of facts without distortion by personal feel- body during pregnancy and the maternal percep-
ings or bias—in the conceptualization of the tion of the unborn child, a survey or correlation
problem accuracy feasibility control of the study may be advised (see Chapter 11). The lit-
experiment internal validity and external valid- erature review should re ect the following:
ity. Statistical principles underlie the many forms When the problem was studied
of control, but it is more important that the The aspects of the problem that were studied
research consumer have a clear conceptual under- Where the problem was investigated
standing of statistics and how they inform the By whom the problem was investigated
research questions. The gaps or inconsistencies in the literature
 CHAPTER 9 Introduction to Quantitative Research 201

successfully carried out. Sometimes, the reality

Helpful Hint of feasibility does not truly sink in until the
A review that incorporates the aspects presented researcher begins the study. When you review a
here allows the consumer to judge the objectivity of the
problem area and therefore whether the design chosen study, you should consider feasibility, including
is suitable for investigating the problem. availability of the participants, timing of the
research, time required for the participants to take
part in the study, costs, and analysis of the data
ACCURACY (Table 9-1). Studies in which researchers are
Accuracy in determining the appropriate design testing feasibility are also called pilot studies (see
is also accomplished through the theoretical Practical Application box).
framework and review of the literature (see Chap-
ters 2 and 5). Accuracy means that all aspects of
a study systematically and logically follow from Practical Application
the research problem. The beginning researcher Steenbeek and colleagues (2009) conducted
is wise to answer a question involving few vari- a descriptive pilot study to determine
ables that does not require the use of sophisticated whether using an interactive journal club would
enhance the acquisition, appraisal, and application of
designs. The simplicity of a research project does nursing research knowledge. If results were positive,
not render it useless or of a lesser value for prac- then the use of this journal could be integrated into
tice. However, although the project is simple, the the nursing curriculum. The results were that the
researcher should not forgo accuracy. The con- nursing students found the journal club sessions
enjoyable and beneficial for helping them interpret
sumer should believe that the researcher chose a research findings and apply research to clinical
design that was consistent with the research practice.
problem and offered the maximum amount of
control.
Many clinical problems have not yet been An example of a feasibility study is that con-
researched, so a preliminary, or pilot, study is a ducted by McGilton and associates (2011), who
wise approach to testing the accuracy of a study examined the feasibility of a patient care com-
design before a larger study is undertaken. A ilot munication intervention. Before conducting a
study is a small, simple study conducted as a larger multicentre study, they wanted to know
prelude to a larger study. The key is the accuracy, more about how staff perceptions of the interven-
validity, and objectivity used by the researcher in tion affected the implementation of the interven-
attempting to answer the question. Accordingly, tion. In addition, the researchers explored changes
you should read various types of research reports in patients’ perceptions of care and psychosocial
and assess whether and how the criteria for each functioning and explored changes in nurses’
step of the research process were followed. Many knowledge of and attitude toward communication
nursing journals publish not only sophisticated with patients. The results indicated that the nurses
clinical research projects but also smaller clinical adhered to the intervention and the response rate
studies whose results can be applied to practice. was high.
Before a large experimental study (such as a
FEASIBILITY randomized clinical trial) is conducted, it is helpful
When you, as a consumer of research, critique to first conduct a pilot study with a small number
the study design, you must also be aware of of participants to determine the feasibility of par-
the pragmatic consideration of feasibility. easi ticipant recruitment, the intervention, the data-
bility is the capability of the study to be collection protocol, the likelihood that participants
202 PART THREE Quantitative Research

TABLE 9-1
PRAGMATIC CONSIDERATIONS IN DETERMINING THE FEASIBILITY OF A RESEARCH PROBLEM
FACTOR PRAGMATIC CONSIDERATION

Time The research problem must be able to be studied within a realistic period of time. All
researchers have deadlines for completion of a project. The scope of the problem
must be circumscribed enough to provide ample time for the completion of the
entire project. Research studies generally take longer than anticipated to complete.
Participant availability The researcher must determine whether a sufficient number of eligible participants
will be available and willing to take part in the study. If a researcher has a
“captive” audience (e.g., students in a classroom), it may be relatively easy to
enlist their cooperation. When a study involves the participants’ independent time
and effort, they may be unwilling to participate when they will receive no
apparent reward for doing so. Other potential participants may have fears about
harm or confidentiality and be suspicious of the research process in general.
Participants with unusual characteristics, such as rare diseases, are often difficult to
locate. People are generally cooperative about taking part in a study, but a
researcher must consider needing a larger participant pool than will actually
participate. At times, when reading a research report, the researcher may note
how the procedures were liberalized or the number of participants was altered—
probably as a result of some unforeseen pragmatic consideration.
Facility and equipment availability All research projects require some kind of equipment, such as questionnaires,
telephones, stationery, stamps, technical equipment, or another apparatus. Most
research projects also require the availability of a facility for the work, such as a
hospital site for data collection, a laboratory space, or a computer centre for data
analysis.
Money Many research projects require some expenditure of money. Before embarking on a
study, the researcher probably itemized the expenses and estimated the total cost
of the project. This estimation of cost provides a clear picture of the budgetary
needs for items such as books, stationery, postage, printing, technical equipment,
telephone and computer charges, and salaries. These expenses can range from
about $200 for a small-scale student project to hundreds of thousands of dollars
for a large-scale federally funded project.
Researcher experience The selection of the research problem should be based on the nurse’s experience and
interest. It is much easier to develop a research study related to a topic that is
either theoretically or experientially familiar. Selecting a problem that is of interest
to the researcher is essential for maintaining enthusiasm when the inevitable
successes and failures occur.
Ethics Research problems that place unethical demands on participants are not feasible for
study. Researchers must take ethical considerations seriously. The consideration of
ethics may affect the choice of the design and the methodology.

will complete the study, the reliability and validity feasibility of data-collection measures, and iden-
of new measurement tools, and the costs of the tify issues related to recruitment and follow-up.
study. Such a pilot study was conducted by Hayward and colleagues made revisions to the
Hayward and colleagues (2007), who investigated inclusion criteria: that is, they used infants’ gesta-
the feasibility of data collection (on the effects of tional age rather than weight. They found that
co-bedding twins on parental self-efficacy) before blinding the data-collection procedure would
beginning a larger multicentre study. The results strengthen the larger study. Redundancy in data
were used to estimate effect size and the organiza- collection was removed. The rate of response to the
tion of staff and bedside care, evaluate the parental questionnaire was poor at 1 month this
 CHAPTER 9 Introduction to Quantitative Research 203

led to adjustments in the follow-up study, in which performed all of the teaching and demonstrations
a small compensation was mailed with the ques- to avoid variance in the intervention from multi-
tionnaire and the research nurse conducted the ple teachers.
1-month discharge follow-up by phone interview. An efficient design can maximize results,
These pragmatic considerations are not pre- decrease errors, and control preexisting condi-
sented as a step in the research process, as are the tions that may affect outcomes. To accomplish
theoretical framework and methods, but they do these tasks, the research design and methods
affect every step of the process and therefore should demonstrate the researcher’s efforts at
should be considered when you assess a study. control, which is important in all designs. When
For example, the student researcher may or may various research designs are critiqued, the issue
not have monies or accessible services. When you of control is always raised but with varying levels
critique a study, note the credentials of the author of exibility. The issues discussed here will
or authors and whether the investigation was part become clearer as you review the various types
of either a student project or a fully funded grant of designs.
project. If the project was a student project, the Control is accomplished by ruling out extrane-
standards of critiquing are applied more liberally ous variables that compete with the independent
than for projects conducted by an experienced variables as an explanation for a study’s outcome.
researcher or clinician with a doctoral degree. An e traneous variable (also called a mediating
Finally, the pragmatic issues raised affect the variable) interferes with the operations of the
scope and breadth of an investigation and there- phenomena being studied (e.g., age and gender).
fore its generalizability. Means of controlling extraneous variables include
the following:
se of a homogeneous sample
CONTROL se of consistent data-collection proce-
A researcher attempts to use a design to maximize dures
the degree of control over the tested variables. Manipulation of the independent variable
Control involves holding the conditions of the Randomization
study constant and establishing specific sampling An investigator might be interested in how a new
criteria, as described by Sobieraj and colleagues smoking cessation program (independent vari-
(2009 see Appendix C). An efficient design can able) affects smoking behaviour (dependent vari-
maximize results, decrease errors, and control able). The independent variable is assumed to
preexisting conditions that may affect outcome. affect the outcome or dependent variable. An
To accomplish these tasks, the research design investigator needs to be relatively sure that the
and methods should demonstrate the researcher’s decrease in smoking is truly related to the smoking
efforts at control. cessation program rather than to another variable,
For example, to test their hypothesis and apply such as motivation.
control, Sobieraj and colleagues (2009) calcu- The following example illustrates and defines
lated a sample size (see Chapter 12) of a minimum these concepts further. Ploeg and associates
of 25 participants per group. Thirty participants (2010) evaluated the effect of an intervention
were recruited to compensate for those who involving a provider-initiated primary care out-
dropped out. The intervention was controlled by reach program in comparison with usual care
asking the participants to demonstrate their among older adults at risk of functional decline.
massage technique to ensure proper administra- To rule out the effects of extraneous variables on
tion. A registered nurse with a master’s degree the quality of life (dependent variable) among the
204 PART THREE Quantitative Research

older adults, demographic information was col- lived in Hamilton, Ontario. The sample was
lected, including gender, marital status, level of therefore homogeneous with regard to age, lan-
education, household income, whether the par- guage, and location of home. This control step
ticipant was living alone, and scores on the mini- limits the generalizability or application of the
mental state examination (MMSE). Although the outcomes to other populations when the outcomes
design of the research study alone does not inher- are analyzed and discussed (see Chapter 17). The
ently provide control, an appropriately designed results can then be generalized only to a similar
study with the necessary controls can increase population of individuals. Homogeneity could be
an investigator’s ability to answer a research considered limiting, but not necessarily because
question. no treatment or program is applicable to all popu-
lations, and educated consumers of research must
Evidence-Informed Practice Tip take into consideration the differences in pop-
As you read a report, assess whether the study ulations.
includes a tested intervention and whether the report
contains a clear description of the intervention and how Helpful Hint
it was controlled. If the details are not clear, the interven- When reviewing studies, remember that it is
tion may have been administered differently among the better to have a “clean” study, whose results can be
participants, which would affect the interpretation of used to make generalizations about a specific popula-
the results. tion, than a “messy” study, whose results may be poorly
or not at all generalizable.

Homogeneous Sampling If the researcher believes that one of the extra-

In the example of smoking cessation, extraneous neous variables is important, it may be included
variables may affect the dependent variable. The in the design. In the smoking cessation example,
characteristics of a study’s participants are com- if individuals are working in an area where
mon extraneous variables. Age, gender, length of smoking is not allowed and this condition is con-
time smoked, amount smoked, and even smoking sidered to be important to the study, the researcher
rules may affect the outcome in the smoking ces- could account for it in the design and set up a
sation example, even though they are extraneous control condition for it. This condition can be
or outside the study’s design. As a control for established by comparing two different work
these and other similar problems, the researcher’s areas: one where smoking is allowed and one
participants should demonstrate ho ogeneity or where it is not. Of importance is that before the
similarity with regard to the extraneous variables data are collected, the researcher should have
relevant to the particular study (see Chapter 13). identified, planned for, and controlled the impor-
Extraneous variables are not fixed but must be tant extraneous variables.
reviewed, and their inclusion in the analyses is
based on the study’s purpose and theoretical base. Constancy in Data Collection
By using a sample of homogeneous participants, Another basic but critical component of control
the researcher has used a straightforward step of is constancy in data-collection procedures. Con
control. stancy refers to the ability of the data-collection
For example, Ploeg and associates (2010) design to hold the conditions of the study to a
ensured the homogeneity of the sample by using cookbook-like recipe. In other words, for the
the following sampling criteria: Participants were purpose of collecting data for the study, each par-
aged 75 years or older, they or their proxy were ticipant is exposed to the same environmental
able to answer questions in English, and they conditions, timing of data collection, data-
 CHAPTER 9 Introduction to Quantitative Research 205

collection instruments, and data-collection proce- vention to only one group within the study but
dures (see Chapter 13). not to the other participants in the study. The
An example of constancy in data collection is first group is known as the e eri ental grou
illustrated in the study by Doran and associates and the other group is known as the control
(2006). The objective of this study was to explore grou or comparison group. In a control group,
which nursing interventions provided during hos- the variables under study are held at a constant
pitalization are associated with patients’ thera- or comparison level. For example, Sinclair and
peutic self-care and functional health outcomes. Ferguson (2009) examined whether combin-
Nurses collected data on patient outcomes using ing classroom learning with simulated learning
the Minimum Data Set and the Therapeutic Self- activities would increase self-efficacy in nursing
Care Scale. Chart audits were conducted by practice. The experimental group received the
research assistants. Constancy was attained by intervention, whereas the control group received
training the research assistants in a half-day the standard 2 hours of lectures per week.
session with a prescribed protocol, which included Experimental and quasiexperimental designs
an introduction to the nursing intervention audit involve manipulation, whereas in nonexperi-
tool, instruction in the operational definition of mental designs, the independent variable is not
each nursing intervention item, and the opportu- manipulated. This lack of manipulation does not
nity to conduct a chart audit with feedback. The decrease the usefulness of a nonexperimental
audit of a research team member was used as the design, but the use of a control group in an experi-
standard against which the research assistants’ mental or quasiexperimental design is related to
chart audits were compared. This type of control the research question and, again, its theoretical
aided the investigators’ ability to draw conclu- framework.
sions, discuss the findings, and cite the need for Blinding is a technique used in experimental
further research in this area. For the consumer, and quasiexperimental research in which the
constancy demonstrates a clear, consistent, and participants are not aware of whether they are
specific means of data collection. receiving the intervention. Double blinding is a
When psychosocial interventions are imple- technique in which both the researchers and the
mented, researchers often describe the training of participants are not aware of who is receiving the
interventionists or data collectors that took place intervention and who is in the control group. For
to ensure constancy. In a study by Kilty and Pren- example, Newby and colleagues (2009) tested the
tice (2010) on the outcomes of a school nurse feasibility of using topical tetracaine, in compari-
referral to a family physician for adolescents son with a placebo, to reduce the pain of intra-
identified with elevated cholesterol or blood pres- muscular immunization injections in infants. The
sure risk factors, interviews were conducted by administration was video-recorded, and a paedi-
research assistants. Training was done to ensure atric nurse assessed the level of pain by watching
that the data collection, recording, and manage- the videos. This nurse was blind as to which
ment methods were consistent and that completed infant received the tetracaine or placebo.
forms were reviewed for quality assurance.
Helpful Hint
Manipulation of the Be aware that the lack of manipulation of the
Independent Variable independent variable does not mean that the study is
A third means of control is manipulation of the weaker. The level of the problem, the amount of theo-
retical work, and the research that has preceded
independent variable. Manipulation refers to the the project affect the researcher’s choice of the design.
administration of a program, treatment, or inter- If the problem is amenable to a design in which the
206 PART THREE Quantitative Research

independent variable can be manipulated, the power of to improve care in the ambulatory care centre. In
a researcher to draw conclusions will increase, provided critiquing this type of study, the issue of control
that all of the considerations of control are equally
addressed. should be applied in a highly exible manner
because of the preliminary nature of the work.
If from a review of a study you determine that
Randomization the researcher intended to conduct a correlational
Researchers may also choose other forms of study (an examination of the relationship between
control, such as randomization. ando i ation or among the variables), then the issue of control
is a sampling selection procedure in which each takes on more importance. Control must be exer-
participant in a population has an equal chance of cised as strictly as possible. At this intermediate
being assigned to either the experimental group level of design, it should be clear to the reviewer
or the control group. Randomization eliminates that the researcher considered the extraneous
bias, aids in the attainment of a representative variables that may affect the outcomes.
sample, and can be used in various designs. John- All aspects of control are strictly applied to
ston and associates (2010) examined whether the studies that use an experimental design. The
use of iPod technology in medical-surgical reader should be able to locate in the research
nursing courses would have an effect on grades. report how the researcher met these criteria:
Some students had their own iPod. Students whether the conditions of the research were con-
without their own iPod were randomly assigned stant throughout the study, the assignment of par-
to one of three experimental groups or to the ticipants was random, and experimental and
control group. control groups were used. Because of the control
Randomization can also be accomplished with exercised in the study, the reader can determine
paper-and-pencil-type instruments. By randomly that all issues related to control were considered
ordering items on the instruments, the investiga- and the extraneous variables were addressed.
tor can assess whether a difference in responses
is correlated with the order of the items. Random-
ization may be especially important in longitudi- Evidence-Informed Practice Tip
Remember that establishing evidence for practice
nal studies, in which bias from giving the same is determined by assessing the validity of each step of
instrument to the same participants on a number the study, assessing whether the evidence assists in
of occasions can be a problem (see Chapter 12). planning patient care, and assessing whether patients
respond to the evidence-informed care.
QUANTITATIVE CONTROL
AND FLEXIBILITY
The same level of control cannot be exercised in INTERNAL AND EXTERNAL VALIDITY
all types of designs. At times, when a researcher Consumers of research must believe that the
wants to explore an area in which little or no results of a study are valid, based on precision,
literature on the concept exists, the researcher and faithful to what the researcher wanted to
will probably use an exploratory design. In this measure. To form the basis of further research,
type of study, the researcher is interested in practice, and theory development, a study must
describing or categorizing a phenomenon in a be credible and dependable. The two important
group of individuals. Richard and colleagues criteria for evaluating the credibility and depend-
(2010) engaged in survey research to identify ability of the results are internal validity and
areas that they considered priorities for change in external validity. Threats to validity are listed in
cancer care. They obtained information on how Box 9-1, and a discussion of each threat follows.
 CHAPTER 9 Introduction to Quantitative Research 207

BOX 9-1 breast-feeding education program on the length

THREATS TO VALIDITY of time of breast-feeding, government-sponsored
breast-feeding promotions on television and in
INTERNAL VALIDITY
newspapers could affect the length of time of
History threats
Maturation effects breast-feeding and would be considered a threat
Testing effects of history. A published example of history affect-
Instrumentation threats ing research findings is Bull and associates’
Mortality (attrition)
Selection bias (2000) study (see Table 9-2).

EXTERNAL VALIDITY Maturation Effects

Selection effects
Reactive effects aturation refers to the developmental, biologi-
Measurement effects cal, or psychological processes that operate within
an individual as a function of time these pro-
cesses are external to the events of the investiga-
tion. For example, suppose that a researcher
Internal Validity wished to evaluate the effect of a specific teach-
Internal validity is the degree to which the ing method on the achievements of baccalaureate
experimental treatment, not an uncontrolled con- students on a skills test. The investigator would
dition, resulted in the observed effects. To estab- record the students’ abilities before and after the
lish internal validity, the researcher rules out teaching method. Between the pretest and the
other factors or threats as rival explanations of the posttest, the students would have grown older
relationship between the variables. Threats to and wiser. The growth or change is unrelated
internal validity may be numerous and are con- to the investigation, and the differences between
sidered by researchers in planning a study and by the two testing periods may be explained by such
consumers before implementing the results in maturation rather than by the experimental
practice (Campbell Stanley, 1996). Research treatment.
consumers should note that the threats to internal Maturation effects could also occur in a study
validity are most clearly applicable to experimen- of the relationship between two methods of teach-
tal designs, but attention to factors that can com- ing about children’s knowledge of self-care mea-
promise outcomes should be considered to some sures. Posttests of student learning must be
degree in all quantitative designs. If these threats conducted relatively soon after the teaching ses-
are not considered, they could negate the results sions are completed. Such a short interval allows
of the research by affecting the design. Threats the investigator to conclude that the results were
to internal validity include history threats, matu- the outcome of the design of the study and not
ration effects, testing effects, instrumentation maturation in a population of children who are
threats, attrition and selection bias. Table 9-2 learning new skills rapidly. Maturation is more
provides examples of these threats. than change that results from an age-related
developmental process maturation can also be
History Threats related to physical changes (see Table 9-2).
Not only the independent variable but also another
specific event may affect the dependent variable, Testing Effects
either inside or outside the experimental setting. Taking the same test repeatedly could in uence
This threat to internal validity is referred to as the participants’ responses the next time the test
history threat In a study of the effects of a is completed. For example, the effect on the
208 PART THREE Quantitative Research

TABLE 9-2
EXAMPLES OF INTERNAL VALIDITY THREATS
THREAT EXAMPLE

History threat Bull, Hansen, and Gross (2000) tested a teaching intervention in one hospital and compared the
outcomes with those at another hospital in which the usual care was given. During the final months
of data collection, the control hospital implemented a critical care pathway for congestive heart
failure; as a result, data from the control hospital (cohort 4) were not included in the analysis.
Maturation effect Koniak-Griffin, Verzemnieks, Anderson, Brecht, Lesser, Kim, et al. (2003) evaluated the 2-year
postbirth infant health and maternal outcomes of an early intervention program by public health
nurses and noted that the lack of change in some of the variables may have been attributable to
the general maturation changes experienced by new mothers rather than to the intervention.
Testing effect A researcher wishes to measure acute pain with a repeated-measures design during a lengthy
procedure. The researcher must consider the results in view of the possible bias of repeating the
pain measurements over a short period of time. The measurements may prime the patients’
responses, and the practice of reporting pain repeatedly on the same instrument during a
procedure may influence the results. Bennett, Lyons, Winters-Stone, and Hanson (2007) evaluated
the effect of a motivational intervention on increasing physical activity in long-term cancer
survivors. Several established instruments were used to measure variables. The measure of
physical activity was obtained through self-report, which the researchers noted to be a possible
limitation. The repeated self-measurements may have primed the patients’ responses and primed
the results.
Instrumentation Inoue, Kakehashi, Oomori, and Koizumi (2004) studied biochemical hypoglycemia in female nurses
threat during shift work in Japan. The nurses determined their own blood glucose levels 12 times, at four
points during each of three shifts. The researchers noted that the study depended on self-testing
and self-reporting of blood glucose levels; thus, documenting the validity of the data was difficult.
Even though the study was well developed and participants were ensured confidentiality, the
researchers still were concerned with this threat to the study’s validity. Holditch-
Davis, Brandon, and Schwartz (2003) examined the development of eight behaviours in preterm
infants. Data collectors were trained and assessed for interrater reliability, which thus precluded the
threat of instrumentation.
Mortality (attrition) Stewart, Reutter, Letourneau, and Makwarimba (2009) tested a support intervention for homeless
youth to optimize peer influence. A total of 70 youths were recruited to the study; 56 completed
some part of the intervention. Attrition over time was a major challenge, resulting in a small
sample and affecting the power of the study.
Selection bias Smith, Corso, Brown, and Cameron (2011) controlled for selection bias by establishing selection
criteria and having two groups: one receiving an intensive brief smoking cessation intervention
(like an experimental group), the other receiving a brief smoking cessation intervention, which was
used as the comparator.

participant’s posttest score as the result of having Sinclair and Ferguson’s (2009) pretests and
taken a pretest is known as a testing e ect The posttests consisted of the same self-efficacy ques-
effect of taking a pretest may sensitize an indi- tionnaire. Whether the significant increase in self-
vidual and improve the score on the posttest. Indi- efficacy resulted from the teaching/learning
viduals generally score higher when they take a strategies or was the effect of taking the test more
test a second time, regardless of the treatment. than once was difficult to determine. Table 9-2
The differences between posttest and pretest provides another example of a testing effect.
scores may be a result not of the independent
variable but rather of the experience gained Instrumentation Threats
through the testing. For example, in a study of the Instru entation threats are changes in the
effect of integrating simulated teaching/learning variables or observational techniques that may
strategies in undergraduate nursing education, account for changes in the obtained measurement.
 CHAPTER 9 Introduction to Quantitative Research 209

For example, a researcher may wish to study from one group than from the other group this
various types of thermometers (e.g., tympanic, effect is known as differential loss of participants
digital, electronic, chemical indicator, plastic strip, In a study of the ways in which a media campaign
and mercury) to compare the accuracy of the affects the incidence of breast-feeding, if most
mercury thermometer with the other temperature- dropouts were non–breast-feeding women, the
taking methods. To prevent instrumentation perception given could be that exposure to the
threats, the researcher must check the calibration media campaign increased the number of breast-
of the thermometers according to the manufactur- feeding women, whereas the effect of experimen-
er’s specifications before and after data collection. tal attrition led to the observed results. See Table
Another example concerns techniques of 9-2 for an example of a study in which mortality
observation or data collection. If a researcher has (attrition) may have in uenced the results.
several raters collecting observational data, all
must be trained in a similar manner. If they are Selection Bias
not similarly trained, or even if they are similarly If precautions are not used to gain a representa-
trained but unable to conduct the study as planned, tive sample, selection bias—the threat to internal
a lack of consistency may occur in their ratings validity that arises when pretreatment differences
therefore, a threat to internal validity will occur. exist between the experimental group and the
Boechler and colleagues (2003) examined control group—could result from the way the par-
whether the amount of caregiving is related to the ticipants were chosen. Selection effects are a
behaviour of a father and his child during a struc- problem in studies in which the individuals them-
tured teaching interaction. The father was selves decide whether to participate in a study.
observed teaching his child to use a standardized Suppose an investigator wishes to assess whether
teaching item, such as grabbing a ring or taking a new smoking cessation program contributes to
the lid off a container. Two trained female observ- smoking cessation. If the new program is offered
ers watched the father and scored the interaction to all smokers, chances are that only individuals
on the Nursing Child Assessment Teaching Scale who are more motivated to stop smoking will take
only consensus scores from both observers were part in the program. Assessment of the effective-
used in the data analysis. (For another example, ness of the program is problematic because the
see Table 9-2.) Although the researcher takes investigator cannot know for certain whether the
steps to prevent problems of instrumentation, the new program encouraged smoking cessation
threat of instrumentation may still occur. When a behaviours or whether only highly motivated
critiquer finds such a threat, it must be evaluated individuals joined the program. To avoid selec-
within the total context of the study. tion bias, the researcher could randomly assign
participants to either the new teaching method
Mortality group or a control group that receives a different
ortality or attrition is the loss of study partici- type of instruction. Table 9-2 provides another
pants from the first data-collection point (pretest) example of selection bias.
to the second data-collection point (posttest). If
the participants who remain in the study are not Helpful Hint
similar to those who dropped out, the results The list of threats to internal validity is not exhaus-
tive. More than one threat can be found in a study,
could be affected. The loss of participants may be depending on the type of study design. Finding a threat
from the sample as a whole, or, in a study that to internal validity in a study does not invalidate the
has both an experimental group and a control results and is usually acknowledged by the investigator
in the “Results” or “Discussion” section of the study.
group, more of the participants may drop out
210 PART THREE Quantitative Research

conditions, and type of observations. These

Evidence-Informed Practice Tip
factors are termed selection effects, reactive
Avoiding threats to internal validity in clinical
research can be difficult. However, this reality does not effects, and testing effects You may notice the
render studies that have threats useless. Take the threats similarity in the names of the factors of selection
into consideration, and weigh the total evidence of a and testing and those of the threats to internal
study for not only its statistical meaningfulness but also
its clinical meaningfulness.
validity. When considering factors as internal
threats, the reader assesses them as they relate to
the independent and dependent variables within
External Validity the study when assessing them as external threats,
E ternal validity concerns the generalizability the reader considers them in terms of the gener-
of an investigation’s findings to additional popu- alizability, or use outside the study with other
lations and to other environmental conditions. To populations and settings.
achieve external validity, variation in the condi- The Critical Thinking Decision Path displays
tions and the types of participants should lead to the ways in which threats to internal and external
the same results. The goal of the researcher is to validity can interact. This path is not, however,
select a design that maximizes both internal and exhaustive with regard to the type of threats
external validity, but attaining this goal is not and their interaction. In comparison with prob-
always possible. If it is not possible, the researcher lems of internal validity, generalizability issues
must attain the minimum criterion of external are typically more difficult to deal with because
validity. they mean that the researcher is assuming that
The factors that may affect external validity other populations are similar to the one being
are related to the selection of participants, study tested.

CRITICAL THINKING DECISION PATH

Potential Threats to a Study’s Validity

Internal Validity External Validity

Selection bias

Mortality (attrition) Selection effects

Maturation effects

Instrumentation threats Measurement effects

Testing effects

Reactive effects
History threats
 CHAPTER 9 Introduction to Quantitative Research 211

street youth in Toronto come from various

Selection Effects locations in Ontario and Canada. (Linton,
Selection concerns the generalizability of the Singh, Turbow, Legg, 2009, p. 389)
results to other populations. An example of selec These remarks are cautionary, but also they also
tion e ects is when the researcher cannot attain point out the usefulness of the findings for prac-
the ideal sample population. At times, the number tice and future research aimed at building the data
of available participants may be low, or they may in these areas.
not be accessible to the researcher. The researcher
may then need to choose a nonprobability method Reactive Effects
of sampling, not a probability method. Therefore, eactivity is defined as the participants’ responses
the type of sampling method used and how par- to being studied. Participants may behave in a
ticipants are assigned to research conditions will certain way with the investigator not because of
affect the generalizability of findings to other the study procedures but merely as an indepen-
groups, or the external validity. In the following dent response to being studied. This response is
quotations, the authors have noted selection also known as the a thorne e ect named after
effects: Western Electric Corporation’s Hawthorne plant,
The results of this study are limited by the where a study of working conditions was con-
small sample size. (Newby, Faschoway, ducted in the 1930s. The researchers developed
Souroroff, 2009, p. 531) several different working conditions, such as
The relatively small sample size of turning up the lights, piping in music loudly or
Chinese youth and low variability of softly, and changing work hours. They found that
peer risky behaviours may account for the no matter what was done, the workers’ productiv-
nonsignificant findings. (Wilgerodt, 2008, ity increased. They concluded that production
p. 404) increased as a result of the workers’ knowing that
Study limitations include the possibility they were being studied rather than because of the
of selection bias. Our sample was chosen experimental conditions.
from two professional organizations whose For example, in a randomized controlled study,
members may have been more likely to Bennett and colleagues (2007) tested the effect of
be concerned about safety, skewing the telephone-based motivational interviewing on
results . . . which limits the generalizability increasing physical activity and improving physi-
of our findings. (Wagner, Capezuti, cal fitness, thus improving health self-efficacy in
Rice, 2009, p. 190) adults living in rural areas. The researchers noted
One primary limitation of our study is that that the physical activity counsellors knew to
we only surveyed street youth in Toronto which group (control or experimental) the partici-
who accessed preventive health services. pants were assigned, which may have allowed the
Therefore, the demographic and behav- control group to become aware of components of
ioural characteristics of street youth in our the experimental treatment. The researchers also
study may not be generalized to street youth noted that each participant also participated in a
elsewhere. Street youth of Toronto tend to 6-minute walk test in which a strict protocol was
come from a variety of racial and ethnic used, but it was also possible that conversation
backgrounds, but in our study, categories of before the test might have in uenced perfor-
ethnicity were limited to three levels Cau- mance rate. The researchers made recommenda-
casian, Blacks and Other therefore, we tions about how to avoid such threats in future
cannot quantify the scope of diversity. Also, studies.
212 PART THREE Quantitative Research

Thomas and colleagues (2010), who assessed examine their attitudes regarding AIDS. The par-
anaemia and red blood cell transfusion practices, ticipants’ responses on follow-up testing may
noted that physicians in the intensive care unit differ from those of individuals who were given
were aware that phlebotomy and transfusion the education program and did not see the pretest.
practices were being recorded. This could Therefore, when a study is conducted and a
have led to temporary modifications in behav- pretest is given, it may prime the participants and
iour during the study, thereby affecting the affect their subsequent answers, which in turn can
outcomes. affect the generalizability of the findings.

Measurement Effects
Helpful Hint
Administration of a pretest in a study affects the When you review a study, be aware of the inter-
generalizability of the findings to other popula- nal and external threats to validity. These threats do not
tions the resulting changes are known as ea render a study useless; instead, they make it more
useful to you. Recognition of the threats allows research-
sure ent e ects Just as pretesting affects the ers to build on data and allows consumers to think
posttest results within a study, pretesting affects through what part of the study can be applied to prac-
the posttest results and generalizability outside tice. Specific threats to validity depend on the type of
the study. For example, suppose a researcher design and generalizations that the researcher hopes to
make.
wants to conduct a study with the aim of changing
attitudes toward acquired immune deficiency
syndrome (AIDS). To accomplish this task, an Other threats to external validity depend on the
education program on the risk factors for AIDS type of design and methods of sampling used by
is incorporated. To test whether the education the researcher but are beyond the scope of this
program changes attitudes toward AIDS, tests are text. Campbell and Stanley (1996) offered detailed
given before and after the teaching intervention. coverage of the issues related to internal and
The pretest on attitudes allows the participants to external validity.

APPRAISING THE EVIDENCE

Quantitative Research
Critiquing the design of a study requires knowledge of the drawn from this chapter. Remember that the criteria are
overall implications of a particular design for the study as applied differently with various designs. Differences in
a whole (see Critiquing Criteria box). Researchers want to application do not mean that the consumer will find a
consider the level of evidence provided by the design and haphazard approach to design but rather that each design
how the study can be used to improve or change practice. has particular criteria that allow the consumer to classify
Minimizing threats to internal and external validity the design by type (e.g., experimental or nonexperimen-
enhances the strength of evidence for any quantitative tal). These criteria must be met and addressed in conduct-
design. The concept of the research design is all-inclusive ing an experiment. The particulars of specific designs are
and parallels the concept of the theoretical framework. addressed in Chapters 10 and 11. The following discussion
The research design is similar to the theoretical frame- primarily pertains to the overall evaluation of a quantita-
work in that it deals with a piece of the research study tive research design.
that affects the whole. This chapter has introduced the The research outcome should demonstrate that an
meaning, purpose, and important factors of design objective review of the literature and the establishment
choice, as well as the vocabulary that accompanies these of a theoretical framework guided the choice of the
factors. Several criteria for evaluating the design can be design. No explicit statement regarding these areas is
 CHAPTER 9 Introduction to Quantitative Research 213

APPRAISING THE EVIDENCE

Quantitative Research—cont’d
made in a research article. A consumer can evaluate the of the variables tested—and thus internally valid—or
design by critiquing the theoretical framework (see whether another explanation is possible. To assess this
Chapter 2) and literature review (see Chapter 5). Is the aspect, you should review the threats to internal validity.
question new and not extensively researched? Has a If the investigator’s study was systematic, was well
great deal of research been conducted on the question, grounded in theory, and followed the criteria for each of
or is the question a new or different way of looking at an the processes, you will probably conclude that the study
old question? Depending on the level of the question, the is internally valid.
investigators make certain choices. These choices enable In addition, you must know whether a study has
researchers to look for differences in a controlled, com- external validity or generalizability to other populations
parative manner. or environmental conditions. External validity can be
The consumer should be alert for the methods that claimed only after internal validity has been established.
investigators use to maintain control (e.g., homogeneity If the credibility of a study (internal validity) has not been
in the sample, consistent data-collection procedures, established, a study has no generalizability to other pop-
manipulation of the independent variable, and random- ulations (external validity). Determination of external
ization). As discussed in Chapter 10, all of these criteria validity is related directly to the sampling method
must be met for an experimental design. As you begin (see Chapter 12). If the study is not representative of any
to understand the types of designs (i.e., experimental, one group or phenomena of interest, external validity
quasiexperimental, and nonexperimental designs such as may be limited or not present. The establishment of inter-
survey and relationship designs), you will find that control nal and external validity requires not only knowledge of
is applied in varying degrees or—as in the case of a the threats to internal and external validity but also
survey study—the independent variable is not manipu- knowledge of the phenomena being studied, which
lated (see Chapter 11). The level of control and its applica- allows critical judgements to be made about the linkage
tions presented in Chapters 10 and 11 provide the of theories and variables for testing. You should find
remaining knowledge for fully critiquing the aspects of a that the design follows from the theoretical framework,
study’s design. literature review, research question, and hypotheses. You
Once you have established whether the necessary should believe, on the basis of clinical knowledge and
control or uniformity of conditions has been maintained, knowledge of the research process, that the investigators
you must determine whether the study is believable or are not, as the expression goes, comparing apples with
valid. You should ask whether the findings are the result oranges.

CRITIQUING CRITERIA

1. Is the type of study design 4. Does the design flow from the 7. What are the threats to external
employed appropriate? proposed research question, validity?
2. Does the researcher use theoretical framework, 8. What are the controls for
the various concepts of literature review, and the threats to external
control that are consistent hypothesis? validity?
with the type of design 5. What are the threats to internal 9. Is the design appropriately linked
chosen? validity? to the levels of evidence
3. Does the design seem to reflect 6. What are the controls for the hierarchy?
feasibility? threats to internal validity?
214 PART THREE Quantitative Research

CRITICAL THINKING CHALLENGES • Control affects not only the outcome of a study
but also its future use. The design should also
reflect how the investigator attempted to control
■ Consider the following statement: “All research threats to both internal and external validity.
attempts to solve problems.” How would you • Internal validity must be established before
support or refute this statement? external validity can be established. Both are
■ As a consumer of research, you recognize that considered within the sampling structure.
control is an important concept in the issue of • No matter which design the researcher chooses, it
should be evident to the reader that the choice
research design. You are critiquing an assigned was based on a thorough examination of the
experimental study as part of your “open-book” research question within a theoretical framework.
midterm examination. From what is written, you • The design, research question, literature review,
cannot determine how the researchers kept the theoretical framework, and hypothesis should all
conditions of the study constant. How does this be interrelated.
• The choice of the design is affected by pragmatic
characteristic affect the study’s use in an issues. At times, two different designs may be
evidence-informed practice model? equally valid for the same question.
■ Box 9-1 lists six major threats to the internal • The choice of design affects the study’s level of
validity of an experimental study. Prioritize them, evidence.
and defend the one that you deem the essential,
or number one, threat to address in a study.
■ You are critiquing the research design of an REFERENCES
assigned study as a consumer of research. How
Bennett, J. A., Lyons, K. S., Winters-Stone K.,
does the research design influence the findings Hanson, G. (2007). Motivational interviewing to
of evidence in the study? increase physical activity in long-term cancer
■ How do threats to external validity contribute to survivors: A randomized controlled trial. Nursing
the strength and quality of evidence provided by Research, (1), 18-27.
the findings of a research study? Boechler, V., Harrison, M. J., Magill-Evans, J.
(2003). Father-child teaching interventions: The
relationship to father involvement in caregiving.
ournal of Pediatric Nursing, , 46-51.
Bull, M. J., Hansen, H. E., Gross, C. R. (2000). A
KEY POINTS professional-patient partnership model of discharge
planning with elders hospitalized with heart failure.
• The purpose of the design is to provide the format Applied Nursing Research, , 19-28.
of masterful and accurate research. Campbell, D., Stanley, J. (1996). E perimental and
• Many types of designs exist. No matter which type uasi-e perimental designs for research. Chicago:
of design the researcher uses, the purpose remains Rand-McNally.
the same. Doran, D., Harrison, M. B., Laschinger, H., Hirdes, J.,
• The research consumer should be able to locate Rukholm, E., Sidani, S., et al. (2006). Relationship
within the study a sense of the question that the
between nursing interventions and outcome achieve-
researcher wished to answer. The question should
be proposed with a plan or scheme for the ment in acute care settings. Research in Nursing
accomplishment of the investigation. Depending ealth, , 61-70.
on the question, the consumer should be able to Hayward, K., Campbell-Yeo, M., Price, S., Morrison,
recognize the steps taken by the investigator to D., Whyte, R., Cake, H., Vine, J. (2007).
ensure control. Co-bedding twins: How pilot study findings guided
• The choice of the specific design depends on the improvements in planning a larger multicenter trial.
nature of the question. To specify the nature of the Nursing Research, , 137-143.
research question, the design must reflect the Holditch-Davis, D., Brandon, D. H., Schwartz, T.
investigator’s attempts to maintain objectivity, (2003). Development of behaviors in preterm infants:
accuracy, pragmatic considerations, and, most
Relation to sleeping and waking. Nursing Research,
important, control.
, 307-317.
 CHAPTER 9 Introduction to Quantitative Research 215

Inoue, K., Kakehashi, Y., Oomori, S., Koizumi, A. important dimension of quality in cancer care.
(2004). Endemic hypoglycemia among female nurses. ournal of Nursing Care uality, (4), 344-351.
Research in Nursing and ealth, , 87-96. Sinclair, B., Ferguson, K. (2009). Integrating
Ireland, S. E., Arthur, H. M., Gunn, E. A., simulated teaching/learning strategies in undergradu-
Oczkowski, W. (2010). Stroke prevention care ate nursing education. International ournal of
delivery: Predictors of risk factor management Nursing Education and Scholarship, (1), Article 7.
outcomes. International ournal of Nursing Studies, Smith, P. M., Corso, L., Brown, K. S., Cameron, R.
, 156-164. (2011). Nurse case-managed tobacco cessation
Johnston, R., Hepworth, J., Goldsmith, M., Lacasse, interventions for general hospital patients: Results of
C. (2010). se of iPod(tm) technology in medical- a randomized clinical trial. Canadian ournal of
surgical nursing courses: Effect on grades. Interna- Nursing Research, (1), 98-117.
tional ournal of Nursing Education Scholarship, Sobieraj, G., Bhatt, M., LeMay, S., Rennick, J.,
(1), 1-17. Johnston, C. (2009). The effect of music on parental
Kilty, H. L., Prentice, D. (2010). Adolescent participation during pediatric laceration repair.
cardiovascular risk factors: A follow-up study. Canadian ournal of Nursing Research, (4),
Clinical Nursing Research, (1), 6-20. 68-82.
Koniak-Griffin, D., Verzemnieks, I. L., Anderson, Steenbeek, A., Edgecombe, N., Durling, J., LeBlanc, A.,
N. L. R., Brecht, M., Lesser, J., Kim, S., et al. (2003). Anderson, R., Bainbridge, R. (2009). sing an
Nurse visitation for adolescent mothers: Two-year interactive journal club to enhance nursing knowledge
infant health and maternal outcomes. Nursing acquisition, appraisal, and application. International
Research, , 127-136. ournal of Nursing Education Scholarship, (1), 1-8.
Linton, A., Singh, M., Turbow, D., Legg, T. J. (2009). Stewart, M., Reutter, L., Letourneau, N.,
Street youth in Toronto: An investigation of demo- Makwarimba, E. (2009). A support intervention to
graphic predictors of high risk behaviors and HIV promote health and coping among homeless youths.
status. ournal of IV AIDS Social Services, , Canadian ournal of Nursing Research, (2),
375-396. 54-77.
McGilton, K., Sorin-Peters, R., Sidani, S., Rochon, E., Thomas, J., Jensen, L., Nahirniak, S., Gibney,
Boscart, V., Fox, M. (2011). Focus on communica- R. T. N. (2010). Anemia and blood transfusion
tion: Increasing the opportunity for successful practices in the critically ill: A prospective cohort
staff-patient interactions. International ournal of review. eart ung, (3), 217-225.
lder People, (1), 13-24. Wagner, L. M., Capezuti, E., Rice, J. C. (2009).
Newby, B. D., Faschoway, G. D., Souroroff, C. I. Nurses’ perceptions of safety culture in long-term
(2009). Tetracaine (Ametop) compared to placebo for care settings. ournal of Nursing Scholarship, (2),
reducing pain associated with intramuscular injection 184-192.
of palivizumab (Synagis). ournal of Pediatric Wilgerodt, M. A. (2008). Family and peer in uences on
Nursing, (6), 529-533. adjustment among Chinese, Filipino and white youth.
Ploeg, J., Brazil, K., Hutchison, B., Kaczorowski, J., Nursing Research, (6), 395-405.
Dalby, D. M., Goldsmith, C. H., Furlong, W.
(2010). Effect of preventive primary care outreach on
health related quality of life among older adults at
FOR FURTHER STUDY
risk of functional decline: Randomised controlled Go to Evolve at http://evolve.elsevier.com/
trial. British Medical ournal, , c1480. Canada/LoBiondo/Research for Audio Glossary, how-to
doi:101136.bmj.c1480 instructions for Writing Proposals for Funding, and
Richard, M. L., Parmar, M. P., Calestagne, P. P., additional research articles for practice in reviewing
McVey, L. (2010). Seeking patient feedback: An and critiquing.
C H A PTER 1 0

Experimental and
Quasiexperimental Designs
Susan Sullivan-Bolyai | Carol Bova | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• List the criteria necessary for inferring cause-and-effect relationships.
• Distinguish the differences between experimental and quasiexperimental designs.
• Define problems with internal validity that are associated with experimental and
quasiexperimental designs.
• Describe the use of experimental and quasiexperimental designs for evaluation research.
• Critically evaluate the findings of selected studies in which cause-and-effect relationships were
tested.
• Apply levels of evidence to experimental and quasiexperimental designs.

KEY TERMS
a priori experimental design posttest–only control group
after-only design formative evaluation design
after-only nonequivalent independent variable quasiexperiment
control group design intervening variable quasiexperimental design
antecedent variable manipulation randomization
attrition mortality Solomon four-group design
control nonequivalent control group summative evaluation
dependent variable design testing effects
evaluation research one-group pretest–posttest time series design
experiment design true experiment

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

216
 CHAPTER 10 Experimental and Quasiexperimental Designs 217

CHAPTER 10 PROVIDED AN OVERVIEW OF the meaning, 1. The causal variable and effect variable must
purpose, and issues related to quantitative research be associated with each other.
design. This chapter provides a discussion of spe- 2. The cause must precede the effect.
cific types of quantitative designs, inasmuch as 3. The relationship must not be explainable by
choosing the correct design is crucial for hypoth- another variable.
esis testing or answering research questions. The When you critique studies in which experimental
design involves a plan, a structure, and a strategy, and quasiexperimental designs were used, the
which guide a researcher in writing the hypothesis primary focus is on the validity of the conclusion
or research questions, conducting the project, and that the experimental treatment, or the inde en
analyzing and evaluating the data. dent variable caused the desired effect on the
Each design has specific characteristics to outcome, or de endent variable The validity of
maintain control: for example, homogeneity in the conclusion depends on how well the researcher
the sample, consistent data-collection procedures, controlled the other variables that may explain
manipulation of the independent variable, and the relationship studied. Thus, the focus of this
randomization. chapter is to explain how various types of experi-
One of the fundamental purposes of scientific mental and quasiexperimental designs control
research in any profession is to determine cause- extraneous variables.
and-effect relationships. Nurses, for example, are The purpose of this chapter is to acquaint you
concerned with developing effective approaches with the issues involved in interpreting studies
to maintaining and restoring wellness. Testing that involved the use of e eri ental design
such nursing interventions to determine how well (characterized by three properties: randomiza-
they actually work—that is, evaluating the out- tion, control, and manipulation) and uasie eri
comes in terms of efficacy and cost effectiveness— ental design (in which random assignment is
is accomplished with the use of experimental and not used, but the independent variable is manipu-
quasiexperimental designs. These designs differ lated, and certain mechanisms of control are
from nonexperimental designs in one important used). Examples of these designs are listed in Box
way: The researcher actively seeks to bring about 10-1. The Critical Thinking Decision Path shows
the desired effect and does not passively observe an algorithm that in uences a researcher’s choice
behaviours or actions. In other words, the of experimental or quasiexperimental design.
researcher is interested not merely in observing
customary patient care but in making something
happen. Experimental and quasiexperimental
BOX 10-1
studies are also important to consider in relation
to evidence-informed practice because they SUMMARY OF EXPERIMENTAL AND
QUASIEXPERIMENTAL RESEARCH DESIGNS
provide level II and level III evidence. The find-
ings of such studies provide the validation of EXPERIMENTAL DESIGNS
1. True experimental (pretest–posttest control group)
clinical practice and the rationale for changing design
specific aspects of practice (see Chapter 20). 2. Solomon four-group design
Experimental designs are particularly suitable 3. After-only design
for testing cause-and-effect relationships because QUASIEXPERIMENTAL DESIGNS
they help eliminate potential alternative explana- 1. Nonequivalent control group design
tions (threats to validity) for the findings. Infer- 2. After-only nonequivalent control group design
ring causality requires that the following three 3. One group (pretest–posttest) design
4. Time series design
criteria be met.
218 PART THREE Quantitative Research

CRITICAL THINKING DECISION PATH

Experimental and Quasiexperimental Design

When the research question deals with, "What happens if . . .?"

Participants can be randomly Participants cannot be randomly

assigned to groups assigned to groups

Experimental design Quasiexperimental design

Pretest data cannot Pretest data can Pretest data cannot Pretest data can
be collected be collected be collected be collected

After-only design Pretest–posttest design After-only design Pretest–posttest design

An example of an RCT was the study by Ploeg

TRUE EXPERIMENTAL DESIGN and colleagues (2010), who compared the effect
An e eri ent is a scientific investigation that of a provider-initiated primary care outreach
makes observations and collects data according intervention with that of usual care among adults
to explicit criteria. A true e eri ent—also at risk for functional decline. They used a 1:1
known as a pretest posttest control group design allocation ratio to assign participants to the inter-
or classic e periment—has three identifying vention or control group. The sequence of alloca-
properties: randomization, control, and manipula- tion was generated in blocks of 8 or 16, so that
tion. These properties allow for other explana- the allocation was balanced in blocks. A table of
tions of the phenomenon to be ruled out and random digits was used to assign block size and
thereby provide the strength of the design for allocation.
testing cause-and-effect relationships.
A research study in which an experimental Randomization
design is used is commonly called a randomi ed ando i ation or random assignment to a
clinical trial (RCT). An RCT or experimental group, is required for a study to be considered a
design is considered to be the best research true experimental design. It involves the assign-
design, the gold standard, for providing infor- ment of participants to either the experimental or
mation about cause-and-effect relationships. An the control group on a purely random basis. In
individual RCT generates level II evidence other words, each participant has an equal and
because only minimal bias is introduced by this known probability of being assigned to any group.
design. The higher level of evidence that a design Random assignment may be performed individu-
produces, the more likely the results are to offer ally or by groups (LeMay, Johnston, Choini re,
an unbiased estimate of the effect of an interven- Fortin, Huber, Fr chette, et al., 2010 Sobieraj,
tion and the more confident you are that the inter- Bhatt, LeMay, Rennick, Johnston, 2009 see
vention will be effective and produce the same Appendix C ). Random assignment to experimen-
results over and over again. tal or control groups allows for the elimination of
 CHAPTER 10 Experimental and Quasiexperimental Designs 219

any systematic bias that may affect the dependent The concepts of control, randomization,
variable being studied. In randomization, it is and manipulation and their application to experi-
assumed that any important intervening variable mental design are sometimes confusing for stu-
(a condition that occurs during the study that dents. These concepts allow researchers to have
affects the dependent variable) will occur in an confidence in the causal inferences they make
equal distribution between the groups (as dis- by allowing them to rule out other potential
cussed in Chapter 9). Randomization minimizes explanations. Consider the use of control, ran-
variance (as discussed in Chapter 12) and domization, and manipulation in the following
decreases selection bias. Participants are ran- example. Moore and colleagues (2009) used a
domly assigned to groups through several proce- cross-sectional RCT to examine whether catheter
dures, such as a table of random numbers or washouts would prevent or reduce catheter block-
computer-generated number sequences. What- age in long-term indwelling urethral catheters.
ever method is used, it is important that the process Participants were recruited from long-term care
be truly random, that it be tamper-proof, and that units and homecare agencies.
the group assignment is concealed. Note that A computer-generated process was used for
random assignment to groups is different from the randomization to the control and experimental
random sampling discussed in Chapter 12. groups. The patients were randomly assigned to
one of three groups: control (usual care, no
Control washout), saline washout, or commercially avail-
Control refers to the introduction of one or more able acidic (Contisol) washout solution. The use
constants into the experimental situation. Control of random assignment meant that all patients who
is acquired by manipulating the causal or inde- met the study criteria had an equal and known
pendent variable, randomly assigning participants chance of being assigned to the control group or
to a group, carefully preparing experimental pro- one of the experimental groups. The use of
tocols, and using comparison groups. In experi- random assignment to groups helps ensure that
mental research, the comparison group is the the three study groups are comparable with regard
control group, or the group that receives the usual to preexisting factors that might affect the
treatment rather than the innovative, experimen- outcome of interest, such as gender, age, length
tal treatment. of stay in the hospital, and stage of cancer. Also,
the researchers in this example checked statisti-
Manipulation cally whether the procedure of random assign-
As discussed previously, experimental designs ment did, in fact, produce groups that were similar
are characterized by the researcher doing some- at baseline.
thing to at least some of the participants. The
experimental treatment is administered to some Evidence-Informed Practice Tip
participants in the study but not to others, or dif- In health care research, the term randomized clini-
ferent amounts of it are administered to different cal trial (RCT) often refers to a true experimental design.
groups. This difference in how the treatment is These designs are being used more frequently in nursing
research, which is critical to evidence-informed practice
provided is the ani ulation of the indepen- initiatives.
dent variable. The independent variable might
be a treatment, a teaching plan, or a medication.
The effect of this manipulation is measured The degree of control exerted over the experi-
to determine the result of the experimental mental conditions in Moore and colleagues’
treatment. (2009) study is illustrated by its detailed
220 PART THREE Quantitative Research

descriptions of how to perform the washouts, with urinary catheter was used. Random assignment to
visual aids on how to properly irrigate the cath- groups helps ensure that the groups are similar
eter. This control helped ensure that all members with regard to these variables so that differences
of the experimental group received similar treat- in the dependent variable may be attributed to the
ment, and it assists the reader in understanding experimental treatment. However, the researcher
the nature of the experimental treatment. The should check, and report, how the groups actually
control group provided a comparison against compared with regard to such variables. An
which the experimental group could be judged. intervening variable is a condition that occurs
In Moore and colleagues’ (2009) study, receiv- during the course of the study and is not part of
ing the saline or Contisol wash were the manipu- the study however, the intervening variable
lated treatments. Patient outcomes were measured affects the dependent variable and can affect the
for all participants, including mean urinary pH, study outcomes. An example of an intervening
incidence of microscopic hematuria and pyuria variable that might have affected the outcomes of
and incidence of urinary tract infections. The Moore and colleagues’ (2009) study is a change
primary outcome variable was mean length of in health care status in any of the participants,
time until first catheter change. By controlling the such as a newly diagnosed medical condition or
experimental intervention with the use of stan- an infection. Thus, if the care provided to patients
dard protocol, Moore and colleagues were able to changed in any major way while the study was
assert that the evidence is insufficient to state being implemented, the results of the study also
whether catheter washout with saline or Contisol would be affected.
is more effective than usual care with no washout
to prevent obstruction and clogging. Types of Experimental Designs
The use of the experimental design allows Several different experimental designs exist
researchers to rule out many of the potential (Campbell Stanley, 1966). Each is based on the
threats to internal validity of the findings, such as classic design called the true e periment, dia-
selection bias, history, and maturation effects grammed in Figure 10-1 Above the description
(see Chapter 9). The strength of the true experi- diagram, symbolic notations are routinely used:
mental design lies in its ability to help the R represents random assignment (for both
researcher control the effects of any extraneous the experimental group and the control
variables—alternative events that could explain group).
the findings—that might constitute threats to signifies observation through data collec-
internal validity. Such extraneous variables can tion on the dependent variable.
be either antecedent or intervening signifies pretest data collection.
The antecedent variable occurs before the represents posttest data collection.
study but may affect the dependent variable and represents exposure to the intervention.
confound the results. Factors such as age, gender, Therefore, in Figure 10-1, note that the partici-
socioeconomic status, and health status might be pants were assigned randomly R to the experi-
important antecedent variables in nursing research mental or the control group. The experimental
because they may affect dependent variables, treatment was given only to participants in the
such as recovery time and ability to integrate experimental group, and the pretests and
health care behaviours. Antecedent variables that posttests are the measurements of the depen-
might have affected the dependent variables dent variables that were made before and after the
in the study by Moore and colleagues (2009) experimental treatment was performed. In all true
included age, gender, and length of time that a experimental designs, participants are randomly
 CHAPTER 10 Experimental and Quasiexperimental Designs 221

A. True or classic experiment

R O1 X O2
Experimental Pretest Experimental treatment Posttest
group

Random assignment

R O1 O2
Control group Pretest Posttest
B. Solomon four-group design
R O1 X O2
Experimental Pretest Experimental treatment Posttest
group

R O1 O2
Control group Pretest Posttest
Random assignment
R X O2
Experimental Experimental treatment Posttest
group

R O2
Control group Posttest
C. After-only design
R X O2
Experimental Experimental treatment Posttest
group

Random assignment

R O2
Control group Posttest
FIGURE 10-1 Comparison of experimental designs.

assigned to groups, an experimental treatment is variable. The researcher then introduces the
introduced to some of the participants, and the experimental variable to one of the groups and
effects of the treatment are observed. The varia- measures the dependent variable again to see
tion in designs primarily concerns the number of whether it has changed. The control group
observations that are made. receives no experimental treatment, but the
As shown in Figure 10-1, participants are ran- dependent variable in that group is also measured
domly assigned to the two groups, experimental later for comparison with the experimental group.
and control, so that antecedent variables are con- The degree of difference between the two groups
trolled. Next, pretest measurements or observa- at the end of the study indicates the confidence
tions are made so that the researcher has a baseline the researcher has that a causal link exists between
for determining the effect of the independent the independent and dependent variables. Because
222 PART THREE Quantitative Research

random assignment and the control inherent in make the participants think more about how they
this design minimize the effects of many threats view themselves so that the next time they fill out
to internal validity, the true experimental design the questionnaire , their self-esteem might
is a strong design for testing cause-and-effect appear to have improved. In reality, however,
relationships. their self-esteem may be the same as it was
However, the design is not perfect. Some before the scores are different only because the
threats cannot be controlled in true experimental participants had previously taken the test. The use
studies (see Chapter 9). People tend to drop out of this design with the two groups that do not
of studies that require their participation over an receive the pretest allows for evaluating the effect
extended period. The in uence over the outcome of the pretest on the posttest in the first two
of an experiment of people dropping out or dying groups. (See Practical Application box for another
is commonly known as attrition or ortality If example of use of the Solomon four-group
the number or type of people who drop out of the design.)
experimental group differs from that of the control Although this design helps evaluate the effects
group, a mortality effect might explain the find- of testing, the threat of mortality remains a
ings. When you read such a work, examine the problem, as in the classic experimental design.
sample and the results carefully to see whether
dropouts or deaths occurred.
Testing e ects—the effects on the scores of a
posttest as the result of having taken a pretest—
Practical Application
Weinrich and colleagues (2007) used the
also can be a problem in these studies because the Solomon four-group design to test an
researcher is usually administering the same test enhanced decision aid for prostate cancer screening
twice, and participants tend to score better the versus standard education with middle-aged men.
second time just by learning the test. Researchers They hypothesized that participants who received the
pretest would have higher posttest knowledge. The
can circumvent this problem in one of two ways: men were first randomly assigned to one of four
They might use different forms of the same test groups: experimental and control groups that
for the two measurements, or they might use a received both pretest and posttest and the other
experimental and control groups that received only
more complex experimental design called the the posttest. They also tested and found no
Solomon four-group design. differences across the four groups in demographics,
The Solo on our grou design shown in family history of prostate cancer, or previous history
of screening prostate examinations, thus ensuring the
Figure 10-1, consists of two groups that are iden- success of the randomization process. The findings
tical to those used in the classic experimental revealed that outcomes varied depending on group
design plus two additional groups: an experimen- assignment (those who took both pretest and
tal after-group and a control after-group. As the posttest had significantly higher scores than did the
posttest–only intervention group) but also depending
diagram shows, all four groups have randomly on whether the men had had previous digital rectal
assigned R participants, as in all experimental examinations.
studies. However, the addition of these latter two
groups helps rule out testing threats to internal
validity that the before- and after-groups may A less frequently used experimental design is
experience. For example, suppose a researcher is the a ter only design shown in Figure 10-1 This
interested in the effects of counselling on the self- design, which is sometimes called the osttest
esteem of patients with chronic illness. Just taking only control grou design is composed of two
a test of self-esteem may in uence how the randomly assigned groups R , but in contrast to
participants report themselves. The items might the true experimental design, neither group is
 CHAPTER 10 Experimental and Quasiexperimental Designs 223

given a pretest or other measures. Again, the an important element in the conduct of experi-
independent variable is introduced to the experi- ments, studies conducted in the field are subject
mental group and not to the control group. to treatment contamination by factors specific to
The process of randomly assigning the partici- the setting that the researcher cannot control.
pants to groups is assumed to be sufficient to However, studies conducted in the laboratory are
ensure a lack of bias so that the researcher can by nature artificial because the setting is created
still determine whether the treatment created for the purpose of research. Thus, laboratory
significant differences between the two groups experiments, although stronger with regard to
and This design is particularly useful internal validity questions than field studies, have
when testing effects are expected to be a major more problems with external validity. For
problem and the number of available participants example, a participant’s behaviour in the labora-
is too limited for a Solomon four-group design. tory may be quite different from the person’s
An example of this design would be a study of behaviour in the real world this dichotomy pres-
an intervention on postoperative pain manage- ents problems in generalizing findings from the
ment, inasmuch as pain cannot be measured laboratory to the real world. Therefore, when you
before surgery and only an after-only design is read research reports, you need to consider the
required. possible effect of the experiment’s setting on the
findings of the study.
Helpful Hint Consider a hypothetical study on different
Remember that mortality is a problem in most types of wound treatment gels and creams for the
experimental studies because data are usually collected management of pressure ulcers. This study could
more than once. The researcher should demonstrate be performed in a laboratory with animals, which
that the groups are equivalent both when they enter the
study and at the final analysis.
would have allowed complete control over the
external environment of the study—a variable that
might be important in studying wound healing.
Field and Laboratory Experiments However, researchers cannot guarantee that the
Experiments also can be classified by setting. results found in a study in a laboratory would be
Field experiments and laboratory experiments applicable to human patients in hospital settings
share the properties of control, randomization, thus, some external validity would be lost.
and manipulation and involve the same design
characteristics but are conducted in different Advantages and Disadvantages of
environments. Laboratory experiments take place the Experimental Design
in an artificial setting created specifically for the As previously discussed, experimental designs
purpose of research. In the laboratory, the are the most appropriate for testing cause-and-
researcher has almost total control over the fea- effect relationships because the design enables
tures of the environment, such as temperature, the researcher to control the experimental situa-
humidity, noise level, and participant conditions. tion. Therefore, experimental designs offer better
Conversely, field experiments are exactly what corroboration than if the independent variable is
the name implies: Experiments that take place in manipulated in such a way that certain conse-
a real, preexisting social setting, such as a hospi- quences can be expected. Such studies are impor-
tal or clinic, where the phenomenon of interest tant because one of nursing’s major research
usually occurs. priorities is documenting outcomes to provide a
Because most experiments in the nursing lit- basis for changing or supporting current nursing
erature are field experiments and control is such practice.
224 PART THREE Quantitative Research

Experimental designs are not commonly used Another problem with experimental designs is
in nursing research, for several reasons. First, that they may be difficult or impractical to perform
experimentation is conducted under the assump- in field settings. It may be quite difficult to ran-
tion that all the relevant variables involved in a domly assign patients on a hospital oor to dif-
phenomenon have been identified. For many ferent groups when they might talk to each other
areas of nursing research, this is simply not the about the different treatments. Experimental
case, and descriptive studies need to be com- procedures also may be disruptive to the usual
pleted before experimental interventions can be routine of the setting. If several nurses are
applied. Second, these designs have some signifi- involved in administering the experimental
cant disadvantages. One problem with an experi- program, it may be impossible to ensure that the
mental design is that many variables important in program is administered in the same way to each
predicting outcomes of nursing care are not ame- participant.
nable to experimental manipulation. It is well Because of these problems in carrying out true
known that health status varies with age and experiments, researchers frequently turn to
socioeconomic status. No matter how careful a another type of research design to evaluate cause-
researcher is, no one can assign participants ran- and-effect relationships. Such designs, because
domly by age or a certain level of income. In they seem experimental but lack some of the
addition, it may be technically possible to manip- control of the true experimental design, are called
ulate some variables, but their nature may pre- uasie periments
clude their actually manipulation.
For example, if a researcher tried to randomly QUASIEXPERIMENTAL DESIGNS
assign groups to study the effects of cigarette uasiexperimental designs are intended to test
smoking and asked the experimental group to cause-and-effect relationships however, in a
smoke two packs of cigarettes a day, that research- quasiexperimental design, full experimental
er’s ethics would be seriously questioned. It is control is not possible. A uasie eri ent is a
also potentially true that such a study would not research design in which the researcher initiates
work because nonsmokers randomly assigned to an experimental treatment, but some characteris-
the smoking group would be unlikely to comply tic of a true experiment is lacking. Control may
with the research task. Thus, sometimes even not be possible because of the nature of the inde-
when a researcher plans to conduct a true experi- pendent variable or the nature of the available
ment, participants dropping out of the study or participants. uasiexperimental designs usually
other factors may, in effect, make the study a lack the element of randomization, as described
quasiexperiment. earlier with McGilton and associates’ (2003)
uasiexperimental designs are considered study. In other cases, the control group may be
when it is not possible to randomly assign partici- missing. However, like experiments, quasiexperi-
pants or when a control group is lacking. For ments involve the introduction of an experimental
example, McGilton and associates (2003) found treatment.
that health care professionals in long-term care In comparison with the true experimental
facilities can be taught how to enhance their design, quasiexperimental designs are used simi-
care without adding staff. To conduct the study, larly. Both types of designs are used when the
randomly assigning residents to nursing staff researcher is interested in testing cause-and-effect
was not feasible therefore, two separate units relationships. However, the basic problem with
were used for the control condition and the the quasiexperimental approach is a weakened
intervention. confidence in making causal assertions. Because
 CHAPTER 10 Experimental and Quasiexperimental Designs 225

of the lack of some controls in the research situ- except that participants are not randomly assigned
ation, quasiexperimental designs are subject to to groups.
contamination by many, if not all, of the threats For example, suppose a researcher is interested
to internal validity discussed in Chapter 9. in the effects of a new diabetes education program
on the physical and psychosocial outcomes of
Types of Quasiexperimental Designs patients with newly diagnosed diabetes. If the
Many different quasiexperimental designs exist. conditions were right, the researcher might be
Only the ones most commonly used in nursing able to randomly assign participants to either the
research are discussed in this book. To illustrate, group receiving the new program or the group
the symbols and notations introduced earlier in receiving the usual program, but for any number
the chapter are used. Refer to the true experimen- of reasons, that design might not be possible (e.g.,
tal design shown in Figure 10-1 and compare it nurses on the unit where patients are admitted
with the none uivalent control grou design might be so excited about the new program that
shown in Figure 10-2. Note that the latter design they cannot help but include the new information
looks exactly like that of the true experiment for all patients). Thus, the researcher has two

A. Nonequivalent control group design

O1 X O2
Experimental Pretest Experimental treatment Posttest
group

O1 O2
Control group Pretest Posttest

B. After-only nonequivalent control group design

X O2
Experimental Experimental treatment Posttest
group

O2
Control group Posttest

C. One group (pretest–posttest) design

O1 X O2
Experimental Pretest Experimental treatment Posttest
group

D. Time series design

O1 O1 X O2 O2
Experimental Pretest Pretest Experimental treatment Posttest Posttest
group
FIGURE 10-2 Comparison of quasiexperimental designs.
226 PART THREE Quantitative Research

choices: to abandon the experiment or to conduct 81 caregivers participated in the study. The
a quasiexperiment. To conduct a quasiexperi- quasiexperimental design used to test the effects
ment, the researcher might find a similar unit that of the intervention showed significant effects on
has not been introduced to the new program and perceived challenge associated with caregiver
study the patients with newly diagnosed diabetes role, control by self, use of social support, and
who are admitted to that unit as a comparison use of problem solving. In the measures taken at
group. The study would then involve the quasiex- the outset of the study, the researcher could
perimental type of design. include a measure of motivation to learn. The
Studies in which both quantitative and qualita- differences between the two groups on this vari-
tive methods are used are called mi ed-methods able could then be tested, and if significant dif-
studies ferences existed, they could be controlled
statistically in the analysis. Nonetheless, the
Helpful Hint strength of the causal assertions that can be made
Remember that researchers often make trade-offs on the basis of such designs depends on the ability
and sometimes use a quasiexperimental design instead of the researcher to identify and measure or
of an experimental design because it may be pragmati- control possible threats to internal validity.
cally impossible to randomly assign participants to
groups. The fact that the design is not “pure” does not Suppose that the researcher did not measure
decrease the value of the study, although the utility of the participants’ responses before the introduc-
the findings may be decreased. tion of the new treatment (or the researcher was
hired after the new program began) but later
The nonequivalent control group design is com- decided that data demonstrating the effect of the
monly used in nursing research studies conducted program would be useful. Perhaps, for example,
in field settings. The basic problem with the design a third party asks for such data to determine
is the weakening of the researcher’s confidence in whether it should pay the extra cost of the new
assuming that the experimental and comparison teaching program. Sometimes the outcomes
groups are similar at the beginning of the study. simply cannot be measured before the interven-
Threats to internal validity, such as selection bias, tion, as with prenatal interventions that are
maturation effects, testing effects, and mortality expected to affect birth outcomes. The study that
(attrition), are possible with this design. However, could be conducted would have an a ter only
the design is relatively strong because the gather- none uivalent control grou design illustrated
ing of the data at the time of the pretest allows in Figure 10-2. This design is similar to the after-
the researcher to compare the equivalence of the only experimental design, but randomization is
two groups on important antecedent variables not used to assign participants to groups. In this
before the independent variable is introduced. design, the two groups are assumed to be equiva-
In Ducharme and colleagues’ (2006) study, the lent and comparable before the introduction of
motivation of caregivers to learn about stress the independent variable Thus, the sound-
management might be important in determining ness of the design and the confidence that the
the effect of this program. The purpose of this researchers can have in the findings depend on
project was to evaluate the implementation and the soundness of this assumption of preinterven-
effects of a stress management intervention for tion comparability. Often, the assumption that
family caregivers of older adults. The interven- the two nonrandomly assigned groups are com-
tion was implemented through an action research parable at the outset of the study is difficult to
design with the collaboration of case managers assert because the validity of the statement
working in community health centers. A total of cannot be assessed.
 CHAPTER 10 Experimental and Quasiexperimental Designs 227

In the example of the teaching program for

patients with newly diagnosed diabetes, measur- Practical Application
ing the participants’ motivation after the teaching Martinez and associates (2007) conducted a
one-group pretest–posttest design study to
program would not reveal whether their motiva-
explore nurses’ perceptions of the effect of the
tions differed before they received the program, 15-minute family interview on the paediatric hospital
and it is possible that the teaching program would admission process. The intervention consisted of two
motivate individuals to learn more about their in-depth teaching sessions and hands-on coaching in
the use of the 15-minute family interview. Each nurse
health problem. Therefore, the researcher’s con- acted as his or her own control by being interviewed
clusion that the teaching program improved phys- twice, before receiving the intervention and after
ical status and psychosocial outcome would be completing six family interviews. Findings revealed
that the nurses perceived the genogram, therapeutic
subject to the alternative conclusion that the questions, and commendations as having a positive
results were an effect of preexisting motivations effect on their ability to conduct family assessment
(selection effect) in combination with greater and family interventions.
learning by participants so motivated (selection-
maturation interaction). Nonetheless, this design quasiexperimental design is called a ti e series
is frequently used in nursing research because design and is illustrated in Figure 10-2. Time
opportunities for data collection are often limited series designs are useful for determining trends,
and because this design is particularly useful as in a study of the effect of adult day programs
when testing effects may be problematic. on family caregivers of elderly relatives (Warren,
Tranmer and Parry (2004) used an after-only Ross-Kerr, Smith, Godkin, 2003). The caregiv-
design to test the effects of an advanced practice ers completed measures of burden, quality of life,
nursing intervention on health-related quality of perceived health, and opinion about institutional-
life, symptom distress, and satisfaction with care ization at four time points (just before patient
among cardiac surgery patients. The intervention admission and 2 weeks, 2 months, and 6 months
was telephone support for 4 weeks after hospital after admission to the day program). Sometimes
discharge. data are collected many times before the intro-
Another quasiexperimental design is a one duction of the treatment to establish a baseline
grou retest osttest design (Figure 10-2), point of reference on outcomes. The experimental
which is used by researchers when only one group treatment is then introduced, and data are col-
is available for study (see Practical Application lected multiple times afterward to determine a
box). Data are collected before and after an exper- change from baseline. The broad range and
imental treatment on this one group of partici- number of data-collection points help rule out
pants. In this type of design, the participants alternative explanations, such as history effects.
act as their own controls, and no randomization However, a testing threat to internal validity is
occurs. Because controls and randomization are ever present because of multiple data-collection
important characteristics that enhance the internal points, and without a control group, the threats of
validity of the study, the evidence generated by selection bias and maturation effects cannot be
the findings of this type of quasiexperimental ruled out (see Chapter 9).
design needs to be interpreted with careful con- To rule out some alternative explanations for
sideration of the design limitations. the findings of a one-group pretest–posttest
An approach used by researchers when only design, researchers typically measure the phe-
one group is available is to study that group over nomenon of interest over a longer period and
a longer period: that is, to test participants before introduce the experimental treatment sometime
an intervention and again afterwards. This during the course of the data-collection period.
228 PART THREE Quantitative Research

Even with the absence of a control group, the the population can suggest that a particular
broader range of data-collection points helps rule explanation is not plausible. Nonetheless, repli-
out threats to validity such as history effects. cating such studies is important to support the
Obviously, the earlier example of teaching causal assertions developed through the use of
patients with diabetes does not lend itself to this quasiexperimental designs.
design because researchers do not have access to The literature on cigarette smoking is an excel-
the patients before the diagnosis. lent example of how findings from many studies,
experimental and quasiexperimental, can be
Helpful Hint linked to establish a causal relationship. A large
One of the reasons replication is so important in number of well-controlled experiments with lab-
nursing research is that many problems cannot be sub- oratory animals randomly assigned to smoking
jected to experimental methods. Therefore, the consis-
tency of findings across many patient populations helps
and nonsmoking conditions have documented
support a cause-and-effect relationship even when an that lung disease does develop in smoking
experiment cannot be conducted. animals. Although such evidence is suggestive of
a link between smoking and lung disease in
humans, it is not directly transferable because
Advantages and Disadvantages of animals and humans are different. Because
Quasiexperimental Designs humans cannot be randomly assigned to smoking
Because of the problems inherent in interpreting and nonsmoking groups for ethical and other
the results of studies with quasiexperimental reasons, researchers interested in this problem
designs, you may wonder why anyone would use must use quasiexperimental data to test their
them. uasiexperimental designs are used fre- hypotheses about smoking and lung disease.
quently because they are practical and feasible, Several different quasiexperimental designs
and the results are generalizable. These designs have been used to study this problem, and all have
are more adaptable to the real-world practice yielded similar results: A causal relationship does
setting than controlled experimental designs. In exist between cigarette smoking and lung disease.
addition, for some hypotheses, these designs may Note that the combination of results from both
be the only way to evaluate the effect of the inde- experimental and quasiexperimental studies led
pendent variable of interest. to the conclusion that smoking causes lung
The weaknesses of the quasiexperimental disease because the studies together meet the
approach involve mainly the inability to establish causal criteria of relationship, timing, and lack of
clear cause-and-effect relationships. However, an alternative explanation.
if the researcher can rule out any plausible The tobacco industry has argued that because
alternative explanations for the findings, such the studies on humans are not true experiments,
studies can lead to increased knowledge about another explanation is possible for the relation-
causal relationships. Researchers have several ships that have been found. For example, these
options for ruling out these alternative explana- relationships suggest that the tendency to smoke
tions. They may control extraneous variables a is linked to the tendency for lung disease to
riori (before initiating the intervention) by develop, and smoking is merely an unimportant
design. intervening variable. The reader needs to review
Researchers can also use methods to control the evidence from studies to determine whether
extraneous variables statistically. In some cases, the cause-and-effect relationship postulated is
common sense knowledge of the problem and believable.
 CHAPTER 10 Experimental and Quasiexperimental Designs 229

design. Both experimental and quasiexperimental

Evidence-Informed Practice Tip
designs (as well as nonexperimental designs) are
Findings from studies with experimental designs
are considered level II evidence, and those from studies used to determine the effect or outcomes of a
with quasiexperimental designs are considered level III program. When these designs are used in evaluat-
evidence. Quasiexperimental designs are lower on the ing a program, the term evaluation research is
hierarchy of evidence because of a lack of a research
control, which limits the ability to establish confident
used. Bigman (1961) listed the following pur-
cause-and-effect statements that influence clinical deci- poses and uses of evaluation research:
sion making. 1. To discover whether and how well the
objectives are being fulfilled
2. To determine the reasons for specific suc-
EVALUATION RESEARCH AND cesses and failures
EXPERIMENTATION 3. To direct the course of the experiment with
As the science of nursing expands and the cost of techniques for its effectiveness
health care rises, nurses and other health care 4. To reveal principles that underlie a success-
professionals have become increasingly con- ful program
cerned with the ability to document the costs and 5. To base further research on the reasons
the benefits of nursing care (see Chapter 1). This for the relative success of alternative
task is a complex process, but at its heart is the techniques
ability to evaluate or measure the outcomes of 6. To redefine the means to be used for attain-
nursing care to inform health care decision ing objectives and to redefine subgoals in
making. Such studies usually are associated with view of research findings
quality assurance, quality improvement, and According to Clarke (2001), the following four
evaluation. Studies of evaluation or quality assur- levels of evaluation research are being high-
ance do exactly what the name implies: They are lighted in health care, especially nursing research:
concerned with the determination of the quality evaluation of the effectiveness of clinical inter-
of nursing and health care and with the assurance ventions evaluation of the effect of new ways of
that the public is receiving high-quality care. delivering health care evaluation of structured
uality assurance and quality improvement in programs aimed at specific patient groups and
nursing are current and important topics for evaluation of the quality of service. In many eval-
nursing care. Many early studies of quality assur- uation research studies, investigators use mixed
ance documented whether nursing care met pre- methods, with both quantitative and qualitative
determined standards. The goal of quality information.
improvement studies is to evaluate the effective- Evaluation studies may be either formative or
ness of nursing interventions and to provide summative. In or ative evaluation a program
direction for further improvement in the achieve- is assessed as it is being implemented usually,
ment of quality clinical outcomes and cost the focus is on evaluation of the process of a
effectiveness. program rather than the outcomes. In su ative
Evaluation research is the use of scientific evaluation the outcomes of a program are
research methods and procedures to evaluate a assessed after completion of the initial program.
program, treatment, practice, or policy. In evalu- Kuehn and colleagues (2011) used a summa-
ation research, analytical means are used to docu- tive evaluation to assess the effect of a North
ment the worth of an activity such as an American nursing exchange program on student
intervention, but such research is not a different cultural awareness. In contrast, Jurasek and
230 PART THREE Quantitative Research

associates (2010) used formative evaluation to evaluation enables researchers to determine not
describe the collaborative development of an ado- only whether care is adequate but also which
lescent epilepsy transition clinic (i.e., for transi- method of care is best under certain conditions.
tion to adult care). They conducted a process Furthermore, such studies can be used to deter-
evaluation (1) to gather information on whether mine whether a particular type of nursing care or
the learning needs of the adolescents were met, intervention is cost effective: that is, that the care
(2) to demonstrate decreased fear associated with or intervention does what it is intended to do but
the transition, (3) to prepare parents for the expec- at lower or equivalent cost. Cost studies are
tations and differences of the adult program, and usually incorporated into the evaluation of an
(4) to determine whether nurses were appropriate intervention. For example, Steel-O’Connor and
program leaders. Data were collected at the end associates (2003) evaluated the effects of pro-
of transition and 2 to 3 months into adult transi- grams of follow-up care by public health nurses
tion. Knowledge related to summative (outcomes) in terms of infant health problems, breast-feeding
and formative (process) evaluation of programs rates, and the use of postpartum health services
is important in translating research into clinical in Ontario. The authors compared the costs asso-
practice. ciated with routine home visiting by a public
The use of experimental and quasiexperimen- health nurse after early obstetrical discharge with
tal designs in studies of quality improvement and the costs associated with a screening telephone

APPRAISING THE EVIDENCE

Experimental and Quasiexperimental Designs
As discussed earlier in the chapter, various designs for describing the design of the study appears in the abstract
research studies differ in the amount of control the and in the “Methods” section of the article. If such a state-
researcher has over the antecedent and intervening vari- ment is not present, the reader should examine the study
ables that may affect the results of the study. True experi- for evidence of the following three characteristics: control,
mental designs, which yield level II evidence, offer the randomization, and manipulation. If all are discussed, the
most possibility for control, whereas nonexperimental design is probably experimental. Conversely, if the study
designs, which yield level IV, V, or VI evidence, offer the involves the administration of an experimental treatment
least. Quasiexperimental designs, which yield level III but does not involve the random assignment of partici-
evidence, offer somewhere in between. Research designs pants to groups, the design is quasiexperimental. Next,
must balance the needs for internal validity and external try to identify which of the variations within these two
validity in order to produce useful results. In addition, types of designs was used. Determining the answer to
judicious use of design requires that the chosen design these questions gives you a head start because inherent
be appropriate to the problem, free of bias, and capable in each design are particular threats to validity, and this
of answering the research question. step makes it easier to critically evaluate the study. The
Questions that you should pose when reading studies next question to ask is whether the researcher required a
that test cause-and-effect relationships are listed in the solution to a cause-and-effect problem. If so, the study is
Critiquing Criteria box. All of these questions should help suited to these designs. Finally, think about the conduct
you judge, with confidence, whether a causal relationship of the study in the setting. Is it realistic to think that the
exists. study could be conducted in a clinical setting without
For studies in which either experimental or quasiex- some contamination?
perimental designs are used, first try to determine the The most important question to ask as you
type of design that was used. Often, a statement read experimental studies is “What else could have
 CHAPTER 10 Experimental and Quasiexperimental Designs 231

APPRAISING THE EVIDENCE

Experimental and Quasiexperimental Designs—cont’d
happened to explain the findings?” Thus, the author control for many plausible alternative explanations. A
must provide adequate accounts of how the procedures well-written report of a quasiexperimental study system-
for randomization, control, and manipulation were atically reviews potential threats to the validity of the
carried out. The study should include a description of findings. Then your work as the reader is to decide
the procedures for random assignment to such a degree whether the author’s explanations make sense. When
that the reader can determine the likelihood for any critiquing evaluation research, you should look for a
one participant to be assigned to a particular group. The careful description of the program, policy, procedure, or
description of the independent variable also should treatment being evaluated. In addition, you may need to
be detailed. The inclusion of this information helps the determine the design used to evaluate the program and
reader decide whether the treatment given to some par- assess the appropriateness of the design for the evalua-
ticipants in the experimental group might differ from tion. Once you have discerned the design, you assess
what was given to others in the same group. In addition, threats to validity for the appropriate design in determin-
threats to validity, such as testing effects and mortality ing the appropriateness of the author’s conclusions in
(attrition), should be addressed. Otherwise, the conclu- relation to the outcomes. As with all research, the results
sions of the study could potentially be erroneous and less of studies with these designs need to be generalizable to
believable to the reader. a larger population of people than was actually studied.
This question of potential alternative explanations or Thus, researchers need to decide whether the experimen-
threats to internal validity for the findings is even more tal protocol eliminated some potential participants and
important when you critically evaluate a quasiexperimen- whether this weakness affected not only internal validity
tal study because these study designs cannot possibly but also external validity.

CRITIQUING CRITERIA

1. What design is used in the 4. Are all threats to validity, 6. Do other limitations related to
study? including mortality (attrition), the design exist that are not
2. Is the design experimental or addressed in the report? mentioned?
quasiexperimental? 5. Whether the experiment was
3. Is the problem one of a conducted in the laboratory or a EVALUATION RESEARCH
cause-and-effect clinical setting, are the findings 1. Do the authors identify a specific
relationship? generalizable to the larger problem, practice, policy, or
4. Is the method used appropriate population of interest? treatment that they will
to the problem? evaluate?
5. Is the design suited to the setting QUASIEXPERIMENTAL DESIGNS 2. Do the authors identify the
of the study? 1. What quasiexperimental design outcomes to be evaluated?
is used in the study, and is it 3. Is the problem analyzed and
EXPERIMENTAL DESIGNS appropriate? described?
1. What experimental design is 2. What are the most common 4. Is the program to be analyzed
used in the study, and is it threats to the validity of the described and standardized?
appropriate? findings of this design? 5. Do the authors identify the
2. How are randomization, 3. What are the plausible measurement of the degree of
control, and manipulation alternative explanations, and change (outcome) that occurs?
applied? have they been addressed? 6. Do the authors determine
3. Are there reasons to believe that 4. Are the author’s explanations of whether the observed outcome
alternative explanations exist for threats to validity acceptable? is related to the activity or to
the findings? 5. What does the author say about one or more other causes?
the limitations of the study?
232 PART THREE Quantitative Research

call. Steel-O’Connor and associates found that

the direct, indirect, and total costs of health ser- KEY POINTS
vices per 100 infants were greater with home
visits and concluded that a routine home visit is • Experimental designs or randomized clinical trials
provide the strongest evidence (level II) in terms of
not always necessary to identify women who whether an intervention or treatment affects
need that service. In an era of health care reform patient outcomes.
and cost containment for health expenditures, • Two types of design commonly used in nursing
research to test hypotheses about cause-and-effect
evaluating the relative costs and benefits of new relationships are experimental and
programs of care is increasingly important. Rela- quasiexperimental designs. Both are useful for the
tively few studies in nursing and medicine have development of nursing knowledge because they
been dedicated to such evaluation, but in terms of test the effects of nursing actions and lead to the
development of prescriptive theory.
outcomes, nursing costs and cost savings are • True experiments are characterized by the ability of
important in future studies. the researcher to control extraneous variation,
manipulate the independent variable, and
randomly assign participants to research groups.
Helpful Hint • Experiments conducted in clinical settings or in the
Think of quality assurance and quality improve- laboratory provide the best evidence in support of
ment projects as research-related activities that enhance a causal relationship because the following three
the ability of nurses to generate cost and quality criteria can be met: (1) the independent and
outcome data. These outcome data contribute to docu- dependent variables are related to each other; (2)
menting the effectiveness of nursing practice. the independent variable chronologically precedes
the dependent variable; and (3) the relationship
cannot be explained by the presence of a third
variable.
• Researchers frequently use quasiexperimental
CRITICAL THINKING CHALLENGES designs to test cause-and-effect relationships
because experimental designs are often impractical
■ Discuss the barriers to nurse researchers in or unethical.
meeting the three criteria of a true experimental • Quasiexperiments may lack either randomization
or the comparison group, or both, which are
design. characteristics of true experiments. Their
■ How is it possible to have a research design that usefulness in studying causal relationships depends
includes an experimental treatment intervention on the ability of the researcher to rule out
and a control group and yet is not considered a plausible threats to the validity of the findings,
true experimental study? How does this affect such as history, selection bias, and maturation and
testing effects.
the usefulness of the findings in an evidence- • The level of evidence (level III) provided by
informed practice? quasiexperimental designs weakens confidence
■ Argue your case for supporting or not supporting that the findings were the result of the intervention
the following claim: “The fact that true rather than extraneous variables.
experimental design is not used does not • The overall purpose of critiquing such studies is to
assess the validity of the findings and to determine
decrease the value of the study, even though it whether these findings are worth incorporating into
may decrease the utility of the findings in the nurse’s personal practice.
practice.” Include examples with your rationale.
■ Respond to the following question: Why are
experimental studies considered the best
evidence for an evidence-informed practice
model? Justify your answer.
 CHAPTER 10 Experimental and Quasiexperimental Designs 233

REFERENCES catheter washouts extend patency time in long-term

indwelling urethral catheters A randomized con-
Bigman, S. K. (1961). Evaluating the effectiveness of trolled trial of acidic washout solution, normal saline
religious programs. Revie of Religious Research, , washout, or standard care. ournal of ound stomy
99-110. Continence Nursing, (1), 82-90.
Campbell, D., Stanley, J. (1966). E perimental and Ploeg, J., Brazil, K., Hutchison, B., Kaczorowski, J.,
uasie perimental designs for research. Chicago: Dalby, D. M., Goldsmith, C. H., Furlong, C.
Rand-McNally. (2010). Effect of preventive primary care outreach on
Clarke, A. (2001). Evaluation research in nursing and health related quality of life among older adults at
health care. Nurse Researcher, (3), 4-14. risk of functional decline: Randomised controlled
Ducharme, F., Lebel, P., Lachance, L., Trudeau, D. trial. British Medical ournal, , c1480.
(2006). Implementation and effects of an individual doi:101136.bmj.c1480
stress management intervention for family caregivers Sobieraj, G., Bhatt, M., LeMay, S., Rennick, J.,
of an elderly relative living at home: A mixed Johnston, C. (2009). The effect of music on parental
research design. Research in Nursing ealth, , participation during pediatric laceration repair.
427-441. Canadian ournal of Nursing Research, , 68-82.
Jurasek, L., Ray, L., uigley, D. (2010). Develop- Steel-O’Connor, K. O., Mowat, D. L., Scott, H. M.,
ment and implementation of an adolescent epilepsy Carr, P. A., Dorland, J. L., Young-Tai, K. F. (2003).
transition clinic. ournal of Neuroscience Nursing, A randomized trial of two public health nurse
(4), 181-189. follow-up programs after early discharge. Canadian
Kuehn, A. F., Chircop, A., Downe-Wamboldt, B., ournal of Public ealth, , 98-103.
Sheppard-LeMoine, D., Wittstock, L., Herbert, R., Tranmer, J. E., Parry, M. J. E. (2004). Enhancing
. . . Critchley, K. (2011). Evaluating the impact of a postoperative recovery of cardiac surgery patients: A
North American nursing exchange program on randomized clinical trial of an advanced practice
student cultural awareness. International ournal of nursing intervention. estern ournal of Nursing
Nursing Education Scholarship, (1), Article 4. Research, , 515-532.
LeMay, S., Johnston, C., Choini re, M., Fortin, C., Warren, S., Ross-Kerr, J., Smith, D., Godkin, D.
Hubert, I., Fr chette, G., . . . Murray, L. (2010). Pain (2003). The impact of adult day programs on family
management interventions with parents in the caregivers of elderly relatives. ournal of Community
emergency department: A randomized trial. ournal ealth Nursing, (4), 209-221.
of Advanced Nursing, , 2442-2449. Weinrich, S. P., Seger, R., Curtsinger, T., Pumphrey, G.,
Martinez, A. M., D’Artois, D., Rennick, J. E. (2007). NeSmith, E. G., Weinrich, M. (2007). Impact of
Does the 15-minute (or less) family interview pretest on posttest knowledge scores with a Solomon
in uence family nursing practice ournal of amily four research design. Cancer Nursing, , E16-E28.
Nursing, (2), 157-178.
McGilton, K. S., O’Brien-Pallas, L. L., Darlington, G.,
Evans, M., Wynn, F., Pringle, D. M. (2003).
FOR FURTHER STUDY
Effects of a relationship-enhancing program of care Go to Evolve at http://evolve.elsevier.com/
on outcomes. ournal of Nursing Scholarship, , Canada/LoBiondo/Research for Audio Glossary, how-to
151-156. instructions for Writing Proposals for Funding, and
Moore, K. N., Hunter, K. F., McGinnis, R., Bacsu, C., additional research articles for practice in reviewing and
Fader, M., Gray, M., . . . Voaklander, D. (2009). Do critiquing.
C H A PTER 1 1

Nonexperimental Designs
Geri LoBiondo-Wood | Judith Haber | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Describe the overall purpose of nonexperimental designs.
• Describe the characteristics of survey and relationship/difference designs.
• Define the differences between survey and relationship/difference designs.
• List the advantages and disadvantages of surveys and each type of relationship/difference design.
• Identify methodological, secondary analysis, and meta-analysis research.
• Identify the purposes of methodological, secondary analysis, and meta-analysis research.
• Discuss relational inferences versus causal inferences as they relate to nonexperimental designs.
• Identify the criteria used to critique nonexperimental research designs.
• Apply the critiquing criteria to the evaluation of nonexperimental research designs as they appear
in research reports.
• Apply levels of evidence to nonexperimental designs.

KEY TERMS
cohort incidence prospective study
correlational study longitudinal study psychometrics
cross-sectional study meta-analysis relationship/difference
descriptive/exploratory survey methodological research study
developmental study nonexperimental research retrospective data
epidemiological study design retrospective study
ex post facto study prediction study secondary analysis
hierarchical linear modelling prevalence survey study
(HLM)

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

234
 CHAPTER 11 Nonexperimental Designs 235

MANY PHENOMENA OF INTEREST AND RELEVANCE to tative research designs is shown in Figure 11-1.
nursing do not lend themselves to an experimen- This chapter divides nonexperimental designs
tal design. For example, nurses studying pain into survey studies and relationship difference
may be interested in knowing the amount of pain, studies, as illustrated in Box 11-1. These catego-
variations in the amount of pain, and patients’ ries are exible nonexperimental studies may be
responses to postoperative pain. The investigator classified in a different way in other sources.
would not design an experimental study that Some studies belong exclusively to one of these
would potentially intensify a patient’s pain just to categories, whereas other studies have the char-
study the pain experience that would be unethi- acteristics of more than one category or more than
cal. Instead, the researcher would use a non- one design label. As you read the research litera-
experimental design to examine the factors that ture, you will often find that researchers who are
contribute to the variability in a patient’s post- conducting a nonexperimental study use several
operative pain experience. Nonexperimental design classifications. This chapter introduces the
research designs are used in studies when the various types of nonexperimental designs, their
researcher wishes to construct a picture of a phe- advantages and disadvantages, the use of nonex-
nomenon examine events, people, or situations perimental research, the issues of causality, and
as they naturally occur or test relationships and the critiquing process as it relates to nonexperi-
differences among variables. Nonexperimental mental research. The Critical Thinking Decision
designs may enable the researcher to understand Path outlines the path to the choice of a nonex-
how a phenomenon occurs at one point or over a perimental design.
period of time.
In experimental research, the independent Evidence-Informed Practice Tip
variable is manipulated in a none eri ental When you critically appraise nonexperimental
research design the independent variable is not studies, be aware of possible sources of bias that can be
manipulated. In nonexperimental research, the introduced at any point in the study.
independent variables have occurred naturally,
and the investigator cannot directly control them
by manipulation. In contrast, in an experimental
BOX 11-1
design, the researcher actively manipulates one or
more variables. The researcher in a nonexperi- SUMMARY OF NONEXPERIMENTAL RESEARCH
DESIGNS
mental design explores relationships or differ-
ences among the variables. Nonexperimental SURVEY STUDIES
• Descriptive
research requires a clear, concise research problem • Exploratory
or hypothesis that is based on a theoretical frame- • Comparative
work. Even though the researcher does not
RELATIONSHIP/DIFFERENCE STUDIES
actively manipulate the variables, the concepts of
• Correlational
control (see Chapter 9) should be considered as • Developmental
much as possible. • Cross-sectional
Researchers do not agree on how to classify • Longitudinal or prospective
• Retrospective or ex post facto
nonexperimental studies. A continuum of quanti-

Nonexperimental Quasiexperimental Experimental

FIGURE 11-1 Continuum of quantitative research designs.

236 PART THREE Quantitative Research

CRITICAL THINKING DECISION PATH

Nonexperimental Design Choices

Assess the nature of the problem being studied

Little is known Test relationship Compare differences Need a

about variables between variables between variables measurement tool

Survey Correlational Methodological

Past Present Present Past Present Present

Descriptive
view view Future view view view Future view

Ex post Retro-
Exploratory Correlational Longitudinal Survey Longitudinal
facto spective

Retro- Cross- Ex post Cross-

Comparative sectional Prospective Prospective
spective facto sectional

Cross-
Correlational
sectional

little is known about the phenomenon. The data

SURVEY STUDIES are used to justify or assess current conditions or
The broadest category of nonexperimental designs to make plans for improvement of conditions.
is the survey study. In a survey study—further ualitative researchers also use the term descrip-
classified as descriptive, e ploratory, or tive in their reports, as in the study by Chojecki
comparative—detailed descriptions of existing and colleagues (2010), who used a qualitative/
variables are collected, and the data are used to descriptive design to explore the perceptions of a
justify and assess current conditions and practices peer learning approach to paediatric clinical edu-
or to make more plans for improving health care cation. You will be able to determine the differ-
practices. When you read research, you will find ence by checking in the analysis and findings
that the terms e ploratory, descriptive, compara- sections, inasmuch as the qualitative descriptive
tive, and survey may be used either alone, inter- study entails the use of the analyses outlined in
changeably, or together to describe the design of Chapter 16, whereas the descriptive correlational
a study (see Table 11-1). For example, investiga- or survey studies entail the use of descriptive and
tors may use a descri tive e loratory survey inferential statistical analyses.
to search for accurate information about the char- In survey studies, the types of variables of
acteristics of particular participants, groups, insti- interest can be classified as opinions, attitudes, or
tutions, or situations or about the frequency of a facts. For example, Sawhney and Sawyer (2008)
phenomenon’s occurrence, particularly when conducted a cross-sectional study in which they
 CHAPTER 11 Nonexperimental Designs 237

used e-mailed surveys to explore and describe the example, if a hospital were contemplating con-
role of nurses working in a specialty practice of verting all patient care units to a case-management
pain management. The results of this study build model, a survey might be conducted to determine
on previous work in which the role of nurses in the attitudes of a representative sample of nurses
pain management was examined. Respondents to in the hospital toward case management. The data
the survey reported several benefits and chal- might provide the basis for projecting the
lenges to the role of these nurses. In another in-service needs of nursing with regard to case
example, to explore nurses’ perceptions of and management. The scope and depth of a survey are
level of satisfaction with the system of medica- a function of the nature of the problem.
tion administration in long-term care, surveys In surveys, investigators attempt only to relate
were given to 301 licensed nurses (Kaasalainen, one variable to another or to assess differences
Agarwal, Dolovich, Papaioannou, Brazil, Akhtar- between variables they do not attempt to deter-
Danash, 2010). mine causation. The two major advantages of
Fact variables include attributes of individuals surveys are the great deal of information that can
such as gender, income level, political and reli- be obtained from a large population in a more
gious affiliations, ethnicity, occupation, and edu- economical manner than face-to-face interviews
cational level. Koren and associates (2010) used and the surprising accuracy of survey research
the data from the annual tracking surveys of nurse information. If a sample is representative of the
practitioners in Ontario conducted by the Centre population (see Chapter 12), a relatively small
for Rural and Northern Health Research for the number of participants can accurately represent
Ministry of Health and Long-term Care to the views of the population.
describe current employment and practice. However, survey studies have several disad-
Data in survey research can be collected vantages. First, the information obtained in a
through a questionnaire or an interview (see survey tends to be superficial. The breadth rather
Chapter 13). For example, Yonge and colleagues than the depth of the information is emphasized.
(2010) administered a Web-based questionnaire Second, conducting a survey requires a great deal
to examine the willingness of nursing students to of expertise in various research areas. The survey
volunteer during a pandemic. Another example is
the study by Hoe Harwood and associates (2010),
who used an online national survey to describe Helpful Hint
nontraditional, innovative clinical placements Research consumers should recognize that a well-
within Canadian nursing education programs constructed survey can provide a wealth of data about
a particular phenomenon of interest, even though cau-
from the perspectives of clinical placement coor- sation is not being examined.
dinators and nurse educators.
Survey researchers study either small or large
samples of participants recruited from defined Evidence-Informed Practice Tip
populations. The sample can be either broad or Evidence obtained from a survey population may
narrow and can be made up of people or institu- be coupled with clinical expertise and applied to a
tions. For example, if a primary care rehabilita- similar population to develop an educational program
tion unit based on a case-management model to enhance knowledge and skills in a particular clinical
area. For example, a survey designed to measure
were to be established in a hospital, researcher nursing staff’s knowledge and attitudes about evidence-
might survey prospective applicants’ attitudes informed practice may yield data that are used to
with regard to case management before the unit develop a staff development course in evidence-
informed practice.
staff members were selected. In a broader
238 PART THREE Quantitative Research

investigator must have skills in sampling tech- workplace empowerment. They found that total
niques, questionnaire construction, interviewing, empowerment was most strongly related to man-
and data analysis to elicit reliable and valid data. agement leadership ability and nurse participa-
Third, large-scale surveys can be time consuming tion in hospital affairs and least strongly to
and costly, although the use of on-site personnel collaborative nurse-physician relationships.
can reduce costs. Correlational studies offer researchers and
research consumers the following advantages:
RELATIONSHIP/DIFFERENCE STUDIES An increased exibility when investigating
Investigators endeavour to trace the relationships complex relationships among variables
or differences between variables that can provide An efficient and effective method of collect-
a deeper insight into a phenomenon. This type ing a large amount of data about a problem
of study can be classified as a relationshi A potential for practical application in clini-
di erence study The following types of cal settings
relationship/difference studies are discussed: cor- A potential foundation for future experi-
relational studies and developmental studies mental research studies
A framework for exploring the relationship
Correlational Studies between variables that cannot be inherently
In a correlational study an investigator exam- manipulated
ines the relationship between two or more vari- The correlational design has a quality of realism
ables. The researcher is not testing whether one and is particularly appealing because it suggests
variable causes another variable or how different the potential for practical solutions to clinical
one variable is from another variable. Instead, the problems. The following, however, are the disad-
researcher is testing whether the variables covary vantages of correlational studies:
in other words, as one variable changes, does a Inability to manipulate the variables of in-
related change occur in the other variable The terest
researcher using this design is interested in quan- No randomization in the sampling proce-
tifying the strength of the relationship between dures, because the study deals with preex-
the variables or in testing a hypothesis about a isting groups therefore, generalizability is
specific relationship. The positive or negative decreased
direction of the relationship is also a central Inability to determine a causal relation-
concern (see Chapter 16 for an explanation of the ship between the variables because of the
correlation between variables). lack of manipulation, control, and random-
In their correlational study, Bailey and col- ization
leagues (2010) explored family member percep- One of the most common misuses of a correla-
tion of informational support, anxiety, satisfaction tional design is the researcher’s conclusion that a
with care, and their interrelationships to guide causal relationship exists between the variables.
further refinement of a local informational support In their study, Bailey and colleagues (2010)
initiative. These researchers were not testing a appropriately concluded that informational sup-
cause-and-effect relationship. port was positively correlated significantly with
Another example of correlational research satisfaction with care. The report concluded with
was the study by Armstrong and associates some very thoughtful recommendations for future
(2009), who tested a theoretical model derived studies in this area by stating that representa-
from Rosabeth Moss Kanter’s theory of tiveness could be improved through probability
 CHAPTER 11 Nonexperimental Designs 239

sampling and sample size if a quasiexperimental TABLE 11-1

study was to be conducted. EXAMPLES OF STUDIES WITH MORE THAN ONE
Correlational studies may be further labelled DESIGN LABEL
descriptive correlational or predictive correla- DESIGN TYPE STUDY’S PURPOSE
tional A study by Plunkett and colleagues (2010) Descriptive with To determine whether there are
is an example of a predictive correlational study. repeated psychosocial and physical benefits
The goal was to predict whether the intention of measures of a self-awareness intervention
for adults with type 1 diabetes
generic BScN students to further pursue graduate and hypoglycaemia (Hernandez,
studies is in uenced by their valuation of gradu- Hume, & Rodger, 2008)
ate studies and perceived self-efficacy for gradu- Descriptive, To examine the associations
ate studies. The researchers concluded that several correlational between depression and
work-related variables such as job
of the variables were predictive of (did not cause) strain, role overload, respect,
intention to pursue graduate studies. social and employer supports,
The inability to draw causal statements should and nurses’ perception of the
quality of care they provide
not lead you to conclude that a nonexperimental (Ohler, Kerr, & Forbes, 2010)
correlational study has a weak design. In terms of Cross-sectional, To describe family member
evidence for practice, researchers—on the basis descriptive, perception of informational
correlational support, anxiety, satisfaction with
of the literature review and their findings—frame care, and their relationships, in
the utility of the results in view of previous order to guide further refinement
research and therefore help establish supportive of a local informational support
evidence of the applicability of the results to a initiative and its eventual
evaluation (Bailey, Sabbagh,
specific patient population. A correlational design Loiselle, Boileau, & McVey, 2010)
is very useful for clinical research studies because Explanatory, To determine the stressors,
many of the phenomena of clinical interest are correlational academic performance, and
learned resourcefulness in
beyond the researcher’s ability to manipulate, baccalaureate nursing students
control, and randomize. (Goff, 2011)
Retrospective, To explore experiences of refugee
Developmental Studies cross-sectional and refugee claimant mothers
with an ill preschooler on
Nonexperimental designs in which a time per- equitable access to health care
spective is used can be further subclassified. A (Wahoush, 2009)
Prospective, To assess for anemia and red blood
develo ental study is concerned not only with
cohort cell transfusion practices in 100
the existing status and the relationship and differ- consecutive adults admitted to a
ences among phenomena at one point in time but general systems intensive care
also with changes that occur as a function of time. unit (Thomas, Jensen, Nahirniak,
& Gibney, 2010)
The following three types of developmental study
designs are discussed here: cross-sectional, lon-
gitudinal or prospective, and retrospective or e
post facto Remember that in the literature, studies Cross-Sectional Studies
may be designated by more than one design In a cross sectional study researchers examine
name. This practice is accepted because many data at one time in other words, the data are col-
studies have elements of several nonexperimental lected on only one occasion with the same par-
designs. Table 11-1 provides examples of studies ticipants rather than with the same participants at
classified with more than one design label. several times.
240 PART THREE Quantitative Research

An example of a cross-sectional study is pro- different ages. By collecting data from each
vided by Richard and associates (2010), who participant at yearly intervals, the investigator
explored patient satisfaction in ambulatory cancer obtains a longitudinal perspective of the diabetic
care. Another cross-sectional study approach is to process.
simultaneously collect data on the study’s vari- In another example of a longitudinal study,
ables from different cohort (participants) groups. Voyer and associates (2005) collected data from
An example of a cross-sectional study with dif- a sample of 138 older adults. The purpose of the
ferent cohort groups was conducted by Fox and study was to describe the mental health status of
colleagues (2010), who used a naturalistic cohort long-term users of benzodiazepines and to
design to determine the relationship between bed compare their status with the mental health of
rest and orthostatic intolerance of sitting in adults older adults who have either begun or stopped
residing in chronic care facilities. Cohorts repre- consuming benzodiazepines over a 1-year period.
sented different amounts of bed rest that were Cross-sectional and longitudinal designs have
occurring naturally: no bed rest, two to four many advantages and disadvantages. When
days of bed rest, and five to seven days of bed assessing the appropriateness of a cross-sectional
rest. In cross-sectional studies, researchers can study versus a longitudinal study, the research
investigate relationships and correlations, differ- consumer should first assess the researcher’s goal
ences and comparisons, or both. For instance, Fox in view of the theoretical framework. For example,
and colleagues (2010) posed research questions in a hypothetical study of infant colic, the
that allowed the researchers to investigate both researchers are investigating a developmental
differences and relationships among and between process therefore, a longitudinal design seems
variables. more appropriate. However, the disadvantages
inherent in a longitudinal design also must be
considered. The period of data collection may be
Evidence-Informed Practice Tip long because of the time the participants take to
Replication of significant findings in nonexperi- progress to each data-collection point. In the
mental studies, with similar or different populations or
both, increases your confidence in the conclusions
infant colic study, it might take the researchers
offered by the researcher and the strength of evidence between 12 and 18 months to collect the data
generated by consistent findings from more than one from the total sample. Threats to internal validity,
study. such as testing and attrition also are ever present
and unavoidable in a longitudinal study (see
Chapter 16). As a result, longitudinal designs are
Longitudinal or Prospective Studies costly in terms of time, effort, and money. More-
In contrast to the cross-sectional design, the over, confounding variables could affect the
longitudinal study or ros ective study (also interpretation of the results. Participants in these
referred to as repeated-measures studies) collects studies may respond in a socially desirable way
data from the same group at different times. Re- that they believe is congruent with the investiga-
searchers also use longitudinal studies to explore tors’ expectations (see the discussion of the Haw-
differences and relationships. For example, the thorne effect in Chapter 9).
investigator conducting a study with children Despite the pragmatic constraints imposed by
with diabetes could use a longitudinal design. In a longitudinal study, the researcher should proceed
that case, the investigator could collect yearly with this design if the theoretical framework sup-
data or monitor the same children over a number ports a longitudinal developmental perspective.
of years to compare changes in the variables at The advantages of a longitudinal study are that
 CHAPTER 11 Nonexperimental Designs 241

participants are monitored separately and thereby affected by the independent variable, and the
serve as their own controls an increased depth of investigator attempts to link current events to
responses can be obtained and early trends in the past events.
data can be analyzed. The researcher can assess When scientists wish to explain causality or
changes in the variables of interest over time the factors that determine the occurrence of
and explore both relationships and differences events or conditions, they prefer to use an experi-
between variables. mental design. However, they cannot always
Cross-sectional studies, in comparison with manipulate the independent variable or use
longitudinal studies, are less time consuming and random assignments. In cases in which experi-
less expensive and are thus more manageable for mental designs cannot be employed, ex post facto
the researcher. Because large amounts of data can studies may be used. E post facto literally means
be collected at one time, the results are more from after the fact. These studies also are
readily available. In addition, the confounding known as causal-comparative studies or com-
variable of maturation, which results from the parative studies. As this design is discussed
passage of time, is not present. However, the further, you will see that ex post facto research is
investigator’s ability to establish an in-depth similar to quasiexperimental research because in
developmental assessment of the interrelation- both, differences between variables are examined
ships of the phenomena being studied is reduced. (Campbell Stanley, 1963).
Thus, the researcher is unable to determine In retrospective studies, a researcher hypoth-
whether the change that occurred is related to the esizes, for example, that variable (cigarette
change that was predicted because the same par- smoking) is related to and a determinant of vari-
ticipants were not monitored over a period of able (lung cancer), but , the presumed cause,
time. In other words, the participants are unable is not manipulated, and participants are not ran-
to serve as their own controls (see Chapter 10). domly assigned to groups. Instead, the researcher
In summary, longitudinal studies begin in the chooses a group of participants who have experi-
present and end in the future, and cross-sectional enced (cigarette smoking) in a normal situation
studies encompass a broader perspective of a and a control group of participants who have not
cross-section of the population at one specific experienced The behaviours, performances, or
time. conditions (lung tissue) of the two groups are
compared in order to determine whether the
exposure to had the effect predicted by the
Evidence-Informed Practice Tip hypothesis. Table 11-2 illustrates this example
The quality of evidence provided by a longitudi-
nal cohort study is stronger than that from other non-
and reveals that although cigarette smoking
experimental designs because the researcher can
determine the incidence of a problem and its possible
causes. TABLE 11-2
PARADIGM FOR THE EX POST FACTO DESIGN
INDEPENDENT
GROUPS (NOT VARIABLE (NOT
Retrospective or Ex Post Facto Studies RANDOMLY MANIPULATED BY DEPENDENT
A retros ective study is essentially the same ASSIGNED) INVESTIGATOR) VARIABLE
as an e ost acto study Epidemiologists pri- Exposed group: X: cigarette Ye: lung cancer
marily use the term retrospective, whereas social cigarette smokers smoking
scientists prefer the term e post facto In either Control group: Yc: no lung cancer
nonsmokers
case, the dependent variable has already been
242 PART THREE Quantitative Research

appears to be a determinant of lung cancer, the the major disadvantage of the retrospective
researcher is still not able to conclude that a design.
causal relationship exists between the variables Another disadvantage of retrospective research
because the independent variable has not been is that an alternative hypothesis may be the reason
manipulated and the participants were not ran- for the documented relationship. If the researcher
domly assigned to groups. obtains data from two existing groups of partici-
Another example of a retrospective study is pants, such as one that has been exposed to and
that of Tourangeau and associates (2006), who one that has not, and the data support the hypoth-
were interested in the effect of the structures esis that is related to , the researcher cannot
and processes of hospital nursing care on the be sure whether or an extraneous variable is the
rate of 30-day mortality among patients with cause of the occurrence of Finding naturally
acute medical conditions. Tourangeau and asso- occurring groups of participants who are similar
ciates collected retros ective data (data that in all ways except for their exposure to the vari-
have already been recorded, such as scores on able of interest is very difficult. The possibility
a standard examination) from several clinical always exists that the groups differ in another
and administrative secondary databases, which way (e.g., in exposure to another lung irritant,
included the Ontario Hospital Insurance Plan, the such as asbestos), which can affect the findings
Ontario Registered Persons Database, and the of the study and produce spurious results. Conse-
Ontario Hospital Reporting System (file 2002– quently, when you read about such a study, you
2003). In self-reports, collected through the need to cautiously evaluate the conclusions drawn
Ontario Nurses Survey 2003, nurses described by the investigator.
the type of clinical unit in which they worked,
their evaluation of patient care quality, their
career intentions, their feelings of burnout, and
Helpful Hint
When you read research reports, you will note
many other work-related variables. The study that, at times, researchers classify a study’s design with
revealed that 30-day mortality rates were lower more than one design label. This classification is correct
in association with hospitals that had a higher because research studies often reflect aspects of more
than one design.
percentage of registered nurse staff, a higher per-
centage of nurses with bachelor’s degrees, more
extensive use of care maps or protocols to guide Longitudinal or prospective (cohort) studies
patient care, and higher rates of burnout among are less common than retrospective studies
nurses. because it can take a long time for the phenom-
The advantages of the retrospective design are enon of interest to become evident in a prospec-
similar to those of the correlational design. The tive study. For example, if researchers were
additional benefit of the retrospective design studying pregnant women who regularly consume
is that it offers a higher level of control than a alcohol, it would take 9 months for the effect of
correlational study. For example, in a cigarette low birth weight in the participants’ infants to
smoking study, the lung tissue samples from non- become evident. The problems inherent in a pro-
smokers and smokers could be compared. This spective study are therefore similar to those of a
comparison would enable the researcher to estab- longitudinal study. However, longitudinal or pro-
lish the existence of a differential effect of ciga- spective studies are considered stronger than ret-
rette smoking on lung tissue. However, the rospective studies because of the degree of control
researcher would remain unable to draw a causal that can be imposed on extraneous variables that
link between the two variables. This inability is might confound the data.
 CHAPTER 11 Nonexperimental Designs 243

causal analysis (Kaplan, 2008 Schumacker

Helpful Hint Lomax, 2004).
Remember that nonexperimental designs can test
relationships, differences, comparisons, or predictions,
depending on the purpose of the study.
Practical Application
Samuels-Dennis and colleagues (2010) used
PREDICTION AND CAUSALITY IN causal modelling to validate a theoretical
NONEXPERIMENTAL RESEARCH model that highlights the process through which
posttraumatic stress disorder (PTSD) develops among
Researchers and research consumers are con- women. The results revealed that mothers’ strains
cerned with the issues of prediction and causality and personal resources played a significant mediating
role in the relationship between cumulative trauma
in explaining cause-and-effect relationships. His- and posttraumatic stress disorder.
torically, researchers have said that only experi-
mental research can support the concept of
causality. For example, nurses are interested in You will also find the terms path analysis,
discovering what causes anxiety in many settings. ISRE , analysis of covariance structures, struc-
If nurses can uncover the causes, they can perhaps tural e uation modelling (SEM), and hierarchi-
develop interventions that would prevent or cal linear modelling L (Raudenbush
decrease the anxiety. Causality makes it neces- Bryk, 2002) used to describe the statistical tech-
sary to order events chronologically therefore, if niques (see Chapter 16) used in these studies.
nurses find in a randomized experiment that event An HLM is a type of regression analysis that
1 (stress) occurs before event 2 (anxiety) and that allows for analysis of hierarchically structured
participants who experienced stress were anxious, data simultaneously at all levels (see the Practical
whereas those in the unstressed group were not, Application box for an example of the use of
then the hypothesis that stress causes anxiety is HLM).
supported. If these results occurred in a nonex-
perimental study in which some participants
underwent the stress of surgery and were anxious, Practical Application
whereas others did not have surgery and were not An example of HLM appeared in a study by
anxious, an association or relationship would be Letourneau and associates (2009), who
examined the influences of fathers’ characteristics on
said to exist between stress (surgery) and anxiety. the behavioural development of children who were
The results of a nonexperimental study, however, exposed to maternal postpartum depression. The
do not imply that the stress of surgery caused the findings indicated that fathers’ workforce participation
in the first 2 years of a child’s life has a long-term
anxiety. effect on the behavioural outcomes of children in
Many variables (e.g., anxiety) that nurse families affected by postpartum depression.
researchers wish to study to explore causation
cannot be manipulated, nor would it be wise to try
to manipulate the variables. However, studies that In a rediction study a model may be tested
can assert a predictive or causal sequence are to assess which independent variables can best
needed. In view of this need, many nurse research- explain the one or more dependent variables in
ers use several analytical techniques that can order to make a forecast or prediction derived
explain the relationships among variables to from particular phenomena. For example, Linton
establish predictive or causal links. These tech- and colleagues (2009) investigated the demo-
niques are called causal modelling (see the Practi- graphic predictors of human immunodeficiency
cal Application box), model testing, and associated virus (HIV) seropositivity status among homeless
244 PART THREE Quantitative Research

youth. Results indicated that there are relation- several studies in which researchers conducted
ships between age and HIV status and between sound tests of theoretical models.
ethnicity and HIV status, although not of a A full description of the techniques and prin-
predictive nature. Goff (2011) conducted an ex- ciples of causal modelling is beyond the scope of
planatory correlational study to explore learned this text.
resourcefulness, stressors, and academic per-
formance in baccalaureate nursing students.
Goff found that levels of personal and academic Evidence-Informed Practice Tip
stressors were evident but were not significant Research studies that entail the use of nonexperi-
predictors of academic performance. Age was a mental designs and provide level IV evidence can build
the foundation for a program of research that leads to
significant predictor of academic performance. experimental designs in which the effectiveness of
In another example, Singh and Cameron (2005) nursing interventions can be tested.
tested a model for predicting the psychosocial
effect of caring for patients who had experienced
a stroke. The variables of the model were devel-
oped from a previous systematic study and further ADDITIONAL TYPES OF
tested in this study. Singh and Cameron explained QUANTITATIVE STUDIES
the development of the model and the premise of Other types of quantitative studies complement
the study. The explanation enables readers of the the science of research. These additional designs
study to clearly understand the purpose and aim of provide a means of viewing and interpreting phe-
the research and the test of the model with regres- nomena to provide further breadth and knowl-
sion analyses. Although Singh and Cameron did edge to nursing science and practice. These
not test a cause-and-effect relationship between types of quantitative studies are methodological
the chosen independent predictor variables and research, meta-analysis, secondary analysis, and
the dependent criterion variable, the study did epidemiological studies.
demonstrate a theoretically meaningful model of
how variables work together in a group in a par- Methodological Research
ticular situation. ethodological research is the development
and evaluation of data-collection instruments,
scales, and techniques. As noted in Chapters 13
Helpful Hint and 14, methodology has a strong in uence on
Nonexperimental clinical research studies have research. The most significant and important
progressed to the point at which prediction models are aspect of methodological research addressed in
used to explore or test relationships between indepen-
dent variables and dependent variables. measurement development is sycho etrics the
theory and development of measurement instru-
ments (such as questionnaires) and measurement
As nurse researchers develop their programs techniques (such as observational techniques)
of research in a specific area, more tests of models through the research process. Thus, psychomet-
will be available. The statistics used in model- rics is concerned with the measurement of a
testing studies are advanced, but the beginning concept, such as anxiety or interpersonal con ict,
research consumer should be able to read the with reliable and valid tools. (See Chapter 14 for
article, understand the purpose of the study, and a discussion of reliability and validity.)
determine whether the model generated was Nurse researchers have used the principles of
logical and developed with a solid basis from the psychometrics to develop and test measurement
literature and past research. This section cites instruments that focus on nursing phenomena.
 CHAPTER 11 Nonexperimental Designs 245

Nurse researchers also use instruments developed items have been developed, the researcher
in other disciplines, such as psychology and soci- assesses the tool’s reliability and validity (see
ology, in which tools have been psychometrically Chapter 14). Various aspects of these procedures
tested. Sound measurement tools are critical for may differ according to the tool’s use, purpose,
the reliability and validity of a study. A study’s and stage of development.
purpose, problems, and procedures may be clear, In an example of methodological research,
and the data analysis may be correct and consis- Watson and associates (2008) documented the
tent, but if the measurement tool has inherent psychometric properties of the Nurses’ Attitudes
psychometric problems, the findings will be ren- Toward Obesity and Obese Patients Scale
dered questionable or of limited utility. (NATOOPS).
The main problem for nurse researchers is Common considerations that researchers
locating appropriate measurement tools. Many of incorporate into methodological research are
the phenomena of interest in nursing practice outlined in Table 11-3. Many more examples
and research are intangible, such as interpersonal of methodological research can be found in
con ict, caring, coping, and maternal-fetal attach- the nursing research literature (Akhtar-Danesh,
ment. The intangible nature of various phenomena, Valaitis, Schofield, nderwood, Martin-Misener,
and yet the need to measure them, places method- Baumann, Kolotylo, 2010 Roberts Ward-
ological research in an important position in re- Smith, 2010 Sidani, Epstein, Bootzin, Moritz,
search. Methodological research differs from other Miranda, 2009 Ward-Griffin, Keefe, Martin-
designs of research. First, it does not include all of Matthews, Kerr, Brown, Oudshoom, 2009). Psy-
the research process steps discussed in Chapter 2. chometric or methodological studies are found
Second, to implement methodical research tech- primarily in journals that report research. The
niques, the researcher must have a sound knowl- ournal of Nursing Measurement is devoted to the
edge of psychometrics or must consult with a publication of information on instruments, tools,
researcher knowledgeable in psychometric tech- and approaches for measurement of variables.
niques. The methodological researcher is not in- The specific procedures of methodological
terested in the relationship of the independent research are beyond the scope of this book, but
variable to a dependent variable or in the effect you are urged to look closely at the tools used in
of an independent variable on a dependent vari- studies.
able. Instead, the methodological researcher is
interested in identifying an intangible construct Meta-analysis
(concept) and making it tangible with a paper-and- eta analysis is a statistical technique, not a
pencil instrument or observation protocol. research design. It is based on a strict scientific
A methodological study includes the following approach in systematic reviews, in which the
steps: results of many studies in a specific area are syn-
Defining the construct, or concept, or be- thesized and statistically summarized to formu-
haviour to be measured late an overall conclusion. Each study is a unit of
Formulating the tool’s items analysis. The goal is to read all of the studies
Developing instructions for users and concerning a particular clinical question and,
respondents using rigorous inclusion and exclusion criteria, to
Testing the tool’s reliability and validity determine which studies are similar and to analyze
A sound, specific, and exhaustive literature review their results, in order to quantify the effectiveness
is necessary to identify the theories underlying of the intervention under study. This method is
the steps in this construct. The literature review more powerful because it is a rigorous process of
provides the basis of item formulation. Once the summarizing evidence rather than estimating the
 TABLE 11-3
COMMON CONSIDERATIONS IN THE DEVELOPMENT OF MEASUREMENT TOOLS
CONSIDERATION COMMENT

The well-constructed scale, test, interview schedule, or other A new tool should be based on a thorough review of
form of index should consist of an objective, standardized previous theoretical and research literature to ensure
measure of samples of a behaviour that has been clearly validity.
defined. Observations should be made on a small but
carefully chosen sampling of the behaviour of interest, thus
creating confidence that the samples are representative.
The tool should be standardized; that is, a set of uniform items Without specific criteria and rating procedures, the
and response possibilities are uniformly administered and evaluations of the items would be based on the
scored. subjective impressions, which may have varied
significantly between observers and conditions.
The items of a measurement tool should be unambiguous; For example, in constructing a tool to measure job
they should be clear-cut, concise, exact statements with only satisfaction, a nurse scientist writes the following
one idea per item. Negative stems or items with negatively item: “I never feel that I don’t have time to provide
phrased response possibilities result in ambiguity in meaning good nursing care.” The response format consists of
and scoring. “Agree,” “Undecided,” and “Disagree.” A response of
“Disagree” will likely not reflect the respondent’s true
intention because of the confusion that is created by
the double-negative phrasing “never . . . don’t.”
The type of items used in any one test or scale should be Mixing true-or-false items with questions that require a
restricted to a limited number of variations. Participants who yes-or-no response and items that provide a response
are expected to shift from one kind of item to another may format of five possible answers can lead to a high
fail to provide a true response as a result of the distraction of level of measurement error.
making such a change.
Items should not provide irrelevant clues. Unless carefully An item that provides a clue to the expected answer
constructed, an item may furnish an indication of the may contain value words that convey cultural
expected response or answer. Furthermore, the correct expectations, such as “A good wife enjoys caring for
answer or expected response to one item should not be her home and family.”
given by another item.
The items of a measurement tool should not be made difficult A test constructed to evaluate learning in an
by requiring unnecessarily complex or exact operations. introductory course in research methods may contain
Furthermore, the difficulty of an item should be appropriate an item that is inappropriate for the designated
to the level of the participants being assessed. Limiting each group, such as “A nonlinear transformation of data to
item to one concept or idea helps accomplish this objective. linear data is a useful procedure before a hypothesis
of curvilinearity is tested.”
The diagnostic, predictive, or measurement value of a tool Two nurse researchers are studying the construct of
depends on the degree to which it serves as an indicator of a quality of life. Each nurse has defined this construct
relatively broad and significant area of behaviour, known as in a different way. Consequently, the measurement
the universe of content for the behaviour. As already tool that each nurse devises will include different
emphasized, a behaviour must be clearly defined before it questions. The questions on each tool will reflect the
can be measured. The definition is developed from the universe of content for quality of life as defined by
universe of content: that is, the information and research each researcher.
findings that are available for the behaviour of interest. The
items should reflect that definition. The extent to which the
test items appear to accomplish this objective is an indication
of the validity of the instrument.
The instrument also should adequately cover the defined For example, few people would be satisfied with an
behaviour. The primary consideration is whether the number assessment of intelligence if the scale were limited to
and nature of items in the sample are adequate. If the three items.
sample has too few items, the accuracy or reliability of the
measure must be questioned. In general, the sample should
have a minimum of 10 items for each independent aspect of
the behaviour of interest.
The measure must prove its worth empirically through tests of The researcher should demonstrate to the reader that
reliability and validity. the scale is accurate and measures what it purports
to measure (see Chapter 14).
 CHAPTER 11 Nonexperimental Designs 247

effect of the results derived from single studies conduct the original analysis of data in the area
alone. Johnston (2005) noted that the meta- but instead synthesizes the data from already pub-
analysis of a number of randomized clinical lished studies by following a set of controlled and
trials (RCTs) gives due weight to the sample systematic steps. Systematic reviews with a meta-
size of the studies included and provides an analysis can be used to synthesize results of both
estimate of treatment effect in other words, a nonexperimental and experimental research
meta-analysis helps determine whether the inter- studies.
vention makes a difference. A systematic review Finally, evidence-informed practice requires
provides the most powerful and useful evidence that you, the research consumer, determine—on
available to guide practice: level I evidence (see the basis of the strength and quality of the evi-
Chapter 3). dence provided by the results of the meta-analysis,
When you critically appraise a systematic coupled with your clinical expertise and patients’
review, some of the questions to consider are the values—whether you would consider a change
following: in practice. For example, in their meta-analysis,
Does the systematic review address a Edwards and associates (2004) addressed the
focused research question clinical question Do psychological interventions
Does the meta-analysis include specific (education, individual cognitive-behavioural or
inclusion and exclusion criteria for judging psychotherapeutic programs, or group support)
the studies improve survival and psychological outcomes
Does a publication bias exist in women with metastatic breast cancer The
Are the included studies homogeneous in results of the systematic review were reported as
terms of purpose, sample, research methods follows:
Are the designs of the studies similar A search was conducted of published and
Are the interventions similar unpublished RCTs in any language that assessed the
Are the outcome measures similar effectiveness of psychological or psychosocial
Think about the systematic review as progres- interventions in women with breast cancer. Data
sively sifting and sorting data until the highest sources included the following: Cochrane Breast
Cancer Group Trials Register, Cochrane Central Register
quality of evidence is used to establish the con-
of Controlled Trials, Medline, CINAHL, PsychINFO, and
clusions. First, the researcher combines the results SIGLE; references of relevant studies and reviews; hand
of all of the studies that focus on a specific ques- searches of relevant journals; and known authors in
tion. The studies considered of lowest quality are the field. Two reviewers assessed the quality of
then excluded, and the quality of the remaining individual RCTs using the Jadad scale and another
studies are reanalyzed. The researcher repeats this method score that was more relevant for trials of
psychological interventions.
process sequentially, excluding studies until only
the studies of highest quality available are The main results of the systematic review indi-
included in the analysis. An alteration in the cate that five studies (N = 636) met the selection
overall results as an outcome of this sifting and criteria in two studies, cognitive-behavioural
sorting process suggests the sensitivity of the group interventions were assessed, and in three
conclusions to the quality of the studies included studies, supportive-expressive group therapy was
in the meta-analysis. assessed. Meta-analysis conducted with a fixed-
Such considerations determine whether it is effects model showed that rates of survival at 1,
reasonable to combine the studies for analysis. 5, or 10 years did not differ between participants
The consumer of research should note that a receiving group psychological interventions and
researcher who conducts a meta-analysis does not those receiving usual care. Cognitive-behavioural
248 PART THREE Quantitative Research

therapy also did not differ from the usual care meaning of caregiving, self-esteem, optimism,
with regard to anxiety (one trial), self-esteem burden, depression, spirituality, and posttrau-
(one trial), or mood state (one trial) at 6 months. matic stress in parents caring for children with a
Supportive expressive group therapy improved life-limiting illness. The original study was a
scores on the Courtauld Emotional Control scale cross-sectional, descriptive study focusing on the
at 8 months (one trial) and reduced reported pain, theoretical construct of posttraumatic growth. In
assessed on a 10-point visual analogue scale another study, Robinson and Molzahn (2007)
(meta-analysis of two trials weighted mean dif- conducted a secondary analysis of Canadian data
ference in reduction of 0.75 95 confidence from a larger international study designed to
interval, 0.63 to 0.86) in comparison with the develop and test a new module for the measure-
usual care. The groups did not differ in mood ment of quality of life of older adults. Data were
states at 10 to 12 months (two trials) or in quality available from a convenience sample of 426 older
of life at 1 year (one trial). Edwards and associ- adults in British Columbia. The purpose of the
ates (2004) concluded that existing data did not study was to explore the relationships between
provide evidence of a survival benefit for women sexual activity and intimacy and the quality of life
with metastatic breast cancer who received group of older adults.
psychological interventions over those who
received the usual care. Evidence on the effects Epidemiological Studies
on various aspects of psychological functioning In an e ide iological study factors affecting the
is mixed. health and illness of populations are examined in
Systematic reviews in which results from mul- relation to the environment. The purview of
tiple RCTs are combined offer stronger evidence public health for many years, epidemiological
(level I) in estimating the magnitude of an effect studies are investigations of the distribution,
for an intervention (see Chapter 3, Figure 3-1). determinants, and dynamics of health and disease.
The strength of evidence provided by systematic In these studies, investigators attempt to link
reviews has become the foundation of evidence- effects with cause however, a clear understand-
informed practice. ing of the causes is often not possible, especially
when the illness or problem has already occurred
Evidence-Informed Practice Tip and the method is to look retrospectively at the
Evidence-informed practice methods, such as sys-
evidence.
tematic reviews, increase a nurse’s ability to manage the Some of the questions that epidemiological
ever-increasing volume of information produced to researchers attempt to answer are Did exposure
develop the best practices that are evidence based. to a certain environment affect health and
Does staff shortage or organizational issues
affect burnout Research cannot answer such
Secondary Analysis questions directly but can establish a statistically
Secondary analysis also is not a design but a significant association between exposure to caus-
form of research in which the previously col- ative factors and disease or the effects of ill
lected and analyzed data from one study are rean- health.
alyzed for a secondary purpose. The original Two frequently conducted types of epidemio-
study may be either an experimental or a nonex- logical studies are studies of revalence (the
perimental design. For example, Schneider and number of people affected by a disease or health
colleagues (2011) conducted a secondary analysis problem) and studies of incidence (the number of
to examine the psychosocial outcomes of the cases occurring in a particular period).
 CHAPTER 11 Nonexperimental Designs 249

APPRAISING THE EVIDENCE

Nonexperimental Designs
The criteria for critiquing nonexperimental designs are Finally, the factor or factors that influence changes in
presented in the Critiquing Criteria box. When you cri- the dependent variable are often ambiguous in nonex-
tique nonexperimental research designs, keep in mind perimental designs. As with all complex phenomena,
that such designs offer the researcher the least amount multiple factors can contribute to variability in the partici-
of control. The first step in critiquing nonexperimental pants’ responses. When an experimental design is not
research is to determine which type of design was used used for controlling some of these extraneous variables
in the study. Often, a statement describing the design of that can influence results, the researcher must strive to
the study appears in the abstract and in the “Methods” provide as much control of these variables as possible
section of the report. If such a statement is not present, within the context of a nonexperimental design.
you should closely examine the report for evidence of When it has not been possible to randomly assign
which type of design was employed. You should be able participants to treatment groups as an approach to con-
to discern that either a survey or a relationship design was trolling an independent variable, the researcher may use
used, as well as the specific subtype. For example, you a strategy of matching participants for identified variables.
would expect an investigation of self-concept develop- For example, in a study of birth weight, pregnant women
ment in children from birth to 5 years of age to be a could be matched with regard to variables such as weight,
relationship study with a longitudinal design. height, smoking habits, drug use, and other factors that
Next, you should evaluate the theoretical framework might influence the birth weights of their infants. The
and underpinnings of the study to determine whether a independent variable of interest, such as the type of pre-
nonexperimental design was the most appropriate natal care, would then be the major difference in the
approach to the problem. For example, in many of the groups. You would then feel more confident that the only
studies on pain discussed throughout this text, the rela- difference between the two groups was the differential
tionship between pain and any of the independent vari- effect of the independent variable because the other
ables under consideration cannot be manipulated. For factors in the two groups were theoretically the same.
such studies, a nonexperimental correlational, longitudi- However, you should also remember that other influen-
nal, or cross-sectional design is appropriate. Investigators tial variables—such as income, education, and diet—
use one of these designs to examine the relationship might have been present but were not considered in
between the variables in naturally occurring groups. matching. Threats to internal and external validity repre-
Sometimes, you may think that it would have been more sent a major influence on the interpretation of a nonex-
appropriate for the investigators to use an experimental perimental study because they impose limitations on the
or a quasiexperimental design. However, you must rec- generalizability of the results.
ognize that pragmatic or ethical considerations also may If you are critiquing one of the additional types of
have guided the researchers in their choice of design (see research discussed, you must first identify the type of
Chapters 6 and 10). research used; then you must understand its specific
You should assess whether the problem is at a level purpose and format. The format and methods of second-
of experimental manipulation. Often, researchers merely ary analysis, methodological research, and meta-analysis
wish to examine whether relationships exist between vary; knowing how they vary allows you to assess
variables. Therefore, when you critique such studies, you whether the process was applied appropriately. Some of
should be able to determine the purpose of the study. If the basic principles of these methods were presented in
the purpose of the study does not include the expecta- this chapter. The specific criteria for evaluating these
tion of a cause-and-effect relationship, you need not look designs are beyond the scope of this text, but the refer-
for one. However, be wary when the researcher in a ences provided can assist you in this process. Even though
nonexperimental study suggests a cause-and-effect rela- the format and methods vary, all research has a central
tionship in the findings. goal: to answer questions scientifically.
250 PART THREE Quantitative Research

CRITIQUING CRITERIA

1. Which nonexperimental design 5. Is the design suited to the effect relationships between the
is used in the study? data-collection methods? variables?
2. In accordance with the 6. Does the researcher present 8. Are alternative explanations for
theoretical framework, is the the findings in a manner the findings possible?
rationale for the type of design congruent with the design 9. How does the researcher
evident? used? discuss the threats to internal
3. How is the design congruent 7. Does the researcher theorize and external validity?
with the purpose of the study? beyond the relational 10. How does the researcher deal
4. Is the design appropriate for the parameters of the findings and with the limitations of the
research problem? erroneously infer cause-and- study?

want to implement it immediately, whereas

Helpful Hint others want to implement it with only some
As you read the literature, you will find studies patients. You think that it should be
with labels such as outcomes research, needs assess-
implemented as a research study. Could
ments, evaluation research, and quality assurance.
These studies are not designs per se; instead, these either of the ways the nurses want to
studies are conducted with either experimental or non- implement the protocol be considered in a
experimental designs. Studies with these labels are research study?
designed to test the effectiveness of health care tech- ■ You are part of a journal club at your hospital.
niques, programs, or interventions. When reading such
Your group has been examining a phenomenon
a research study, you should assess which design was
used and whether the principles of the design, sampling specific to your patient population and noticed
strategy, and analysis are consistent with the study’s that 20 correlational studies on the topic have
purpose. been published. Your group decides to perform
a meta-analysis of the data. What steps need
to be considered in performing the meta-
CRITICAL THINKING CHALLENGES analysis? What level of evidence would you
expect to obtain with this method? Explain
■ Discuss which type of nonexperimental design your answer.
might help validate the defining characteristics of
a particular nursing diagnosis you use in practice.
Do you think it is possible for nurses and
patients to serve as the participants in this type KEY POINTS
of study?
■ The midterm group (five-student) assignment for • Nonexperimental research designs are used in studies
your research class is to critique an assigned that make an account of events as they naturally
quantitative study. To proceed, you must first occur. The major difference between
nonexperimental and experimental research is that in
decide the study’s overall type. You think it is an nonexperimental designs, the independent variable
ex post facto nonexperimental design, whereas is not actively manipulated by the investigator.
the other students think it is an experimental • Nonexperimental designs can be classified either as
design because the study has several explicit survey studies or as relationship/difference studies.
hypotheses. How would you convince the other • Survey studies and relationship/difference studies
are both descriptive and exploratory in nature.
students that you are correct? • In survey research, the investigator collects detailed
■ You are completing your senior practicum on a descriptions of existing phenomena and uses the
surgical step-down unit. The nurses completed data either to justify current conditions and
an evidence-informed practice protocol for practices or to make more intelligent plans for
patient-controlled analgesics. Some of the nurses improving them.
 CHAPTER 11 Nonexperimental Designs 251

informational support, anxiety, and satisfaction

• In relationship/difference studies, researchers with care. Intensive and Critical Care Nursing, ,
endeavour to explore the relationships or
114-122.
differences between variables in order to provide
deeper insight into the phenomena of interest. Campbell, D. T., Stanley, J. C. (1963). E perimental
• In correlational studies, researchers examine and uasi-e perimental designs for research.
relationships. Chicago: Rand-McNally.
• Developmental studies are further divided into Chojecki, P., Lamarre, J., Buck, M., St-Sauveur, I.,
categories of cross-sectional, longitudinal Eldaoud, N., Purden, M. (2010). Perceptions of a
(prospective), and retrospective (ex post facto) peer learning approach to pediatric clinical education.
studies. International ournal of Nursing Education Scholar-
• Methodological research, secondary analysis, and ship, (1), Article 39. Epub 2010 Oct 29.
meta-analysis are examples of other means of Edwards, A. G., Hailey, S., Maxwell, M. (2004).
adding to the body of nursing research. Both the
Psychological interventions for women with meta-
researcher and the reader must consider the
advantages and disadvantages of each design. static breast cancer. Cochrane Database System
• Nonexperimental research designs do not enable Revie , (2), CD004253.
the investigator to establish cause-and-effect Fox, M. T., Sidani, S., Brooks, D. (2010). The
relationships between the variables. You must be relationship between bed rest and sitting orthostatic
wary of nonexperimental studies in which intolerance in adults residing in chronic care
researchers make causal claims about the findings, facilities. ournal of Nursing and ealthcare of
unless a causal modelling technique is used. Chronic Illness, (3), 187-196. doi:10.1111/j/1752-
• Nonexperimental designs offer the researcher the 9824.2010.01058.x
least amount of control. Threats to validity represent Goff, A.-M. (2011). Stressors, academic performance,
a major influence on the interpretation of a
and learned resourcefulness in baccalaureate nursing
nonexperimental study because they impose
limitations on the generalizability of the results and, students. International ournal of Nursing Education
as such, should be fully assessed by the critical reader. Scholarship, (1), Article 1.
• The critiquing process is directed toward evaluating Hernandez, C. A., Hume, M. R., Rodger, N. W.
the appropriateness of the selected (2008). Evaluation of a self-awareness intervention
nonexperimental design in relation to factors such for adults with type 1 diabetes and hypoglycemia
as the research problem, the theoretical awareness. Canadian ournal of Nursing Research,
framework, the hypothesis, the methodology, and (3), 38-56.
the data analysis and interpretation. Hoe Harwood, C., Reimer-Kirkham, S., Sawatzky, R.,
• Although nonexperimental designs do not provide Terblanche, T., Van Hofwegen, L. (2010).
the highest level of evidence (level I), they do provide
Innovation of community clinical placements: A
a wealth of data that become useful for formulating
both level I and level II studies that are aimed at Canadian survey. International ournal of Nursing
developing and testing nursing interventions. Education Scholarship, (1), Article 28.
Johnston, L. (2005). Critically appraising quantitative
evidence. In B. M. Melnyk E. Fineout-Overholt
(Eds.), Evidence-based practice in nursing and
healthcare (pp. 79-126). Philadelphia: Lippincott
REFERENCES Williams Wilkins.
Akhtar-Danesh, N., Valaitis, R. K., Schofield, R., Kaasalainen, S., Agarwal G., Dolovich, L.,
nderwood, J., Martin-Misener, R., Baumann, A., Papaioannou, A., Brazil, K., Akhtar-Danesh, N.
Kolotylo, C. (2010). A questionnaire for assessing (2010). Nurses’ perceptions of and satisfaction
community health nurses’ learning needs. estern with the medication administration system in
ournal of Nursing Research, (8), 1055-1072. long-term-care homes. Canadian ournal of
Armstrong, K., Laschinger, H., Wong, C. (2009). Nursing Research, (4), 58-79.
Workplace empowerment and magnet hospital Kaplan, D. (2008). Structure e uation modeling:
characteristics as predictors of patient safety climate. oundations and e tensions. Thousand Oaks, CA:
ournal of Nursing Care uality, (1), 55-62. Sage.
Bailey, J. J., Sabbagh, M., Loiselle, C. G., Boileau, J., Koren, I., Mian, O., Rukholm, E. (2010). Integration
McVey, L. (2010). Supporting families in the of nurse practitioners into Ontario’s primary health
IC : A descriptive correlational study of care system: Variations across practice settings.
252 PART THREE Quantitative Research

Canadian ournal of Nursing Research, (2), Schumacker, R. E., Lomax, R. C. (2004). A begin-
48-69. ner s guide to structural e uation modeling. Hills-
Letourneau, N., Duffett-Leger, L., Salmani, M. dale, NJ: Erlbaum.
(2009). The role of paternal support in the behav- Sidani, S., Epstein, D. R., Bootzin, R. R., Moritz, P.,
ioural development of children exposed to postpartum Miranda, J. (2009). Assessment of preferences for
depression. Canadian ournal of Nursing Research, treatment: Validation of a measure. Research in
(3), 86-106. Nursing and ealth, , 419-438.
Linton, A., Singh, M., Turbow, D., Legg, T. J. (2009). Singh, M., Cameron, J. (2005). The psychosocial
Street youth in Toronto: An investigation of demo- aspects of caregiving to stroke patients. A N, (1),
graphic predictors of high risk behaviors and HIV 18-24.
status. ournal of IV AIDS Social Services, , Thomas, J., Jensen, L., Nahirniak, S., Gibney, R. T.
375-396. (2010). Anemia and blood transfusion practices in the
Ohler, M., Kerr, M. S., Forbes, D. A. (2010). critically ill: A prospective cohort review. eart
Depression in nurses. Canadian ournal of Nursing ung, (3), 217-225.
Research, (3), 66-82. Tourangeau, A. E., Doran, D. M., McGillis-Hall, L.,
Plunkett, R. D., Iwasiw, C. L., Kerr, M. (2010). The O’Brien-Pallas, L., Pringle, D., Tu, J. V. (2006).
intention to pursue graduate studies in nursing: A Impact of hospital nursing care on 30-day mortality
look at BScN students’ self-efficacy and value for acute medical patients. ournal of Advanced
in uences. International ournal of Nursing Educa- Nursing, , 32-44.
tion Scholarship, (1), Article 23. Voyer, P., Cappeliez, P., P rodeau, G., Pr ville, M.
Raudenbush, S., Bryk, A. (2002). ierarchical linear (2005). Mental health for older adults and benzodiaz-
models: Applications and data analysis methods (2nd epine use. ournal of Community ealth Nursing, ,
ed.). Newbury Park, CA: Sage. 213-229.
Richard, M. L., Parmar, M. P., Calestagne, P. P., Wahoush, E. O. (2009). Equitable health-care access:
McVey, L. (2010). Seeking patient feedback: An The experiences of refugee and refugee claimant
important dimension of quality in cancer care. mothers with an ill preschooler. Canadian ournal of
ournal of Nursing Care uality, (4), 344-351. Nursing Research, (3), 186-206.
Roberts, C. A., Ward-Smith, P. (2010). Choosing a Ward-Griffin, C., Keefe, J., Martin-Matthews, A.,
career in nursing: Development of a career search Kerr, M., Brown, J-B., Oudshoom, A. (2009).
instrument. International ournal of Nursing Development and validation of the Double Duty
Education Scholarship, (1), Article 2. Caregiving Scale. Canadian ournal of Nursing
Robinson, J. G., Molzahn, A. E. (2007, March). Research, (3), 108-128.
Sexuality and quality of life. ournal of erontologi- Watson, L., Oberle, K., Deutscher, D. (2008).
cal Nursing, (3), 19-27. Development and psychometric testing of the Nurses’
Samuels-Dennis, J., Ford-Gilboe, M., Wilk, P., Avison, Attitudes Toward Obesity and Obese Patients
W. R., Ray, S. (2010). Cumulative trauma, (NATOOPS) scale. Research in Nursing and ealth,
personal and social resources, and post-traumatic , 586-593.
stress symptoms among income-assisted single Yonge, O., Rosychuk, R. J., Bailey, T. M., Lake, R.,
mothers. ournal of amily Violence, (6), 603-617. Marrie, T. J. (2010). Willingness of university nursing
doi:10.1007/s10896-010-9323-7 students to volunteer during a pandemic. Public
Sawhney, M., Sawyer, J. (2008). A cross-sectional ealth Nursing, (2), 174-180.
study of the role of Canadian nurses with a specialty
practice in pain management. Acute Pain, ,
151-156.
FOR FURTHER STUDY
Schneider, M., Steele, R., Cadell, S., Hemsworth, D. Go to Evolve at http://evolve.elsevier.com/
(2011). Differences on psychosocial outcomes Canada/LoBiondo/Research for Audio Glossary, how-to
between male and female caregivers of children with instructions for Writing Proposals for Funding, and
life-limiting illnesses. ournal of Pediatric Nursing, additional research articles for practice in reviewing
(3), 186-199. doi:10.1016/j.pedn.2010.01.007 and critiquing.
253

PA RT FOUR
PART ONE Part Title

Processes Related
to Research
12 Sampling
13 Data-Collection Methods
14 Rigour in Research
15 Qualitative Data Analysis
16 Quantitative Data Analysis
17 Presenting the Findings
RESEARCH VIGNETTE Does IPE, when two or
more professions learn with,
Developing and Implementing a from and about each other to
Framework for Interprofessional improve collaboration and
the quality of care (Centre
Education in Health Sciences for the Advancement of In-
Education: Can Nursing Contribute? terprofessional Education,
2002), result in the forma-
tion of a more collaborative
nursing practitioner
Julie Gaudet, RN, MN (Admin.) on knowledge and skills related
Professor Does IPE lead to entry-to-
to collaboration, less is known
School of Nursing
about the long-term effect of IPE practice nurses’ adoption of
Centre for Health Sciences an inclusive view of a health
George Brown College on student behaviour, postregis-
Toronto, Ontario tration collaborative practice, and care professional
patient care (Hammick, Freeth, The answers to these questions
Koppel, Reeves, Barr, 2007). have since preoccupied my
As a teacher at George Brown In my former role as a mental thoughts and those of many of my
College since 2002, I have always health clinical nurse, it was colleagues. As a beginning step to
been interested in educational natural for me to work collabora- meeting the identified need for
research. When I was asked to par- tively with multidisciplinary staff, IPE, a team of committed leaders
ticipate in the creation of interpro- including nurses, psychiatrists, and faculty members embarked on
fessional education (IPE) curricula psychologists, dietitians, physio- a developmental journey that
in health sciences at the college, I therapists, occupational therapists, involved the creation of an early
was eager to get involved because and a myriad of other clinical con- curriculum framework for IPE
of my belief that interdisciplinary sultants. From the outside look- and four key IPE learning out-
collaboration is key to achieving ing in this type of collaborative comes (George Brown College,
high-quality patient care. working scenario does not appear 2004) for all students attending
Through my work on a variety unique to the mental health setting programs in the Centre for Health
of IPE projects, I discovered that but is true of all specialties. Con- Sciences. The IPE learning out-
although the concept of learning versely, professionals who work comes that follow were designed
in an interdisciplinary context as or spend a considerable amount of to enable students to accomplish
a preparatory step for real world time in psychiatric settings would the following objectives:
practice intuitively makes sense to argue that clinical discussions, Appraise the relationship
everyone, students generally lack and those related to the therapeu- between their own profession
the understanding of each mem- tic milieu, decision making, and and the background, scope, and
ber’s role on a health care team team dynamics have, at times, a roles of other health care pro-
and have limited opportunities to different feel than within more fessionals
interact with students from other traditional clinical settings. Evaluate their ability to work
health care professions while in Although reality in this context in a team
their respective programs. More- is highly subjective to the interpre-
over, at times, professional bound- tation of the individuals involved, In the Centre for Health Sciences at
ary issues are misinterpreted as my curiosity related to the follow- George Brown College, the term
interprofessional education is used to
barriers to learning collaboratively. ing research questions was, of
include all such learning in academic and
Although a growing body of evi- course, expected and grounded in work-based settings before and after
dence exists on the short-term my past experience within a psy- qualification an inclusive view of a
student outcomes of IPE focused chiatric interdisciplinary context: professional is thereby adopted.

254
 Participate collaboratively as a Interprofessional learning ac- of 7 years to gather insight into
health team member to support tivities such as simulation educa- the perspectives of student par-
patients’ achievement of their tion allows learners to gain a better ticipants and from faculty who
expected health outcomes understanding of interprofessional facilitated those IPE/simulation
Assess the effect of the broader teamwork and to re ect on their learning activities. Evaluation data
legislative and ethical frame- attitudes toward working in an from these projects highlighted
work on interprofessional prac- interprofessional team (Reeves, students’ overall positive attitudes
tice Lewin, Espin, warenstein, toward IPE, toward people from
In 2004, our team recognized 2010). These spaces have enabled other professions, and toward one
that specific learning outcomes a variety of opportunities for inter- another, and teaching faculty
were paramount for the increased professional learning among stu- gained considerable experience
IPE focus and that appropriate dents of the different disciplines, with this type of facilitation
interprofessional learning spaces who previously used learning however, additional research is
were needed for this type of edu- spaces in more insular ways. needed to determine whether
cation to be sustained. Significant A series of IPE programs and learning in interprofessional simu-
renovations were undertaken to pilot projects have been designed lated contexts, real-life contexts,
create interdisciplinary learning to expose both faculty and students or both will result in nursing stu-
spaces, which included a clinical to this form of learning. For dents’ ability to practise in more
simulation practice environment. example, in one project, nursing collaborative ways as they enter
According to the Society for Sim- and dental health students actively the practice setting and whether it
ulation in Healthcare (2010), learned blood pressure monitoring will ultimately lead to safer patient
health care simulation education skills and oral cleaning techniques outcomes or to more holistic care.
is a form of learning that mimics from each other. In another, nursing Because nurses make up the bulk
real world health care scenarios in students collaborated with students of practitioners on a health care
safe environments: from the dental hygiene, fitness and team in acute and chronic care set-
lifestyle management, health infor- tings, researchers must determine
mation management, and health whether the health outcomes and
Simulation is the imitation or informatics programs to develop a well-being of patients can be
representation of one act or common patient assessment tool. A improved as a result of training in
system by another. Healthcare
more recent project involved this interdisciplinary team-based
simulations can be said to have
four main purposes—education,
health sciences students and inter- approach.
assessment, research, and health professional health care providers In order to move the interpro-
system integration in facilitating (from Revera Inc., a long-term and fessional and simulation educa-
patient safety. residential care service provider in tion and research agendas forward
Each of these purposes may be Toronto) jointly involved in an in the Centre for Health Sciences
met by some combination of role interprofessional ethical decision- at George Brown College, and
play, low and high tech tools, and making workshop. This workshop, building on the initial learning
a variety of settings from tabletop structured in a way that fostered outcomes work (George Brown
sessions to a realistic full mission open discussions among the partic- College, 2004), my colleagues
environment. Simulations may ipants in diverse health care roles, and I have further developed a
also add to our understanding of
enabled both students and health framework by which to formally
human behavior and true-to-life
settings in which professionals
care professionals to apply the integrate IPE into our curriculum.
operate. The link that ties together framework to ethical issues and This framework involves six
all these activities is the act of dilemmas relevant in the resident major foundations of interprofes-
imitating or representing some population of older adults. sional care and their associated
situation or process from the These pilot projects provided levels of competencies: (1) health
simple to the very complex. several opportunities over a span care systems and professions, (2)

255
biomedical and social models of align with the professional entry learning theory (e.g., adult learn-
health, (3) communications, (4) to practice competencies (where ing theory, experiential learning,
collaboration, (5) con ict manage- relevant), and to the breadth, depth, problem-based learning, behav-
ment, and (6) critical thinking. and complexity of knowledge and ioural approaches, hybrid ap-
The specific competencies are vocational outcomes identified in proaches)
organized into steps or levels. For the Ministry of Training, Col- Pilot implementation projects
example, in relation to health care leges, and niversities’ Frame- began September 2011 and
systems and professions, we have work for Programs of Instruction. involve the BScN and Practical
defined two main levels that In addition, the planning for this Nursing (PN) students (and pos-
encompass three competencies: program will need to account for sibly students from the Gerontol-
nderstanding the scopes of other challenges, such as individ- ogy/Activation program). In 2012,
practice and the legislation ual program structure and length, we plan on expanding the number
(related to one’s own pro- and for broader expectations about of programs involved until all
fession and those of others) the health care system in order to our students are receiving the IPE
Social and political histories ensure the success of the frame- curriculum and participating in
of professions (understand- work implementation. the centre-wide interprofessional
ing power relationships and In terms of next steps, we plan activities.
imbalances that produce to identify the appropriate level of Engaging in IPE curriculum
barriers to effective inter- competency for each of our pro- development and research has
professional practice) grams. For example, students in a been tremendously rewarding. I
The competency ladder is as 1-year dental assisting program have witnessed first-hand how
follows: will not be expected to achieve the willing and eager various students
1. Demonstrates an under- same competencies as those in a and faculty colleagues are to col-
standing of one s o n pro- 4-year BScN program. We also laborate toward a common goal.
fession, history, legislation, plan to develop and integrate cur- Although learning what is unique
regulation, and scope of riculum into the programs to to nursing is paramount for devel-
practice achieve these competencies. We oping professional competence,
2. Appraises the relationship will establish Centre-wide IPE optimal patient-centred holistic
bet een one’s own profes- Days to expose all students in the care cannot be achieved indepen-
sion and the background/ Centre for Health Sciences to IPE dently of the contributions of
history, scope, and roles of content and experiential learning. other health care professionals.
other health care profes- Last, but perhaps most important, The pursuit of excellence in team-
sions we will begin to study and answer based collaborative practice will
3. Critically assesses the im- challenging but important opera- take time, but nursing is a driving
pact of social history, and tional and process questions, such force toward achieving this goal
broader political, legisla- as these: How do we get students ■
tive, and ethical frame- from different programs/years to-
works on one’s own gether to practise interprofession-
profession and on interpro- ally How will we overcome REFERENCES
fessional practice operational and logistical chal- Centre for the Advancement of
In the Centre for Health Sci- lenges in designating specific days Interprofessional Education. (2002).
ences at George Brown College, each semester for IPE initiatives De ning IPE. Retrieved from http://
the successful application of the How relevant and adequate are the www.caipe.org.uk.
George Brown College. (2004,
framework and the competency foundations of interprofessional
December 7). The future of health
levels on the competency ladder to care and their associated levels science education: rame or
be achieved by a specific program of competencies Are the com- for inter-professional education.
will, to a large extent, need to petencies adequately grounded in Toronto: Author.

256
Hammick, M., Freeth, D., Koppel, I., health professions Paper presented at Society for Simulation in Healthcare.
Reeves, S., Barr, H. (2007). A best IPE Ontario 2010, Toronto. (2010). hat is health care
evidence systematic review of Reeves, S., Lewin, S., Espin, S., simulation Retrieved from http://
interprofessional education. Medical warenstein, M. (2010). www.ssih.org/about-simulation.
Teacher, , 461–467. Interprofessional team or for health
Johnston, C., Toni, G. (2010, January). and social care. Oxford, K:
A frame or for IPE in non-degree Wiley-Blackwell.

257
C H A PTER 1 2

Sampling
Judith Haber | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Identify the purpose of sampling.
• Define population, sample, and sampling.
• Compare a population and a sample.
• Discuss the eligibility criteria for sample selection.
• Define nonprobability sampling and probability sampling.
• Identify the types of strategies for both nonprobability and probability sampling.
• Identify the types of qualitative sampling.
• Compare the advantages and disadvantages of specific nonprobability and probability sampling
strategies.
• Discuss the contribution of nonprobability and probability sampling strategies to the strength of
evidence provided by study findings.
• Discuss the factors that influence determination of sample size.
• Discuss the procedure for drawing a sample.
• Identify the criteria for critiquing a sampling plan.
• Use the critiquing criteria to evaluate the “Sample” section of a research report.

KEY TERMS
accessible population multistage sampling sampling
cluster sampling network sampling sampling frame
convenience sampling nonprobability sampling sampling interval
data saturation pilot study sampling unit
delimitations population simple random sampling
effect size probability sampling snowball effect sampling
element purposive sampling stratified random sampling
eligibility criteria quota sampling systematic sampling
heterogeneity random selection target population
homogeneous representative sample theoretical sampling
matching sample

258
 CHAPTER 12 Sampling 259

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

SAMPLING IS THE PROCESS OF SELECTING representa- Essentially, researchers sample representative

tive units of a population for study in a research segments of the population because sampling the
investigation. Although sampling is a complex entire population of interest to obtain relevant
process, it is a familiar one. In their daily lives, information is rarely feasible or necessary.
people gather knowledge, make decisions, and This chapter will familiarize you with the basic
formulate predictions on the basis of sampling concepts of sampling as they pertain to the prin-
procedures. For example, nursing students may ciples of quantitative and qualitative research
make generalizations about the overall quality of designs, nonprobability and probability sampling,
nursing professors as a result of their exposure to sample size, and the related critiquing process.
a sample of nursing professors during their under-
graduate programs. Patients may make general- SAMPLING CONCEPTS
izations about a hospital’s food or quality of
nursing care during a 1-week hospital stay. Population
Limited exposure to a limited portion of these A o ulation is a well-defined set that has certain
phenomena forms the basis of people’s conclu- specified properties or characteristics from which
sions, so much of their knowledge and many of data can be gathered and analyzed. A population
their decisions are based on their experience with can be composed of people, animals, objects, or
samples. events. For example, if a researcher is studying
Researchers also derive knowledge from sam- undergraduate nursing students, the type of edu-
ples. Many questions in scientific and naturalistic cational preparation of the population must be
research cannot be answered without the use of specified. In this example, the population consists
sampling procedures. For example, when the of undergraduate students enrolled in a generic
effectiveness of a new education intervention for baccalaureate nursing program. Examples of
diabetic patients is tested, the intervention is other possible populations might be all female
administered to a sample of the population. The patients admitted to a certain hospital for lumpec-
researcher must come to some conclusions tomies for treatment of breast cancer during 2008,
without giving the intervention to the entire popu- all children with asthma in the province of
lation of diabetic patients. To obtain the experi- Alberta, or all men and women with a diagnosis
ences or outcomes of engaging in this education, of schizophrenia in North America. These exam-
the researcher needs to select the appropriate ples illustrate that a population may be broadly
sampling strategy in accordance with the research defined and potentially involve millions of people,
design and question. This is done to avoid errone- or it may be narrowly specified to include only a
ous conclusions or making generalizations from few people.
a nonrepresentative sample. Thus, research meth- When you read a research report, you should
odologists have expended considerable effort to consider whether the researcher has identified
develop sampling theories and procedures that the population descriptors that form the basis for
produce accurate and meaningful information. the inclusion (eligibility) or exclusion criteria
260 PART FOUR Processes Related to Research

(delimitations) that are used to select the sample Examples of exclusion criteria, or deli ita
from the array of all possible units, whether tions (characteristics that restrict the population
people, objects, or events. Consider the popula- to a homogeneous group of participants), include
tion previously defined as undergraduate nursing gender, age, marital status, socioeconomic status,
students enrolled in a generic baccalaureate religion, ethnicity, level of education, age of chil-
program. Would this population include both dren, health status, and diagnosis. In a study of
part-time and full-time students Would it include the differences in sleep complaints in adults with
students who had previously attended another varying levels of bed rest residing in extended-
nursing program What about international stu- care facilities for chronic disease management,
dents At which level (first year through senior Fox and colleagues (2010) established the follow-
year) would students qualify As much as possi- ing exclusion criteria: ability to ambulate out of
ble, the researcher must specifically delineate the bed without physical assistance, receiving pallia-
exact criteria used to decide whether an individ- tive care, admission for short-term rehabilitation,
ual would be classified as a member of a given and being in the acute illness phase. These exclu-
population. The population descriptors that sion criteria, or delimitations, were selected
provide the basis for inclusion (eligibility) criteria because of their potential effect on the accurate
should be evident in the sample in other words, evaluation of the varying levels of bed rest among
the characteristics of the population and the patients in extended care.
sample should be congruent. The degree of con- In a randomized clinical trial, Smith and asso-
gruence is evaluated to assess the representative- ciates (2011) tested the efficacy of two smoking
ness of the sample. For example, if a population cessation interventions: one intensive and one
is defined as full-time, Canadian-born, senior- brief. The inclusion criteria were as follows: age
level nursing students enrolled in a generic bac- of 18 years or older, tobacco use in the past 30
calaureate nursing program, the sample would be days, minimum 36-hour stay, telephone access,
expected to re ect these characteristics. and willingness to be randomly assigned to a
Think about the concept of inclusion criteria, group and to quit smoking. Exclusion criteria
or eligibility criteria (characteristics of a popula- were as follows: enrollment in another smoking
tion that meet requirements for inclusion in a cessation trial, being pregnant, the presence
study), applied to a research study in which the of medical complications (e.g., receiving pallia-
participants are patients. For example, in an tive care, unstable condition, or institutionaliza-
investigation of the effects of music on dyspnea tion), inability to speak English, communication
during exercise in individuals with chronic difficulties, substance abuse, and a psychiatric
obstructive pulmonary disease (COPD), the par- history.
ticipants had to meet all of the following inclu- The heterogeneity or dissimilarities, of a
sion (eligibility) criteria: sample group inhibits the researchers’ ability to
1. A confirmed medical diagnosis of COPD interpret the findings meaningfully and to make
(i.e., chronic bronchitis, emphysema, or generalizations. It is much wiser to study only
both) one ho ogeneous group—that is, a group with
2. Ability to speak and read English limited variation in attributes or characteristics,
3. Ability to ambulate independently as in the aforementioned study—or to include
4. Experiencing dyspnea at least once a week specific groups as distinct subsets of the sample
5. An increase in the level of dyspnea of at and study the groups comparatively, as was the
least two points on the Borg scale after a case in Ostry and colleagues’ (2010) study. These
6-minute walk researchers sought to determine mental health
 CHAPTER 12 Sampling 261

outcomes in sawmill workers in British Colum- from patients in the Mount Sinai Hospital ante-
bia. They compared rural workers with urban partum clinic or all patient charts on file at a day
workers and, after controlling for socioeconomic surgery centre. A population can be defined in a
variables, concluded that rural sawmill workers variety of ways. Of importance is that the basic
had better outcomes. unit of the population must be clearly defined
Remember that exclusion criteria or delimita- because the generalizability of the findings is a
tions are not established in a casual or meaning- function of the population criteria.
less way but are established to control for
extraneous variability or bias. Each exclusion cri- Evidence-Informed Practice Tip
terion should have a rationale, presumably related Consider whether the sample selection was
to a potential contaminating effect on the depen- biased, thereby influencing the validity of the evidence
provided by the outcomes of the study.
dent variable. Carefully established sample exclu-
sion criteria, or delimitations, increase the
precision of the study and contribute to accuracy Samples and Sampling
while constraining the generalizability or trans- Sampling is a process of selecting a portion or
ferability of the findings (see Chapter 9). subset of the designated population to represent
The population criteria establish the target the entire population. A sa le is a set of ele-
o ulation that is, the entire set of cases about ments that make up the population an ele ent is
which the researcher would like to make gener- the most basic unit about which information is
alizations. A target population might include all collected. The most common element in nursing
undergraduate nursing students enrolled in research is individuals, but other elements (e.g.,
generic baccalaureate programs in Canada. places or objects) can form the basis of a sample
Because of time, money, and personnel, however, or population. For example, a researcher plans a
using a target population is often not feasible. An study to compare the effectiveness of different
accessible o ulation—one that meets the popu- nursing interventions on reducing falls in older
lation criteria and is available—is used instead. adults in long-term care facilities. Four facilities,
For example, an accessible population might in each of which a different treatment protocol is
include all full-time generic baccalaureate stu- used, are identified as the sampling units—not the
dents attending school in Manitoba. Pragmatic nurses themselves or the treatment alone. A sam-
factors must also be considered in identifying a pling unit can be an organization, a group, or an
potential population of interest. individual person.
The purpose of sampling is to increase the
Helpful Hint efficiency of a research study. Examining every
Often, researchers do not clearly identify the element or unit in the population would not be
population under study, or the population is not clari- feasible. When sampling is done properly, the
fied until the “Discussion” section, when an effort is
made to discuss the group (population) to which the researcher can draw inferences and make gener-
study findings can be generalized. alizations about the population without examin-
ing each unit in the population.
A population is not restricted to human partici- In qualitative research, the results can have
pants. The population may consist of hospital good generalizability to the population under
records blood, urine, or other specimens taken study. Sampling procedures that entail the formu-
from patients at a clinic historical documents or lation of specific criteria for selection ensure that
laboratory animals. For example, a population the characteristics of the phenomena of interest
might consist of all urine specimens collected will be, or are likely to be, present in all of the
262 PART FOUR Processes Related to Research

elements being studied. The researcher’s efforts the samples cannot be estimated. In other words,
to ensure that the sample is representative of the ensuring that every element has a chance for
target population provide a stronger position from inclusion in the nonprobability sample is not pos-
which to draw conclusions from the sample find- sible. In robability sa ling some form of
ings that are generalizable to the population (see random selection is used when the sample units
Chapter 9). are chosen. This type of sample enables the
After reviewing a number of research studies, researcher to estimate the probability that each
you will recognize that samples and sampling element of the population will be included in the
procedures vary in terms of merit. The foremost sample. Probability sampling is the more rigorous
criterion in evaluating a sample is its representa- sampling strategy used in quantitative research
tiveness. A re resentative sa le has key char- and is more likely to result in a representative
acteristics that closely approximate those of the sample.
population. If 70 of the population in a study of The remainder of this section is devoted to a
childrearing practices consisted of women and discussion of different types of nonprobability
40 were full-time employees, a representative and probability sampling strategies. A summary
sample should re ect these characteristics in the of sampling strategies appears in Table 12-1. You
same proportions. may refer to this table as the various nonprobabil-
The representativeness of a sample cannot be ity and probability strategies are discussed in the
guaranteed without access to a database about the following sections. Note that if there is bias in
entire population. Because it is difficult and inef- sampling, it will distort the analysis and the find-
ficient to assess an entire population, the ings of the study.
researcher must employ sampling strategies that
minimize or control for sample bias. If an appro- Helpful Hint
priate sampling strategy is used, the sample Research articles are not always explicit about the
data will almost always enable a reasonably accu- type of sampling strategy that was used. If the sampling
strategy is not specified, assume that in a quantitative
rate understanding of the phenomena under study, a convenience sample was used and that in a
investigation. qualitative study, a purposive sample was used.

Evidence-Informed Practice Tip Nonprobability Sampling

Determining whether the sample is representative Because of a lack of random selection, the non-
of the population being studied in journal articles will
influence both your interpretation of the evidence pro-
probability sampling strategy is less generaliz-
vided by the findings and your decision making about able than probability sampling because it tends
the findings’ relevance to the your patient population to produce less representative samples. Such
and practice setting. samples are more feasible for the researcher to
obtain, however, and most samples—in nursing
research and the research of other disciplines—
TYPES OF SAMPLES are nonprobability samples. When a nonprobabil-
Sampling strategies are generally grouped into ity sample re ects the target population through
two categories: nonprobability sampling and the careful use of inclusion and exclusion criteria,
probability sampling In non robability sa you can have more confidence in the representa-
ling elements are chosen through nonrandom tiveness of the sample and the external validity
methods. The drawback of this strategy is that of the findings. The major types of nonprobability
each element’s probability of being included in sampling are convenience, uota, net or , and
 CHAPTER 12 Sampling 263

TABLE 12-1
SUMMARY OF SAMPLING STRATEGIES
EASE OF DRAWING A
SAMPLING STRATEGY REPRESENTATIVE SAMPLE RISK OF BIAS REPRESENTATIVENESS OF THE SAMPLE

NONPROBABILITY
Convenience Very easy Greater than in any other Because samples tend to be
sampling strategy self-selecting;representativeness
is questionable
Quota Relatively easy Contains an unknown source Builds in some representativeness by
of bias that affects external using knowledge about the
validity population of interest
Purposive Relatively easy Bias increases with greater Very limited ability to generalize
heterogeneity of the because the sample is handpicked
population;conscious bias from a quantitative view, but this
is also a danger but is offset approach is necessary for the
with maximal variation qualitative researcher to choose
participants on the basis of the
phenomenon under study
Network Can be easy if the Minimal if a thorough Represents the event, incident, or
network is accessible sampling plan is developed experience being studied
Theoretical Requires a two-stage Minimal if a thorough Typically begins with another type
process;can be sampling plan is developed of sampling, such as convenience
prolonged or criterion sampling aimed at
variation in the phenomenon,
and thus represents aspects of the
theory being constructed

PROBABILITY
Simple random Laborious Low Maximized;the probability of
nonrepresentativeness decreases
with increased sample size
Stratified random Time-consuming Low Enhanced
Cluster Less time consuming Subject to more sampling Less representative than simple or
than simple or errors than is simple or stratified sampling
stratified sampling stratified sampling
Systematic More convenient and Bias in the form of Less representative if bias occurs as a
efficient than is simple, nonrandomness can be result of coincidental
stratified, or cluster inadvertently introduced nonrandomness
sampling

purposive sampling strategies Convenience and volunteers, the first 25 patients admitted to a
quota sampling can be used in both quantitative certain hospital with a particular diagnosis, all of
and qualitative research, whereas network and the people who enrolled in a certain program
purposive sampling are used mostly in qualitative during the month of September, or all of the stu-
research and are discussed later in this chapter. dents enrolled in a certain course at a particular
university during 2005. The participants are con-
venient and accessible to the researcher hence
Convenience Sampling the term convenience sample
Convenience sa ling is the use of the most In studying the association between parental
readily accessible persons or objects as partici- anxiety and compliance with preoperative require-
pants in a study. The participants may include ments for pediatric outpatient surgery, Chahal and
264 PART FOUR Processes Related to Research

colleagues (2009) used convenience sampling to osteoporosis in postmenopausal women who had
obtain a sample size of 203. Sinclair and Ferguson completed treatment (except for tamoxifen) for
(2009) used a convenience sample of 250 col- breast cancer, and for whom hormone replace-
laborative baccalaureate nursing students to study ment therapy was contraindicated, Waltman and
the effect of integrating simulated teaching and colleagues (2003) recruited participants from
learning strategies in undergraduate nursing edu- breast cancer support groups and through physi-
cation. The advantage of a convenience sample is cian referrals and local television and radio
that it is easy for the researcher to obtain partici- announcements. To assess the degree to which
pants. The researcher may need to be concerned a convenience sample approximates a random
only with obtaining a sufficient number of par- sample, a researcher can compare the conve-
ticipants who meet the same criteria. nience sample data with the known demographic
Linton and associates (2009) also obtained a information and by examining variability around
convenience sample from health and social the mean. In this manner, the researcher checks
service agencies servicing homeless youth. Chio- for the representativeness of the convenience
vitti (2006) used convenience sampling as a first sample and the extent to which bias is or is not
stage of sampling in a qualitative, grounded evident (Cochran, 1977 Sousa, auszniewski,
theory study in order to explore nurses’ meaning Musil, 2004).
of caring with patients in acute psychiatric hospi- Because recruiting research participants is a
tal settings. problem that confronts many nurse researchers,
The major disadvantage of a convenience innovative recruitment strategies are sometimes
sample is that the risk of bias is greater than in used. For example, a researcher may offer to pay
any other type of sample (see Table 12-1). Because the participants for their time. A relatively new
convenience samples entail voluntary participa- method of accessing and recruiting participants is
tion, the probability that researchers will recruit through online computer networks (e.g., disease-
people who feel strongly about the issue being specific chat rooms and bulletin boards).
studied is increased, which may favour certain In evaluating a research report, you should
outcomes of the study. The problem of bias is recognize that the convenience sample strategy,
related to the tendency of convenience samples although the most common, is the weakest form
to be self-selecting in other words, the researcher of sampling strategy in quantitative research in
obtains information only from the people who terms of generalizability. When a convenience
volunteer to participate. In this case, the follow- sample is used, researchers should analyze and
ing questions must be raised: interpret the data cautiously. When you critique a
What motivated some of the people to par- research study in which this sampling strategy
ticipate and others not to participate was used, you should be skeptical about the exter-
What kind of data would have been obtained nal validity of the findings (see Chapter 9).
if nonparticipants had also responded
How representative of the population are Quota Sampling
the people who did participate uota sa ling refers to a form of nonprobabil-
For example, a researcher may stop people on a ity sampling in which knowledge about the popu-
street corner to ask their opinion on an issue lation of interest is used to ensure some
place advertisements in the newspaper or put representativeness about the sample (see Table
signs in local churches, community centres, or 12-1). Through quota sampling, the researcher
supermarkets to recruit volunteers for a particular identifies a particular strata of the population, and
study. To study the prevention or treatment of the quota sample proportionally represents the
 CHAPTER 12 Sampling 265

TABLE 12-2
NUMBERS AND PERCENTAGES OF STUDENTS IN STRATA OF A QUOTA SAMPLE OF 5000 GRADUATES OF
NURSING PROGRAMS IN A PARTICULAR CITY
CATEGORIES DIPLOMA GRADUATES ASSOCIATE DEGREE GRADUATES BACCALAUREATE GRADUATES

Strata 1000 (20%) 2000 (40%) 2000 (40%)

Quota sample 100 200 200

strata. For example, the data in Table 12-2 reveal the problem of overrepresentation or underrepre-
that of the 5000 nurses in a particular city, 20 sentation of certain segments of a population in a
are diploma graduates, 40 are post–RN degree sample.
graduates, and 40 are baccalaureate graduates. An example of nonproportional quota sam-
Each of these strata should be proportionately pling is the study by Fox and colleagues (2010),
represented in the sample. In this case, the who examined differences in sleep complaints
researcher used a proportional quota sampling among adults with varying amounts of bed rest
strategy and decided to include 10 of a popula- who were residing in extended-care facilities for
tion of 5000 (i.e., 500 nurses). On the basis of the chronic disease management. The three cohorts
proportion of each stratum in the population, 100 (comparative, moderate, and high) re ected dif-
diploma graduates, 200 post-RN graduates, and ferent amounts of bed rest that were naturally
200 baccalaureate graduates were the quotas occurring. To ensure equal representation of the
established for the three strata. The researcher different amounts of bed rest, nonproportional
recruited participants who met the eligibility cri- quota sampling was used.
teria of the study until the quota for each stratum The characteristics chosen to form the strata
was filled. In other words, once the researcher are selected according to a researcher’s judge-
obtained the necessary 100 diploma graduates, ment on the basis of knowledge of the population
200 post-RN graduates, and 200 baccalaureate and the literature review. The criterion for selec-
graduates, the sample was complete with regard tion should be a variable that re ects important
to both the research design and other pragmatic differences in the independent variables under
matters, such as economy. investigation. Age, gender, religion, ethnicity,
The researcher systematically ensures that pro- medical diagnosis, socioeconomic status, level of
portional segments of the population are included completed education, and occupation are among
in the sample. An example is the study by Kaas- the variables that are likely to be important
alainen and Crook (2004), who evaluated the in stratifying samples in nursing research
ability of older adult residents of a long-term care investigations.
facility to report their pain. The researchers strati- In critiquing a research strategy, you seek to
fied a sample of 130 residents according to their determine whether the sample strata appropri-
level of cognitive impairment: cognitively intact, ately re ect the population under consideration
mildly impaired, moderately impaired, and and whether the variables used are homogeneous
extremely impaired. The quota sample is not ran- enough to ensure a meaningful comparison. Even
domly selected (i.e., once the proportional strata when the researcher has addressed these factors,
have been identified, the researcher obtains par- you must remember that a quota strategy is a
ticipants until the quota for each stratum has been nonprobability sample and thus includes an
filled), but it does increase the representativeness unknown source of bias that affects the external
of the sample. This sampling strategy addresses validity. The people who choose to participate
266 PART FOUR Processes Related to Research

may not be typical of the population in terms In si le rando sa ling the researcher
of the variables being measured, and assessing defines the population (a set), lists all units of the
the possible biases that may be operating is not population (a sa ling ra e), and selects a
possible. When the phenomena being investi- sample of units (a subset) from which the sample
gated are relatively similar within the population, will be chosen. For example, if Canadian hospi-
the risk of bias may be minimal however, tals specializing in the treatment of cancer were
in heterogeneous populations, the risk of bias is the sampling unit, a list of all such hospitals
greater. would be the sampling frame. If certified adult
nurse practitioners constituted the accessible pop-
Evidence-Informed Practice Tip ulation, a list of those nurses would be the sam-
When you think about applying study findings to pling frame.
your clinical practice, consider whether the participants Once a list of the population elements has been
in the sample are similar to your own patients. developed, the best method of selecting a sample
is to employ a table of random numbers contain-
ing columns of digits, as shown in Figure 12-1.
Probability Sampling Such tables can be generated by computer pro-
The primary characteristic of probability sam- grams. After assigning consecutive numbers to
pling is the random selection of elements from units of the population, the researcher starts at
the population. In rando selection each any point on the table of random numbers and
element of the population has an equal and inde- reads consecutive numbers in any direction (i.e.,
pendent chance of being included in the sample. horizontally, vertically, or diagonally). When a
In the hierarchy of evidence, probability sam- number is read that corresponds with the written
pling represents the strongest type of sampling unit on a card, that unit is chosen for the sample.
strategy. That means there is greater confidence The investigator continues to read until a sample
that the sample is representative rather than of the desired size is drawn. The advantages of
biased and that it more closely re ects the simple random sampling are as follows:
characteristics of the population of interest. Four The sample selection is not subject to the
commonly used probability sampling strategies conscious biases of the researcher.
are simple random sampling, strati ed random The representativeness of the sample is
sampling, cluster sampling, and systematic maximized in relation to the population
sampling characteristics.
Random selection of sample participants The differences in the characteristics of the
should not be confused with random assignment sample and the population are purely a
of participants. As discussed in Chapter 10, ran- function of chance.
domi ation refers to the assignment of partici- The probability of choosing a nonrepresen-
pants to either an experimental or a control group tative sample decreases as the size of the
on a purely random basis. sample increases.
Boscart (2009) used simple random sampling
Simple Random Sampling to select 30 patients from three different units to
Simple random sampling is a laborious and care- evaluate the effect of a brief, focused educational
fully controlled process. Because the principles interaction on the quality of verbal interactions
of simple random sampling are incorporated in between nursing staff and patients in a chronic
the more complex probability designs, the prin- care facility. You must remember that although
ciples of this strategy are presented. a researcher may use a carefully controlled
 CHAPTER 12 Sampling 267

1000 random integers between 0 and 99

40 23 0 29 10 94 17 58 12 85 13 25 80 84 72 74 54 63 55 31
32 98 59 23 74 97 51 42 21 87 48 64 54 38 84 68 14 17 35 48
84 34 84 14 53 65 67 37 2 45 84 21 71 34 10 80 72 27 11 13
86 37 24 89 23 4 44 40 72 81 44 69 25 44 34 34 34 75 50 50
50 58 85 8 22 24 73 20 63 35 60 87 91 92 96 80 19 22 87 24
1 87 43 82 9 31 40 88 33 28 82 73 18 6 48 64 59 45 34 3
21 19 42 76 84 67 29 68 8 66 93 89 96 28 12 14 38 47 52 65
32 66 33 21 81 97 39 76 67 27 97 22 76 89 41 11 91 29 6 66
16 82 42 75 35 42 92 90 77 24 21 8 36 16 5 54 89 51 57 85
74 32 63 65 93 96 18 36 82 72 39 69 37 97 51 17 36 71 38 30
50 94 4 66 17 37 10 53 8 29 67 74 88 38 11 59 60 91 56 17
71 47 81 18 53 98 7 87 29 37 22 93 13 6 95 7 95 71 14 6
71 93 48 16 33 19 46 21 60 44 52 91 52 58 10 9 41 31 35 18
20 94 13 99 45 6 53 54 1 25 79 28 1 48 36 26 68 37 59 7
75 22 69 56 62 40 64 45 40 99 94 14 98 84 22 38 24 87 43 71
16 87 41 0 88 83 11 37 71 78 22 39 43 37 75 84 84 11 55 58
92 90 80 2 30 37 84 55 56 50 3 71 24 13 62 74 82 44 90 32
96 89 31 32 37 45 70 67 80 55 58 9 55 60 61 55 86 44 27 77
38 29 36 94 65 39 56 29 29 65 88 13 71 38 71 8 81 66 31 44
20 6 61 66 90 13 70 60 92 53 87 49 34 42 14 47 75 33 26 9
63 44 94 21 14 13 41 80 39 72 29 3 25 89 44 88 13 49 18 58
13 32 93 90 31 75 86 95 18 51 61 59 84 95 67 54 40 30 29 63
26 35 48 81 19 24 36 36 76 16 46 5 93 41 97 46 79 54 95 49
89 74 96 95 94 69 31 60 16 69 76 42 28 71 69 34 46 55 20 42
50 39 28 64 20 68 60 33 92 82 61 70 5 68 95 88 12 85 18 94
55 86 5 96 87 69 75 93 54 79 0 57 45 8 86 59 25 21 9 29
75 35 1 2 86 62 70 83 85 13 97 37 13 73 16 38 36 23 54 11
74 50 1 77 87 92 68 87 57 36 17 47 0 97 78 72 72 45 54 51
34 24 35 13 26 42 22 75 47 2 34 87 15 50 65 27 5 72 28 68
73 33 42 65 91 24 44 84 71 55 70 1 27 30 8 61 65 61 18 92
7 55 12 6 61 17 23 95 91 58 60 30 35 61 34 27 75 44 35 64
10 94 18 4 3 19 21 37 28 55 76 25 10 29 80 64 8 81 20 32
20 48 92 87 95 58 57 73 42 1 12 81 94 85 63 97 24 19 93 51
81 10 92 49 70 15 76 4 36 92 62 99 78 32 86 74 43 22 98 46
66 67 82 94 67 75 16 88 84 98 0 52 37 0 43 9 0 51 2 62
84 92 36 11 3 52 44 65 45 67 97 86 92 2 50 5 93 66 73 40
36 29 98 46 88 23 28 44 8 71 69 43 53 16 87 21 56 23 37 24
15 11 82 30 59 94 23 30 40 25 87 26 24 30 44 53 33 65 72 55
89 57 49 79 83 88 42 45 41 93 38 24 15 80 97 18 61 12 13 42
23 36 65 9 64 26 93 37 26 44 42 17 45 68 27 77 74 56 49 34
9 93 90 61 45 40 75 85 64 66 36 89 72 43 99 90 92 10 10 85
53 94 30 31 62 92 82 30 94 56 40 4 50 53 9 74 87 2 36 36
18 69 77 38 89 78 30 68 71 92 22 93 91 74 52 1 97 69 71 42
50 20 76 36 6 20 75 56 36 5 14 70 9 78 23 33 91 33 25 72
30 46 1 10 16 72 69 26 94 39 80 36 36 68 92 74 22 74 41 42
59 47 7 92 77 55 2 12 5 24 0 30 25 62 83 36 92 96 36 75
93 22 3 20 82 44 16 69 98 72 30 57 77 15 90 29 32 38 3 48
9 55 27 41 40 94 77 14 54 10 25 75 1 74 72 15 69 80 33 58
70 8 3 5 46 89 28 86 40 6 25 40 81 26 63 97 87 48 26 41
19 6 89 31 80 60 13 89 17 69 38 93 58 55 54 69 74 33 8 55

FIGURE 12-1 A table of random numbers.

268 PART FOUR Processes Related to Research

sampling procedure that minimizes error, no of a research article must exercise caution in gen-
guarantee exists that the sample will be represen- eralizing from reported findings, even when
tative. Factors such as sample heterogeneity and random sampling is the stated strategy, if the
participant dropout may jeopardize the represen- target population has been difficult or impossible
tativeness of the sample despite the most strin- to list completely.
gent random sampling procedure. In examining
single mothers’ adverse and traumatic experi- Stratified Random Sampling
ences and posttraumatic stress symptoms, Strati ed rando sa ling requires that the
Samuels-Dennis and associates (2010) used population be divided into strata or subgroups.
simple random sampling to recruit 2400 single The subgroups or subsets that the population is
mothers from the active caseload of the provin- divided into are homogeneous. An appropriate
cial social assistance program. No simple answer number of elements from each subset is randomly
exists as to an acceptable response rate of surveys, selected on the basis of the proportion in the
but researchers consider the accuracy of their population. The goal of this strategy is to achieve
results, whether the population is heterogeneous, a greater degree of representativeness. Stratified
and the sorts of biases introduced by the number random sampling is similar to the proportional
of responses received. In Samuels-Dennis and stratified quota sampling strategy discussed
associates’ study, 247 single mothers completed earlier in this chapter. The major difference is that
the survey, which re ects a response rate of stratified random sampling involves a random
11.3 . The author states that this extremely low selection procedure for obtaining sample partici-
response rate diminishes our ability to generalize pants. Figure 12-2 illustrates the use of stratified
these findings to the wider population of income- random sampling.
assisted single mothers (p. 20). The population is stratified according to any
An assessment of the differences between number of attributes, such as age, gender, ethnic-
responders and nonresponders can also provide ity, religion, socioeconomic status, or level of
valuable information, such as length of survey, education completed. The variables selected to
degree of responder fatigue, and relevance of make up the strata lead to subgroups that share
questions (Berk, 2003). one or more of the attributes being studied (see
The major disadvantage of simple random Practical Application box). The following ques-
sampling is that it is a time-consuming and inef- tions can be asked in the selection of a stratified
ficient method of obtaining a random sample. sample:
(Consider the task of listing all baccalaureate Does a critical variable or attribute exist
nursing students in Canada.) With random that provides a logical basis for stratifying
sampling, it may also be impossible to obtain the sample
an accurate or complete listing of every element Does the population list contain sufficient
in the population. Imagine trying to obtain a list information about the attributes that will be
of all completed suicides in Toronto for 2001. used to divide the sample into subsets
Although suicide may have been the cause of Is it appropriate for each subset to be equal
death, another cause (e.g., cardiac failure) often in size, or is it more appropriate for each
appears on the death certificate. It would be dif- subset to be proportionally stratified on the
ficult to estimate how many elements of the target basis of the proportion of each subset in the
population would be eliminated from consider- population
ation. Bias would definitely be an issue, despite If proportional sampling is being used, is
the researcher’s best efforts. Thus, the evaluator the number of participants in each subset
 CHAPTER 12 Sampling 269

Entire
5000 Registered nurses in City X
population

Classification

Proportional STRATUM 1 STRATUM 2 STRATUM 3

stratified Diploma school Associate degree Baccalaureate
graduates graduates graduates
Population 20% = 1000 adults 40% = 2000 40% = 2000

Random Selection

Sample: 100 200 200

Randomized
selection
of 10% of Diploma-school Associate-degree Baccalaureate
each graduates graduates graduates
stratum
FIGURE 12-2 Participant selection through the use of a proportional stratified random sam-
pling strategy.

sufficient as a base for meaningful com- In one study (Akhtar-Danesh et al., 2010), a
parisons proportional random sampling was based on one
Once the subset comparison has been deter- strata—location of practice—in assessing com-
mined, are random procedures used for munity nurses’ learning needs. The authors
selection of the sample sampled 500 community health nurses: 350 from
Ontario and 150 from Nova Scotia, representing
equal proportions from each province.
As illustrated in Table 12-1, a stratified random
Practical Application sampling strategy has the following advantages:
A stratified random sampling plan was used (1) The representativeness of the sample is
by Ulrich and colleagues (2006) to identify
ethical concerns and conflicts of nurse practitioners enhanced and (2) the risk of bias is low (i.e., the
and physician assistants in relation to managed care researcher has a valid basis for making com-
and the factors that influence ethical conflict. A parisons among subsets if information about the
self-administered questionnaire was mailed to a
stratified sample of 3900 nurse practitioners and critical variables has been available). A third
physician assistants in the United States. Ulrich and advantage is that the researcher is able to over-
colleagues calculated the sample size for each sample a disproportionately small stratum to
stratum to detect a correlation of 0.2, allowing for
adjust for the researchers’ underrepresentation,
75% eligibility, 40% response rate, and three
different state practice environments (excellent or statistically weigh the data accordingly, and con-
favourable, acceptable, and limiting or restricting) tinue to make legitimate comparisons.
and allowing for a 10% difference in proportions The obstacles encountered by a researcher
among strata with 95% confidence and 80% power.
in using this strategy include (1) the difficulty
270 PART FOUR Processes Related to Research

of obtaining a population list containing com- ACNPs’ time must be spent in providing care
plete critical variable information (2) the time- directly to patients in acute or critical care prac-
consuming effort of obtaining multiple tices and (3) the participants must be in full-time
enumerated lists (3) the challenge of enrolling employment at the hospital. The second-stage
proportional strata and (4) the time and money sampling strategy calls for random selection of
involved in carrying out a large-scale study with a two ACNPs from each hospital who meet the
stratified sampling strategy. In critiquing the eligibility criteria.
study, you must question the appropriateness of When multistage sampling is used in relation
this sampling strategy for the problem under in- to large national surveys, provinces are used as
vestigation. the first-stage sampling unit, followed by succes-
To examine the effect of biographical factors sively smaller units (such as counties, cities, dis-
(i.e., gender, age, nursing education, current posi- tricts, and blocks) as the second-stage sampling
tion at work, and length of time in current work- unit and then households as the third-stage sam-
place) on job satisfaction, Curtis (2008) recruited pling unit.
a stratified random sample from all eligible reg- Sampling units or clusters can be selected by
istrants with the Irish Nursing Board on the basis simple random or stratified random sampling
of these factors. It is appropriate for the researcher methods (see Practical Application box). Suppose
to strive to represent all strata proportionately in that the hospitals described in the preceding
the study sample. example are grouped into four strata according to
size (i.e., number of beds) as follows: (1) 200 to
Multistage Sampling (Cluster Sampling) 299 (2) 300 to 399 (3) 400 to 499 and (4) 500
ultistage sa ling or cluster sa ling or more. Stratum 1 comprises 25 of the popula-
involves a successive random sampling of units tion stratum 2 comprises 30 of the population
(clusters) that meet sample eligibility criteria this stratum 3 comprises 20 of the population and
sampling progresses from large to small. A sa stratum 4 comprises 25 of the population. Thus,
ling unit is an element or set of elements used either a simple random or a proportional stratified
for selecting the sample. The first-stage sampling sampling strategy can be used to randomly select
unit consists of large units or clusters. The second- hospitals that would proportionately represent the
stage sampling unit consists of smaller units or population of hospitals in the provincial nurses’
clusters. Third-stage sampling units are even association list.
smaller. The main advantage of cluster sampling, as
Consider an example in which a sample of illustrated in Table 12-1, is that it is considerably
nurse practitioners is desired. The first sampling more economical in terms of time and money
unit is a random sample of hospitals, obtained than other types of probability sampling, particu-
from a provincial nurses’ association list, that larly when the population is large and geographi-
meet the eligibility criteria (e.g., size, type). The cally dispersed or when a sampling frame of the
second-stage sampling unit consists of a list of elements is not available. However, cluster sam-
acute care nurse practitioners (ACNPs) practising pling has two major disadvantages: (1) More
at each hospital selected in the first stage (i.e., the sampling errors tend to occur than with simple
list obtained from the vice president for nursing random or stratified random sampling, and (2) the
at each hospital). The criteria for inclusion in the appropriate handling of the statistical data from
list of ACNPs are as follows: (1) Participants cluster samples is very complex.
must be certified ACNPs with at least 2 years’ In critiquing a research report, you need to
experience as an ACNP (2) at least 75 of the consider whether the use of cluster sampling is
 CHAPTER 12 Sampling 271

justified in light of the research design, as well as from every tenth hospital room for a study on
other pragmatic matters, such as economy. patient satisfaction with nursing care. In the hos-
pital where the study was being conducted, every
tenth room happened to be a private room.
Practical Application Patients in private rooms might respond differ-
Vlack and colleagues (2007) obtained survey ently regarding their satisfaction than patients in
estimates of immunization coverage for semiprivate rooms. Because of the nonrandom
indigenous 2-year-old Australian children in arrangement of the rooms, bias may be
Queensland and compared these estimates with
those from the national Immunization Register. To introduced.
select a survey sample, they first stratified 153 Second, the first element or member of the
geographical areas in Queensland according to their sample must be selected randomly. In this case,
accessibility, creating four strata (from “highly
the researcher—who has a population list, or
accessible” to “very remote”), and then randomly
selected 30 of them for a total target sample of 210 sampling frame—first divides the population N
children: 7 eligible children from each area. This by the size of the desired sample n to obtain the
sample represented 6% of the estimated population. sampling interval width The sa ling inter
val is the standard distance between the elements
chosen for the sample. For example, to select a
Systematic Sampling sample of 50 family nurse practitioners from a
Syste atic sa ling is a sampling strategy that population of 500 family nurse practitioners, the
involves the selection of every th case drawn sampling interval would be as follows:
from a population list at fixed intervals, such as
500
every tenth member listed in the directory of the k= = 10
50
College and Association of Registered Nurses of
Alberta (CARNA). Systematic sampling might Essentially, every tenth case on the family nurse
be used to recruit every th person who enters practitioner list would be sampled. Thus, if the
a hospital lobby or who is hospitalized with a starting point was participant 5, then next person
diagnosis of acquired immune deficiency syn- chosen would be 15th, then 25th, etc.
drome (AIDS) in 2005. When systematic sam- Once the sampling interval has been deter-
pling is used, the population must be narrowly mined, the researcher uses a table of random
defined (e.g., as consisting of all people entering numbers (see Figure 12-1) to obtain a starting
or leaving the hospital lobby) for the sample to point for the selection of the 50 participants. If
be considered a probability sample. If older adults the population size is 500 and a sample size of 50
were sampled systematically on entering a hospi- is desired, a number between 1 and 500 is ran-
tal lobby, the resulting sample would not be a domly selected as the starting point. In this
probability sample because not every older adult instance, if the first number is 51, the family nurse
would have a chance of being selected. As such, practitioners corresponding to numbers 51, 61,
systematic sampling can sometimes represent a 71, and so forth would be included in the sample
nonprobability sampling strategy. of 50.
Systematic sampling strategies can be Another procedure recommended in many
designed, however, to fulfill the requirements of texts is to randomly select the first element
a probability sample. First, the listing of the popu- from within the first sampling interval. If the
lation (sampling frame) must be random in rela- sampling interval is 5, a number between 1 and 5
tion to the variable of interest. For example, is selected as the random starting point. For
suppose that participants were being selected example, the number 3 is randomly chosen.
272 PART FOUR Processes Related to Research

Keeping in mind the sampling interval of 5, the eralizability are drastically altered when a non-
next elements selected would correspond to the probability sample is involved.
numbers 8, 13, 18, and so on, until the sample For example, in their study, Cho and associates
was obtained. Although this procedure is techni- (2003) sought to determine the effects of nurse
cally correct, choosing a random starting point staffing (that is, ratio of number of Registered
from across the total population of elements is Nurses to Registered Practical Nurses) on adverse
more attractive because every element has a events, including morbidity, mortality, and
chance to be chosen for the sample during the first medical costs. The study used two existing data-
selection step. bases: California Hospital Financial Data and
Systematic sampling and simple random sam- 1997 data for the state of California released by
pling are essentially the same type of procedure. the Agency for Healthcare Research and uality
The advantage of systematic sampling is that the (AHR ). In the selection of hospitals and patients,
results are obtained in a more convenient and the researchers strived to create a sample that
efficient manner (see Table 12-1). The disadvan- included homogeneous hospital and patient
tage of systematic sampling is that bias in the groups while representing the majority of the
form of nonrandomness can be inadvertently target population. Hospitals were stratified by
introduced into the procedure. This problem may ownership, hospital size, teaching affiliation, and
occur if the population list is arranged so that a location nurse staffing was stratified by type of
certain type of element is listed at intervals that care unit (e.g., medical-surgical acute care,
coincide with the sampling interval. For example, medical-surgical intensive care, and coronary
if every tenth nursing student on a population list care). Patient characteristics were stratified by
of all types of nursing students in Ontario was a diagnostic related group and selected demo-
baccalaureate student and the sampling interval graphic variables. Because randomization was
was 10, baccalaureate students would be over- not used at any phase of this multilevel sampling
represented in the sample. procedure, you would consider this study to be a
Cyclical uctuations are also a factor in sys- nonprobability stratified sample with the external
tematic sampling. For example, if a list is kept of validity limitations of that sampling strategy (see
nursing students using the college library each Chapter 9).
day to do computer literature searches, a biased
sample would probably be obtained if every Evidence-Informed Practice Tip
seventh day, such as Sunday, is chosen as the The sampling strategy, whether probability or
sampling interval: in the case of Sunday, because nonprobability, must be appropriate for the study design
and evaluated in relation to the level of evidence pro-
probably fewer and perhaps different nursing stu- vided by the design.
dents use the library on Sundays than on week-
days. Therefore, caution must be exercised about
departures from randomness because they affect Special Sampling Strategies
the representativeness of the sample and, as a Several special sampling strategies are used in non-
result, the external validity of the study. probability sampling. atching is a special strat-
You should note whether a satisfactory random egy used to construct an equivalent comparison
selection procedure was performed. If random- sample group by filling it with participants who are
ization was not used, the systematic sampling similar to each participant in another sample group
may have become a nonprobability quota sample. in terms of preestablished variables, such as age,
You need to be cognizant of this issue because gender, level of education, medical diagnosis, or
the implications related to interpretation and gen- socioeconomic status. Theoretically, any variable
 CHAPTER 12 Sampling 273

other than the independent variable that could strategy used for locating samples that are diffi-
affect the dependent variable should be matched. cult or impossible to locate in other ways. This
In reality, the more variables matched, the more sampling strategy takes advantage of social net-
difficult it is to obtain an adequate sample size. works and the tendency of friends to share char-
Matching was used in a study that sought to acteristics. When a few participants with the
determine whether unmarried adolescent mothers necessary eligibility criteria are found, the
and married adult mothers differ in terms of sat- researcher asks for their assistance in getting in
isfaction with inpatient postpartum nursing care. touch with other people with similar characteris-
In this sample, adolescent and adult postpartum tics that meet these criteria.
mothers were matched in terms of parity, mode Chen (2010) recruited participants from pur-
of delivery, infant health status, and infant feeding posive sampling and then, in her second round of
method (Peterson DiCenso, 2002). When an recruitment, used network sampling by asking
organization or institution composes the sampling participants whether they knew any other inter-
unit, matching may also be an important consid- ested women who would be part of the study. The
eration. For example, in a study of the effect of study was about understanding the cultural
an ankle-strengthening and walking exercise context of Chinese mothers’ perceptions of breast-
program on improving fall-related outcomes in feeding and infant health in Canada.
older adults, Schoenfelder and Rubenstein (2004) Today, online computer networks, as described
recruited participants from 10 private, urban in the following section on purposive sampling,
nursing homes in eastern Iowa. Participants were can be used to assist researchers in recruiting
matched in pairs by scores on the Risk Assess- participants who are otherwise difficult to locate,
ment for Falls Scale II and then the members of thereby taking advantage of the networking or
each pair were randomly assigned to opposite snowball effect. The Critical Thinking Decision
groups (the intervention or control condition). Path illustrates the relationship between the
type of sampling strategy and the appropriate
Nonprobability Sampling Strategies Used generalizability.
in Qualitative Research
Because nonprobability sampling is the best
method of obtaining individuals who are key Purposive Sampling
informants of a phenomenon, these sampling Pur osive sa ling is an increasingly common
methods are widely used in qualitative research. strategy in which the researcher’s knowledge of
As described in Chapter 7, qualitative research the population and its elements is used to hand-
methods are conducted to gain both insights into pick the cases to be included in the sample. The
and in-depth meaning about experiences, inci- researcher usually selects participants who are
dents, or events. In qualitative research, the sam- considered typical of the population.
pling procedure is governed by the methodology For example, in a qualitative study (see
used. Many sampling strategies are used in quali- Appendix A), Seneviratne and associates (2009)
tative sampling, but the most common approaches used a subset of purposive sampling, maximum
are net or sampling, purposive sampling, and variation purposive sampling, to locate health
theoretical sampling care professionals in their ethnographic study.
Maximum variation purposive sampling is the
Network Sampling process of deliberately selecting a heteroge-
Net or sa ling sometimes referred to as neous sample and observing commonalities in
sno ball e ect sa ling or sno balling, is a their experiences (Morse, 1994, p. 229).
274 PART FOUR Processes Related to Research

CRITICAL THINKING DECISION PATH

Assessing the Relationship between the Type of Sampling Strategy and
the Appropriate Generalizability

Assess the relationship between the type of

sampling strategy and the appropriate
generalizability

Broader generalizability Limited generalizability

Probability Nonprobability
sampling strategies sampling strategies

Matching tests
of difference random
assignment

Simple Stratified
Multistage
random random Systematic
(cluster)
sampling sampling

Convenience Quota Purposive

Networking

Snowballing

A purposive sample is also used when a highly Delaware Valley Twin Study, Meininger and
unusual group is being studied, such as a popula- associates (1998) examined the differential effect
tion with a rare genetic disease (e.g., Tay-Sachs of cardiovascular risk factors that have the poten-
disease). In this case, the researcher would tial to respond to environmental and lifestyle
describe the sample characteristics precisely to modification. They recruited participant families
ensure that the reader will have an accurate from Mothers of Twins clubs and schools in
picture of the participants in the sample. This type the Philadelphia metropolitan area. Same-sex
of sample can also be used to study the differen- monozygotic and dizygotic twin pairs who met
tial effect of risk factors in a specific population the eligibility criteria were recruited into this
longitudinally. For example, in the longitudinal study.
 CHAPTER 12 Sampling 275

In another situation, the researcher may wish The collection of descriptive data (e.g., as
to interview individuals who re ect a particular in qualitative studies) with which research-
characteristic. For example, Martin-Misener and ers seek to describe the lived experience of
colleagues (2010) studied the role of primary a particular phenomenon (e.g., postpartum
health care nurse practitioners in rural Nova depression, caring, hope, or surviving child-
Scotia. Nine chairpersons, six women and three hood sexual abuse)
men, were selected through the use of purposive The focus of the study population when it
sampling to form a group representative of the is related to a specific diagnosis (e.g., type
health boards in each district health authority. 1 diabetes, multiple sclerosis), a specific
Today, computer networks (e.g., online ser- condition (e.g., legal blindness, terminal
vices) can be of great value in helping researchers illness), or a specific demographic charac-
access and recruit participants for purposive teristic (e.g., same-sex twin pairs)
samples. For instance, Valaitis and associates Many types of purposive sampling exist (Miles
(2008) used an online method to invite 91 schools Huberman, 1994 Patton, 2010), but the follow-
of nursing to participate in their study of the bar- ing three types of cases are the most often used:
riers and enablers in uencing the integration of 1. Typical cases: cases that are normal or
community health nursing content in baccalaure- average among those being studied
ate education in Canada. 2. Deviant or extreme cases: cases that repre-
The researcher who uses a purposive sample sent unusual manifestations of the phenom-
assumes that errors of judgement in overrepre- enon of interest
senting or underrepresenting elements of the 3. Confirming or disconfirming cases: cases
population in the sample will tend to balance each that are exceptions, that represent variation,
other. The validity of this assumption, however, or for which an initial elaborate analysis is
cannot be determined objectively. You must be necessary
aware that the more heterogeneous the popula- In any type of purposive sampling, sampling is
tion, the greater the chance that bias is introduced stopped when data saturation occurs: that is,
in the selection of a purposive sample. As indi- when the information being shared with the
cated in Table 12-1, conscious bias in the selec- researcher becomes repetitive.
tion of participants remains a constant concern. Criterion sampling is also a form of purposive
Therefore, the findings from a study involving a sampling. The researcher needs to have a set of
purposive sample should be regarded with criteria for a sample, and all cases that meet these
caution. As with any nonprobability sample, the criteria are selected. It is important that the crite-
ability to generalize is very limited. The follow- ria are established so that cases that are chosen
ing are several instances when a purposive sample will yield rich data relevant to the research
may be appropriate: problem being explored: for example, all patients
The effective pretesting of newly developed who were in a smoking cessation program and
instruments with a purposive sample of have resumed smoking. This criterion would
diverse types of people enable an understanding of what is needed to
The validation of a scale or test with a support individuals who wish to quit smoking.
known-groups technique
The collection of exploratory data in relation Theoretical Sampling
to an unusual or highly specific population, Theoretical sampling is associated with grounded
particularly when the total target population theory research. As you learned in Chapter 8, the
remains unknown to the researcher goal of grounded research is theory generation
276 PART FOUR Processes Related to Research

thus, a theoretical sampling strategy is used to lation (i.e., a common vs. a rare health
fully elaborate and validate variations in the data problem)
by finding examples of a theoretical construct The projected cost of using a particular
(Sandelowski, 1995). In theoretical sa ling sampling strategy
the researcher selects experiences that will help The sample size should be determined before the
test ideas and gather complete information about study is conducted. A general rule is always to
developing concepts. Sampling is stopped when use the largest sample possible. The larger the
theory saturation or redundancy occurs. sample, the more likely it is to be representative
Theoretical sampling was used by Young and of the population smaller samples produce less
associates (2007), who evaluated the relative accurate results.
effectiveness of telephone and videophone An exception to the rule about sample size is
follow-up for children and families after a child’s the ilot study which is a small sample study
surgery for scoliosis. Young and associates used conducted as a prelude to a larger scale (parent)
memo writing to guide the initial coding of con- study. The pilot study typically is conducted with
cepts and the theoretical sampling process. The similar methods and procedures that both yield
views, situations, and experiences of different preliminary data for determining the feasibility of
family members were compared, and data from conducting a larger scale study and establish that
the same individuals were gathered at different sufficient scientific evidence exists to justify sub-
times. Young and associates engaged in constant sequent, more extensive research.
comparative analyses as new categories arose. Hertzog (2008) summarized methods for jus-
tifying sample sizes on the basis of the aim of the
Helpful Hint pilot study. This author suggests that a sample
Look for a brief discussion of a study’s sampling size as small as 10 to 15 participants per group
strategy in the “Methods” section of a research article. may be sufficient for the decisions being made.
Some articles have a separate subsection with the
heading “Sample,” “Participants,” or “Study Partici-
For pilot studies involving group comparisons, 10
pants.” A statistical description of the characteristics of to 20 participants per group may be enough. On
the actual sample often does not appear until the the other hand, if a researcher is developing or
“Results” section of a research article. testing an instrument, it is suggested that each
group comprise 35 to 40 participants.
For example, Hayward and colleagues (2007)
SAMPLE SIZE: QUANTITATIVE conducted a pilot study, Co-Bedding Twins:
No single rule can be applied to the determination How Pilot Study Findings Guided Improvements
of a sample’s size. When researchers estimate in Planning a Larger Multicenter Trial, using a
sample size, they must consider many factors, sample with 70 babies per group. The researchers
such as the following: analyzed preliminary data to estimate effect size,
The type of design used determine staff and bedside care organization,
The type of sampling procedure used evaluate feasibility of data-collection measures,
The type of formula used for estimating the and identify issues related to recruitment and
optimal sample size follow-up before conducting a multicentre study.
The degree of precision required The principle of larger is better holds true
The heterogeneity of the attributes under for both probability and nonprobability samples.
investigation Results based on small samples (fewer than 10
The relative frequency at which the phe- participants) tend to be unstable the values uc-
nomenon of interest occurs in the popu- tuate from one sample to the next. Small samples
 CHAPTER 12 Sampling 277

tend to increase the probability of obtaining a the averages get closer to the population value,
markedly nonrepresentative sample. As the and the differences in the estimates between
sample size increases, the mean more closely samples A and B also get smaller. Large samples
approximates the population values thus, fewer permit the principles of randomization to work
sampling errors are introduced. effectively (i.e., to counterbalance atypical values
An example of this concept is illustrated by a in the long run).
study in which the average monthly consumption The sample size can be estimated with the use
of sleeping pills was investigated for patients on of a statistical procedure known as po er analy-
a rehabilitation unit after a cerebrovascular acci- sis (see Chapter 16). A simple example illustrates
dent. The data in Table 12-3 indicate that the this concept. Suppose that a researcher wants to
population consisted of 20 patients whose average determine the effect of nurse preoperative teach-
consumption of sleeping pills was 15.2 per month. ing on patient postoperative anxiety. Patients are
The population of 20 patients was divided into randomly assigned to an experimental group or a
sets of two simple random samples with sizes of control group. How many patients should be used
2, 4, 6, and 10. Each sample average in the right in the study When using power analysis, the
column represents an estimate of the population researcher must estimate how large a difference
average, which is known to be 15.15. In most will be observed between the groups (i.e., the
cases, the population value was unknown to the difference in the mean amount of postoperative
researchers, but because the population is so anxiety after the experimental preoperative teach-
small, it could be calculated. In Table 12-3, note ing program). This difference is called the e ect
that with a sample size of two, the estimate might si e If a small difference is expected, the sample
have been wrong by as much as eight sleeping must be large (in this case, 196 patients in each
pills in sample 1B. As the sample size increases, group) to ensure that the differences will be
revealed in a statistical analysis. If a medium-size
TABLE 12-3
difference is expected, the total sample size would
COMPARISON OF POPULATION AND SAMPLE
VALUES AND AVERAGES IN STUDY OF SLEEPING be 128 (64 in each group). When expected differ-
PILL CONSUMPTION ences are large, a small sample size can ensure
NUMBER OF SLEEPING that differences will be revealed through statisti-
PILLS CONSUMED cal analysis.
NUMBER (VALUES EXPRESSED
An example is illustrated by the study of Fox
IN GROUP GROUP MONTHLY) AVERAGE
and colleagues (2010), who examined differences
20 Population 1, 3, 4, 5, 6, 7, 9, 15.2 in sleep complaints among patients with varying
11, 13, 15, 16, 17,
19, 21, 22, 23, amounts of bed rest. Before data collection, they
25, 27, 29, 30 conducted a power analysis based on prior
2 Sample 1A 6, 9 7.5 research. With an alpha value set at .05 and the
2 Sample 1B 21, 25 23.0 power set at .80, the power analysis indicated that
4 Sample 2A 1, 7, 15, 25 12.0 21 participants were required in each of the three
4 Sample 2B 5, 13, 23, 29 17.5 cohorts in order to detect a large effect. Alpha is
6 Sample 3A 3, 4, 11, 15, 21, 25 13.3 the probability of making a Type I error (rejecting
6 Sample 3B 5, 7, 11, 19, 27, 30 16.5 the null hypothesis when the null hypothesis
10 Sample 4A 3, 4, 7, 9, 11, 13, 17, 13.8 is true).
21, 23, 30 Power analysis is an advanced statistical tech-
10 Sample 4B 1, 4, 6, 11, 15, 17, 13.8 nique that is commonly used by researchers and
19, 23, 25, 27
is a requirement for external funding. When
278 PART FOUR Processes Related to Research

power analysis is not used, research studies may

be based on samples that are too small, which Helpful Hint
may lead to a lack of support for the researcher’s Remember to look for some rationale about
the sample size and the strategies that the researcher
hypotheses and to a type I error (rejecting a null has used (e.g., matching, test of differences on demo-
hypothesis when it should have been accepted) graphic variables) to ascertain or build in sample
in other words, the researcher finds significant representativeness.
results when none exist (see Chapter 16). A
researcher may also commit a type II error case study may be sufficient for a phenomeno-
(accepting a null hypothesis when it should have logical study. When you critique a study, you
been rejected) if the sample is too small in other must note how the researcher has explained the
words, the sample is too small to detect treatment sampling plan and what limitations have been
effects (see Chapter 16). noted. Participants are added to the sample until
Despite the principles related to determining data saturation is reached (i.e., new data no longer
sample size that have been identified in this emerge during the data-collection process). The
chapter, you should be aware that large samples fittingness of the data is a more important concern
do not ensure representativeness or accuracy. A than the representativeness of participants (see
large sample cannot compensate for faulty Chapter 14).
research design. The proportion of the population
that is sampled does not provide a guarantee of
accurate results. Accurate results can be obtained Evidence-Informed Practice Tip
from only a small fraction of a large population. Research designs and types of samples are often
For example, a 10 probability sample of a pop- linked. You would expect to see experimental designs
in which probability sampling strategies were used;if a
ulation containing 1500 elements will yield more nonprobability purposive sampling strategy is used to
precise results than will a nonprobability 0.01 recruit participants to such a study, you would expect
sample of a population with 100,000 elements. the participants to then be randomly assigned to inter-
vention and control groups.
You should evaluate the sample size in terms
of (1) how representative the sample is of the
target population and (2) to which population
the researcher wishes to generalize the results of SAMPLING PROCEDURES
the study. The goal of sampling is to gather a The criteria for selecting a sample vary according
sample as representative as possible with as few to the sampling strategy. Regardless of which
sampling errors as possible. strategy is used, the procedure must be systemati-
cally organized. Such organization will eliminate
SAMPLE SIZE: QUALITATIVE the bias that occurs when sample selection is
In qualitative research, no power analyses are carried out inconsistently. Bias in sample repre-
conducted a priori to determine sample size sentativeness and generalizability of findings are
requirements. According to Sandelowski (1995), important sampling issues that have generated
sample size is determined by the purpose and type national concern.
of the sampling and the research method to be For example, many of the landmark adult
used. Morse (1994) recommended about six par- health studies (e.g., the Framingham Heart Study
ticipants for phenomenological studies and about and the Baltimore Longitudinal Study of Aging)
30 to 50 cases for ethnographies and grounded historically excluded women as participants. The
theory studies. These suggestions do not consti- findings of these studies were generalized from
tute a hard-and-fast rule because a one-person men to all adults despite the lack of female
 CHAPTER 12 Sampling 279

representation in the samples. Similarly, the use entire group of people or objects about whom the
of largely Euro-American participants in medica- researcher wants to establish conclusions or make
tion clinical trials limits the identification of generalizations) must be identified. The target
variant responses to drugs in other ethnic or population may consist, for example, of all female
racially distinct groups (Campinha-Bacote, 1997). patients with a first-time diagnosis of breast
Findings based on Euro-American data cannot be cancer, all children with asthma, all pregnant
generalized to Punjabis, Chinese, West Indians, teenagers, or all doctoral students in Canada.
or any other cultural group. Consequently, careful Next, the accessible portion of the target popu-
identification of the target population is a crucial lation must be delineated. An accessible popula-
step in the process. tion might consist of all nurse practitioners in the
In order to establish conclusions about psycho- province of New Brunswick, all male patients
social stressors related to all patients with a with AIDS admitted to a certain hospital during
first-time myocardial infarction, both men and 2001, all pregnant teenagers in a specific prenatal
women must be included in the target population. clinic, or all children with rheumatoid arthritis
To establish conclusions about the incidence under care at a specific hospital specializing in
of extrapyramidal side effects of haloperidol the treatment of autoimmune diseases.
(Haldol) in a psychiatric ward among Chinese Then a sampling plan or a protocol for actually
patients in comparison with Euro-Americans, the selecting the sample from the accessible popula-
target population must be diverse. Sometimes, tion is formulated. The researcher makes deci-
however, the target population must be gender sions about how participants will be approached,
specific, as when breast or prostate cancer or how the study will be explained, and who—the
aspects of pregnancy or menopause are studied. researcher or a research assistant—will select the
Several general steps (Figure 12-3) ensure sample. Regardless of who implements the sam-
the identification of a consistent approach by the pling plan, consistency in how it is done is of
researcher. Initially, the target population (i.e., the paramount importance. In reading a research
report, you want to find a description of the
sample, as well as the sampling procedure, in the
Step 1: study. On the basis of the appropriateness of what
Identify target population has been reported, you can make judgements
about the soundness of the sampling protocol,
which, of course, will affect the interpretations of
Step 2:
the findings.
Finally, once the accessible population and
Delineate the accessible population
sampling plan have been established, permission
is obtained from the institution’s research board,
which is commonly referred to as the research
Step 3:
ethics board This permission provides free access
Develop a sampling plan
to the desired population.
When an appropriate sample size and sam-
pling strategy have been used, the researcher
Step 4:
can feel more confident that the sample is repre-
Obtain approval from research ethics board sentative of the accessible population however,
FIGURE 12-3 Summary of the general sampling proce- it is more difficult to feel confident that the acces-
dures. sible population is representative of the target
280 PART FOUR Processes Related to Research

population. Are nurse practitioners in New Bruns- servative about making sweeping claims in rela-
wick representative of all nurse practitioners in tion to the findings.
Canada It is impossible to know for sure.
Researchers must exercise judgement when Helpful Hint
assessing typicality. nfortunately, no guidelines Remember to evaluate the appropriateness of the
for making such judgements exist, and critiquers generalizations made about the findings of a quantita-
tive study in view of the target population, the accessi-
have even less basis to make such decisions. ble population, the type of sampling strategy, and the
The best rule to use when evaluating the repre- sample size. In qualitative research, evaluate the trans-
sentativeness of a sample and its generalizability ferability of the findings on the basis of the research
design and its sampling strategy and size.
to the target population is to be realistic and con-

APPRAISING THE EVIDENCE

Sample
The criteria for critiquing the sampling technique of a description of the accessible population, the sample, or
study are presented in the Critiquing Criteria box. You (the both. For example, if a researcher states that 100 partici-
reader) and the researcher approach the “Sample” section pants were randomly selected from a population of men
of a research report with different perspectives. You must and women older than 65 and with a diagnosis of COPD
raise the following two questions: who were treated in a respiratory rehabilitation program
1. If this study were to be replicated, is enough infor- at a particular hospital during 2001, you can specifically
mation available about the nature of the popula- evaluate the parameters of the population. The demo-
tion, the sample, the sampling strategy, and the graphic characteristics of the sample (e.g., age, gender,
sample size for another investigator to carry out diagnosis, ethnicity, religion, and marital status) should
the study? also be presented in either a tabular or a narrative
2. Are the previously mentioned factors appropriate summary because they provide further explication about
for the particular research design, and, if not, the nature of the sample and enable you to evaluate the
which factors require modification, especially if the sampling procedure more accurately. For example, in
study is to be replicated? their study on Canadian nurses’ reports of workplace vio-
Sampling is considered to be one important aspect of lence, Lemelin and associates (2009) presented detailed
the methodology of a research study. Thus, data pertain- data summarizing demographic variables of importance.
ing to the sample usually appear in the “Methodology” These data are reproduced as follows:
section of the research report. The sampling content pre-
sented should reflect the outcome of a series of decisions Analysis of the sociodemographic variables showed
that the average age of respondents was 32.63 years
based on sampling criteria appropriate to the design of
65.0% of respondents had less than 10 years of
the study, as well as the options and limitations inherent
nursing experience, and 68.5% of respondents had
in the context of the investigation. The following discus- college-level nursing education. (Lemelin, Bonin, &
sion highlights several sampling criteria that you should Duquette, 2009, p. 158)
consider when you evaluate the merit of a sampling strat-
egy in relation to a specific research study. This example illustrates how a detailed description of
Initially, the parameters or attributes of the study the sample both provides a frame of reference for the
population should clearly specify to what population the study population and sample and generates questions to
findings may be generalized. In general, the target popu- be raised. For example, note the variability in the range
lation of the study is not specifically identified by the of mean risk-adjusted mortality rate. When this demo-
researcher, but the nature of it is implied in the graphic sample information is available, you are able to
 CHAPTER 12 Sampling 281

APPRAISING THE EVIDENCE

Sample—cont’d
question a sampling strategy that does not also account A well-designed, carefully controlled study with a non-
for the differential effect of the size or type of hospital on probability sampling strategy can yield accurate and
the mortality rate, which are factors that might affect the meaningful findings that make a significant contribution
difference in the mortality rate. Also helpful is the to nursing’s scientific body of knowledge. As the critiquer,
researcher’s rationale for having elected to study one type you must ask a philosophical question: “If it is not pos-
of population versus another. For example, why did sible or appropriate to conduct an experimental or
Lemelin and associates (2009) use only nurses in Quebec quasiexperimental investigation with the use of probabil-
in their study? ity sampling, should the study be abandoned?” The
In a research study in which a nonprobability sam- answer usually suggests that it is better to perform the
pling strategy is used, it is particularly important to fully investigation and be fully aware of the limitations of the
describe the population and the sample in terms of who methodology than not to acquire the potential knowl-
the study participants were, how they were chosen, and edge. The researcher is always able to move on to sub-
the reason they were chosen. If these criteria are adhered sequent studies that either replicate the initial study or
to, the degree of heterogeneity or homogeneity of the entail the use of more stringent design and sampling
sample can be determined. The use of a homogeneous strategies to refine the knowledge derived from a nonex-
sample minimizes the amount of sampling error intro- perimental study.
duced, a problem particularly common in nonprobability The greatest difficulty in nonprobability sampling
sampling. stems from the fact that not every element in the popula-
Next, the defined representativeness of the population tion has an equal chance of being represented in the
should be examined. Probability sampling is clearly the sample. Therefore, some segment of the population will
ideal sampling procedure for ensuring the representative- probably be systematically underrepresented. If the pop-
ness of a study population. Use of random selection pro- ulation is homogeneous with regard to critical character-
cedures (e.g., simple random, stratified random, cluster, istics, systematic bias will not be an important problem.
or systematic sampling strategies) minimizes the occur- Few of the attributes that researchers are interested in,
rence of conscious and unconscious biases that affect the however, are sufficiently homogeneous to render sam-
researcher’s ability to generalize about the findings from pling bias an irrelevant consideration.
the sample to the population. You should be able to iden- Next, the sampling plan’s suitability to the research
tify the type of probability strategy used and determine design should be evaluated. In experimental and quasiex-
whether the researcher adhered to the criteria for a par- perimental designs, some form of random selection
ticular sampling plan. In experimental and quasiexperi- or random assignment of participants to groups is used
mental studies, you must also know whether or how the (see Chapter 10). In critiquing the report, you evaluate
participants were assigned to groups. If the criteria have whether the researcher adhered to the principles of
not been followed, you have a valid reason for being random selection and assignment. Lack of adherence to
skeptical about the proposed conclusions of the study. such principles compromises the representativeness of
Random selection is the ideal in establishing the rep- the sample and the external validity of the study. The
resentativeness of a study population; more often, following are questions that you might pose in relation
however, realistic barriers (e.g., institutional policy, inac- to this issue:
cessibility of participants, lack of time or money, and • Has a random selection procedure (e.g., a table of
current state of knowledge in the field) necessitate the random numbers) been identified?
use of nonprobability sampling strategies. Many impor- • Has the appropriate random sampling plan been
tant research problems that are of interest to nurses do selected? In other words, has a proportional strati-
not lend themselves to experimental design and proba- fied sampling plan been selected instead of a
bility sampling, particularly qualitative research designs. simple random sampling plan in a study in which

Continued
282 PART FOUR Processes Related to Research

APPRAISING THE EVIDENCE

Sample—cont’d
three distinct occupational levels appear to be criti- The importance of such examples lies in understand-
cal variables for stratification? ing that this type of statement meets the criteria stated at
• Has the particular random sampling plan been the beginning of the paragraph and should be evident in
carried out appropriately? In other words, if a the research report. Other considerations with regard to
cluster sampling strategy was used, did the sam- sample size, especially when the sample size appears to
pling units logically progress from the largest to the be small or inadequate and no rationale is stated for the
smallest? size, are as follows:
Random sampling should not be regarded as a perfect • How will the sample size affect the accuracy of the
method of obtaining a representative sample. Some- results?
times, bias is inadvertently introduced even when • Are any subsets or cells of the sample overrepre-
random selection is used. In many nonexperimental sented or underrepresented?
designs, nonprobability sampling strategies are used. For • Are any of the subsets so small as to limit mean-
such studies, you can ask whether a nonexperimental ingful comparisons?
design and a related nonprobability sampling plan were • Has the researcher examined the effect of attrition
most appropriate. Sometimes, if the researchers had used on the results?
another type of design or sampling plan, they could have • Has the researcher recognized and identified any
constructed a stronger study that would have produced limitations posed by the size of the sample?
findings that were more generalizable and more reliable. Essentially, these criteria necessitate that you carefully
In critiquing, however, you are rarely in a position to scrutinize several important elements pertaining to
know what factors entered into the decision to plan one sample size that have implications for the generalizability
type of study rather than another. of the findings. Keep in mind that in reports of qualitative
You should then determine whether the sample size studies, neither the predetermining nor the method of
is appropriate and its size is justifiable. The researcher determining the sample size will be discussed. Rather, the
usually indicates in a research article how the sample size sample size depends on the methodology used and is a
was determined;a similar indication is also seen com- function of data saturation (see Chapter 8).
monly in doctoral dissertations. The method of arriving at With qualitative research designs, you apply criteria
the sample size and the rationale should be briefly men- related to sampling strategies that are relevant for a par-
tioned. For example, a researcher may state the following: ticular type of qualitative study. In general, sampling strat-
A power analysis for the multiple regression with 2 egies are purposive because the study of specific
independent variables (p = .05 and power = .80) phenomena in their natural setting is emphasized;any
determined that a sample of 76 was required to participant belonging to a specified group is considered
detect a moderate effect size of 0.15. A sample size to represent that group. For example, in the qualitative
of 300 was surveyed to offset the possibility of a study by Seneviratne and associates (2009;see Appendix
low response rate typically found with mailed A), the specified group was nurses on an acute stroke
surveys, movement of nurses from employers, and unit. The researchers’ goal was to understand the nature
any errors in the registry list. (Armstrong, Laschinger, of nursing on this type of unit—that is, the typicality or
& Wong, 2009, p. 57)
atypicality of the observed events and behaviours—and
In Lemelin and associates’ (2009) study on workplace to reveal nurses’ perceptions of the contexts of caring for
violence reported by Canadian nurses, a power analysis survivors of acute stroke.
was conducted, and Finally, the “Sample” section of the research report
it was determined that 252 respondents were should provide evidence that the rights of human partici-
needed for a medium effect size (.25;alpha = .05; pants have been protected. You will evaluate whether
power = .95). (p. 158) permission was obtained from an institutional research
 CHAPTER 12 Sampling 283

APPRAISING THE EVIDENCE

Sample—cont’d
ethics board that reviewed the study with regard to main- appropriateness of the sampling strategy become crucial
taining ethical research standards (see Chapter 6). For elements in the analysis and interpretation of data, in the
example, the research ethics board examines the research conclusions derived from the findings, and in the gener-
proposal to determine whether the introduction of an alizability of the findings from the sample to the popula-
experimental procedure may be potentially harmful and tion. As stated earlier in this chapter, the major purpose
therefore undesirable. You also examine the report for of sampling is to increase the efficacy of a research study
evidence of the participants’ informed consent, as well as by representing the particular population so that not
protection of their confidentiality or anonymity. Research every element need be studied, while producing the find-
studies that do not demonstrate evidence of having met ings that can be generalized from the sample to the popu-
these criteria are highly unusual. Nevertheless, you will lation. You must demonstrate that the sampling strategy
want to be certain that ethical standards that protect used provided a valid basis for the findings and their
sample participants have been maintained. generalizability.
Many factors must be considered when you critique
the “Sample” section of a research report. The type and

CRITIQUING CRITERIA

1. Have the sample characteristics 6. Would it be possible to 11. Does the researcher identify the
been completely described? replicate the study limitations in generalizability of
2. Can the parameters of the population? the findings from the sample to
study population be inferred 7. How was the sample selected? the population? Are those
from the description of the Is the method of sample limitations appropriate?
sample? selection appropriate? 12. Is the sampling strategy
3. To what extent is the sample 8. What kind of bias, if any, appropriate for the design of
representative of the population is introduced by this the study and level of evidence
as defined? method? provided by the design?
4. Are criteria for eligibility in 9. Is the sample size appropriate? 13. Does the researcher indicate
the sample specifically How is it substantiated? how replication of the study
identified? 10. Does the researcher indicate with other samples would
5. Have sample delimitations that the rights of participants provide increased support for
been established? have been ensured? the findings?

CRITICAL THINKING CHALLENGES How is this possible? Would they have used a
nonprobability or a probability sample? If you
■ A research classmate asks the instructor the agree that this is a legitimate sampling
following question: “Why isn’t it better to study technique, present both the advantages and the
an entire population of patients with lung cancer disadvantages;if you disagree, indicate your
instead of using the research technique of rationale.
sampling?” How would you answer this ■ Your research class is having a debate on
question? Include examples that will help the probability sampling versus nonprobability
student see your point of view. sampling with regard to desirability and
■ In the report of a quasiexperimental study, the feasibility. You are assigned to present the
researchers indicated that they used a advantages of nonprobability sampling in nursing
convenience sample with random assignment. research. What arguments would you use?
284 PART FOUR Processes Related to Research

■ Discuss the principle of “larger is better” and its

probability sampling include simple random,
relationship to network sampling and the sample stratified random, cluster, and systematic sampling.
size of qualitative studies. Include in your • Sample size is a function of the type of sampling
discussion the concept of data saturation and the procedure being used, the degree of precision
use of computer technology. required, the type of sample estimation formula
being used, the heterogeneity of the study
■ Your research classmate is arguing that a random
attributes, the relative frequency of occurrence of
sample is always better, even if it is small and the phenomena under consideration, and the cost.
represents only one site. Another student is • Criteria for selecting a sample vary according to the
arguing that a very large convenience sample sampling strategy. Systematic organization of the
representing multiple sites can be very sampling procedure minimizes bias. The target
population is identified, the accessible portion of
significant. Which classmate would you defend,
the target population is delineated, permission to
and why? conduct the research study is obtained, and a
sampling plan is formulated.
• In critiquing a research report, you evaluate the
sampling plan for its appropriateness in relation to
the particular research design.
KEY POINTS • The completeness of the sampling plan is
examined with regard to the potential replicability
• Sampling is a process in which representative units of the study. In critiquing, you evaluate whether
of a population are selected for study. Researchers the sampling strategy is the strongest plan for the
select representative segments of the population particular study under consideration.
because selecting entire populations of interest to • An appropriate systematic sampling plan will
obtain accurate and meaningful information is maximize the efficiency of a research study. It will
rarely feasible or necessary. increase the accuracy and meaningfulness of the
• Researchers establish eligibility criteria;these are findings and enhance the generalizability of the
descriptors of the population and provide the basis findings from the sample to the population.
for inclusion into a sample. Eligibility criteria can
include age, gender, socioeconomic status, level of
education, religion, and ethnicity.
• The researcher must identify the target population REFERENCES
(i.e., the entire set of cases about which the
Akhtar-Danesh, N., Valaitis, R. K., Schofield, R.,
researcher would like to make generalizations).
Because of pragmatic constraints, however, the nderwood, J., Martin-Misener, R., Baumann, A.,
researcher usually uses an accessible population Kolotylo, C. (2010). A questionnaire for assessing
(i.e., one that meets the population criteria and is community health nurses’ learning needs. estern
available). ournal of Nursing Research, (8), 1055-1072.
• A sample is a set of elements that make up the Armstrong, K., Laschinger, H., Wong, C. (2009).
population. Workplace empowerment and magnet hospital
• A sampling unit is the element or set of elements characteristics as predictors of patient safety
used for selecting the sample. The foremost climate. ournal of Nursing Care uality, (1),
criterion in evaluating a sample is the 55-62.
representativeness or congruence of characteristics
Berk, R. (2003). Regression analysis: A constructive
with the population.
• Sampling strategies consist of nonprobability and criti ue. Thousand Oaks, CA: Sage.
probability sampling. Boscart, V. M. (2009). A communication intervention
• In nonprobability sampling, the elements are for nursing staff in chronic care. ournal of Advanced
chosen by nonrandom methods. Types of Nursing, (9), 1823-1832.
nonprobability sampling include convenience, Campinha-Bacote, J. (1997). nderstanding the
quota, and purposive sampling. in uence of culture. In J. Haber, B. Krainovich-
• Probability sampling is characterized by the Miller, A. L. McMahon, P. Price-Hoskins,
random selection of elements from the population. (Eds.), Comprehensive psychiatric nursing (5th ed.,
In random selection, each element in the pp. 70-95). St. Louis: Mosby.
population has an equal and independent chance
Chahal, N., Manlhiot, C., Colapinto, K., Van Alphen, J.,
of being included in the sample. Types of
McCrindle, B. W., Rush, J. (2009). Association
 CHAPTER 12 Sampling 285

between parental anxiety and compliance with in uences on cardiovascular disease risk factors in
preoperative requirements for pediatric outpatient adolescents. Nursing Research, , 11-18.
surgery. ournal of Pediatric ealth Care, (6), Miles, M. B., Huberman, A. M. (1994). ualitative
372-377. data analysis: An e panded sourceboo (2nd ed.).
Chen, W. (2010). nderstanding the cultural context of Thousand Oaks, CA: Sage.
Chinese mothers’ perceptions of breastfeeding and Morse, J. M. (1994). Designing funded qualitative
infant health in Canada. ournal of Clinical Nursing, research. In Denzin, N. K., Lincoln, Y. S. (Eds.),
, 1021-1029. andboo of ualitative research (pp. 220-235).
Chiovitti, R. F. (2006). Nurses’ meaning of caring with Thousand Oaks, CA: Sage.
patients in acute psychiatric hospital settings: A Ostry, A., Maggi, S., Hershler, R., Chen, L.,
grounded theory study. International ournal of Hertzman, C. (2010). Mental health differences
Nursing Studies, (2), 203-223. among middle-aged sawmill workers in rural
Cho, S. H., Ketefian, S., Barkauskas, V. H., Smith, compared to urban British Columbia. Canadian
D. G. (2003). The effects of nurse staffing on adverse ournal of Nursing Research, (3), 84-100.
events, morbidity, mortality, and medical costs. Patton, M. . (2010). Evaluation checklists. Retrieved
Nursing Research, , 71-79. from http://www.wmich.edu/evalctr/checklists/
Cochran, W. G. (1977). Sampling techni ue (3rd ed.). checklist topics.
New York: Wiley. Peterson, W., DiCenso, A. (2002). A comparison of
Curtis, E. A. (2008). The effects of biographical adolescent and adult mothers’ satisfaction with their
variables on job satisfaction among nurses. British postpartum nursing care. Canadian ournal of
ournal of Nursing, (3), 174-180. Nursing Research, , 117-127.
Fox, M. T., Sidani, S., Brooks, D. (2010). Differences Samuels-Dennis, J., Ford-Gilboe, M., Ray, S. (2010).
in sleep complaints in adults with varying levels of Single mother’s adverse and traumatic experiences
bed days residing in extended care facilities for and post-traumatic stress symptoms. ournal of
chronic disease management. Clinical Nursing amily Violence, (1), 9-20.
Research, (2), 181-202. Sandelowski, M. (1995). Sample size in qualitative
Hayward, K., Campbell-Yeo, M., Price, S., Morrison, research. Research in Nursing ealth, , 179-183.
D., Whyte, R., Cake, H., Vine, J. (2007). Schoenfelder, D. P., Rubenstein, L. M. (2004). An
Co-bedding twins: How pilot study findings guided exercise program to improve fall-related outcomes in
improvements in planning a larger multicenter trial. elderly nursing home residents. Applied Nursing
Nursing Research, , 137-143. Research, (1), 21-31.
Hertzog, M. A. (2008). Consideration in determining Seneviratne, C. C., Mather, C. M., Then, K. L.
sample sizes for pilot studies. Research in Nursing (2009). nderstanding nursing on an acute stroke
ealth, , 180-191. unit: Perceptions of space, time and interprofessional
Kaasalainen, S., Crook, J. (2004). An exploration of practice. ournal of Advanced Nursing, (9),
seniors’ ability to report pain. Clinical Nursing 1872-1881. doi: 10.1111/j.1365-2648.2009.05053.x
Research, , 199-215. Sinclair, B., Ferguson, K. (2009). Integrating
Lemelin, L., Bonin, J.-P., Duquette, A. (2009). simulated teaching/learning strategies in undergradu-
Workplace violence reported by Canadian nurses. ate nursing education. International ournal of
Canadian ournal of Nursing Research, (3), Nursing Education and Scholarship, (1), Article 7.
152-167. Smith, P. M., Corso, L., Brown, K. S., Cameron, R.
Linton, A., Singh, M., Turbow, D., Legg, T. J. (2009). (2011). Nurse case-managed tobacco cessation
Street youth in Toronto: An investigation of demo- interventions for general hospital patients: Results of
graphic predictors of high risk behaviors and HIV a randomized clinical trial. Canadian ournal of
status. ournal of IV AIDS Social Services, , Nursing Research, (1), 98-117.
375-396. Sousa, V. D., Zauszniewski, J. A., Musil, C. M. (2004).
Martin-Misener, R., Reilly, S. M., Vollman, A. R. How to determine whether a convenience sample
(2010). Defining the role of primary health care nurse represents the population. Applied Nursing Research,
practitioners in rural Nova Scotia. Canadian ournal 17(2), 130-133.
of Nursing Research, (2), 30-47. lrich, C. M., Danis, M., Ratcliffe, S. J., Garrett-Mayer,
Meininger, J. C., Hayman, L. L., Coates, P. M., E., Koziol, D., Soeken, K. L., Grady, C. (2006)
Gallagher, P. R. (1998). Genetic and environmental Ethical con ict in nurse practitioners and physician
286 PART FOUR Processes Related to Research

assistants in managed care. Nursing Research, (6), Testing an intervention for preventing osteoporosis in
391-401. postmenopausal breast cancer survivors. ournal of
Valaitis, R. V., Rajsic, C. J., Cohen, B., Stamler, L. L., Nursing Scholarship, , 333-338.
Meagher-Stewart, D., Froude, S. A. (2008). Young, L., Siden, H., Tredwell, S. (2007). Post-
Preparing the community health nursing workforce: surgical telehealth support for children and family
Internal and external enablers and challenges care-givers. ournal of Telemedicine and Telecare,
in uencing undergraduate nursing programs in , 15-19.
Canada. International ournal of Nursing Education
Scholarship, (1), Article 22.
Vlack, S., Foster, R., Menzies, R., Williams, G.,
Shannon, D., Riley, I. (2007). Immunisation
coverage of ueensland indigenous two-year-old
FOR FURTHER STUDY
children by cluster sampling and by register. Go to Evolve at http://evolve.elsevier.com/
Australian Ne ealand ournal of Public ealth, Canada/LoBiondo/Research for Audio Glossary, how-to
(1), 67-72. instructions for Writing Proposals for Funding, and
Waltman, N. L., Twiss, J. J., Ott, C. D., Gross, G. J., additional research articles for practice in reviewing
Lindsey, A. M., Moore, T. E., Berg, K. (2003). and critiquing.
C H A PTER 1 3

Data-Collection Methods
Susan Sullivan-Bolyai | Carol Bova | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Define the types of data-collection methods used in nursing research.
• List the advantages and disadvantages of each of these methods.
• Compare how specific data-collection methods contribute to the strength of evidence in a research
study.
• Critically evaluate the data-collection methods used in published nursing research studies.

KEY TERMS
biological measurement intervention physiological measurement
closed-ended item intervention fidelity questionnaire
concealment interview reactivity
consistency Likert-type scale records or available data
content analysis measurement scale
debriefing objective scientific observation
external criticism open-ended item social desirability
internal criticism operational definition systematic
interrater reliability operationalization

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

287
288 PART FOUR Processes Related to Research

NURSES USE ALL OF THEIR SENSES when collecting 14). This aspect of the research process necessi-
data from the patients to whom they provide care. tates painstaking effort from the researcher. Thus,
Nurse researchers also have many ways to collect the process of evaluating and selecting the avail-
information about their research participants. able tools to measure variables of interest is
Both the data collected when they perform patient crucial for the potential success of the study. In
care and the data collected for the purpose of this section, the selection of measures and the
research are objective and systematic. Ob ective implementation of the data collection process are
means that the data must not be in uenced by the discussed. An algorithm that in uences a research-
person who collects the information, and syste er’s choice of data collection methods is dia-
atic means that the data must be collected in the grammed in the Critical Thinking Decision Path.
same methodical way by each person involved Information about phenomena of interest to
in the collection procedure. The methods that nurses can be collected in many different ways.
researchers use to collect information about par- Nurses are interested in the biological and physi-
ticipants are the identifiable and repeatable opera- cal indicators of health (e.g., blood pressure and
tions that define the major variables being studied. heart rate), but they are also interested in complex
O erationali ation is the process of translat- psychosocial questions presented by patients.
ing the concepts of interest to a researcher into Psychosocial variables, such as anxiety, hope,
observable and measurable phenomena. The social support, and self-concept, may be mea-
same information may be collected in a number sured by several different techniques, such as
of ways. For example, Horgas and colleagues observation of behaviour, self-reports of feelings,
(2008) defined disability as having both physical or self-reports about attitudes in interviews or
and social functional limitations, and they opera- questionnaires. To study variables of interest,
tionally measured this variable by using the Sick- researchers also may use data that have already
ness Impact Profile. been collected for another purpose, such as
This purpose of this chapter is to familiarize records, diaries, or other media.
you with the various ways in which researchers The data-collection method must be appropri-
collect information from and about participants. ate to the problem, the hypothesis, the setting, and
The chapter provides nurse readers with the tools the population. For example, Van Cleve and asso-
for evaluating the selection, use, and practicality ciates (2004) were interested in studying the pain
of the various ways to collect data. experiences of children across multiple age
groups. Because they were studying children aged
MEASURING VARIABLES OF INTEREST 4 to 17 years, the same type of instrument could
To a large extent, the success of a study depends not be used for all the children. To deal with the
on the quality of the data-collection methods reading and development levels of the children,
chosen and employed. Researchers have many they used tested age-appropriate instruments. For
types of methods available for collecting informa- example, the children who were 4 to 13 years of
tion from participants in research studies. ea age used the poker chip tool, in which the child
sure ent is the assignment of numbers to objects chooses from one to four red chips to represent a
or events according to rules determining which little bit to the most pain the child can experi-
measurement to use in a particular investigation ence. The older children, however, used the ado-
may be the most difficult and time-consuming lescent paediatric pain tool, in which words and
step in the study design. In addition, nurse graphics are used to rate the pain experience.
researchers have an array of quality instruments Selection of the data collection method begins
with adequate reliability and validity (see Chapter during the literature review. As noted in Chapter 5,
 CHAPTER 13 Data-Collection Methods 289

CRITICAL THINKING DECISION PATH

Data Collection Methods

Is the concept to be studied . . .

Physiological data? Complex environmental data? Self-report data?

Use physiological instrument Use observation Discrete content?

Discrete content? Yes No

Yes No Questionnaire Unstructured

or structured interview
interview
Observational tool Field notes

one purpose of the literature review is to provide physiological. If researchers are interested in
clues about instrumentation. As the literature studying stressors, they must first define what
review is conducted, the researcher begins to they mean by the concept of stressor, both
explore how previous investigators defined and conceptually and operationally. uality-of-life
operationalized variables similar to those of inter- research is popular with researchers from many
est in the current study. The researcher uses this disciplines, including nursing. Definitions of
information to define conceptually the variables quality of life may be related to health function-
to be studied. Once a variable has been defined ing, life satisfaction, or well-being.
conceptually, the researcher returns to the litera- uality of life may also be interpreted in a
ture to define the variable operationally: that is, general way (well-being) or be related specifi-
describe how a concept is measured and what cally to a type of illness. Therefore, if researchers
instruments are used to capture the essence of the are interested in studying quality of life, they
variable. This o erational de nition translates need first to define what they mean by the concept
the conceptual definition into behaviours or ver- of quality of life. For example, Molzahn (2007)
balizations that can be measured for the study. In was interested in quality of life as it relates to
this second literature review, the researcher spirituality in later life. Molzahn wrote that this
searches for measurement instruments that might concept has many definitions but chose the World
be used as is or adapted for use in the study. If Health Organization’s definition: uality of life
instruments are available, the researcher must is the individuals’ perception of their position in
obtain the author’s permission for their use. life in the context of the culture and value systems
The following examples illustrate the relation- in which they live and in relation to their goals,
ship of conceptual and operational definitions. expectations, standards and concerns (as cited
Stress research is of interest to researchers from in Molzahn, 2007, p. 35). According to this
many disciplines, including nursing. Definitions conceptual definition, the researcher would use
of stressors may be psychological, social, or a quality-of-life instrument specifically about
290 PART FOUR Processes Related to Research

spirituality to determine the perceived quality of trained and supervised. To ensure consistency in
life of participants in the study. If another data collection, sometimes referred to as inter
researcher disagreed with this definition or was vention delity (Santacroce, Maccarelli, Grey,
more interested in the quality of life of people 2004), researchers must train data collectors in
with a specific illness or the quality of life of the methods to be used in the study so that each
children, a different instrument may be more data collector acquires the information in the
appropriate. same way. Information about how to observe, ask
Sometimes no suitable measuring device questions, and collect data often is included in a
exists, and so the researcher must then decide kind of cookbook protocol or manual for the
how important the variable is to the study and research project. A researcher needs to spend time
whether a new device should be constructed. The developing the protocol and training data collec-
construction of new instruments for data collec- tors to gather data systematically and reliably.
tion that have reasonable reliability and validity Comments about their training and the consis-
(see Chapter 14) is a difficult task. If no suitable tency with which they collected data for the study
measuring device exists, the researcher may should be provided by the researcher.
decide not to study a variable, or the researcher An example of intervention fidelity is given in
may decide to invest time and energy in instru- the study by Ratner and colleagues (2004), who
ment development. Either decision is acceptable, examined the efficacy of a smoking cessation
depending on the goals of the study and the goals intervention for patients about to undergo elective
of the researcher. surgery. The researchers used several ways to
ensure fidelity: (1) structured and rigorous train-
Helpful Hint ing of research staff (2) role playing to evaluate
Remember that the researcher may not always the competence of the study’s registered nurses
present complete information about the way the data (3) checks every 3 to 6 months to assess the
were collected, especially when established tools were
used. To learn about the tool that was used, the reader
extent of drift in role playing (4) regular staff
may need to consult the original article that described meetings to review the protocol and to address
the use or development of the tool. complex situations (5) checklists for every timed
intervention to ensure that all components of the
Whether the researcher uses available methods intervention were covered and (6) a video record-
or creates new ones, once the variables have been ing of an enactment of the intervention protocol
operationally defined in a manner consistent with for review by the registered nurses.
the aims of the study, the population to be studied, Another example of the importance of training
and the setting, the researcher decides how the data collectors appears in the study by Doran and
data-collection phase of the study will be imple- associates (2006) on data collection of nursing-
mented. This decision concerns how the instru- sensitive outcomes in acute care and long-term
ments for data collection will be given to the care settings. Doran and associates needed to
participants. Consistency is the most important assess nursing-sensitive outcomes accurately.
issue in this phase. Staff nurses were trained using didactic content
Consistency in data collection means that the and case studies on how to collect data on patient
method used to collect data from each participant outcomes by research assistants. To assess inter-
in the study is exactly the same or as close to the rater reliability, the research assistants (raters)
same as possible. Consistency can minimize the conducted an independent assessment of three to
bias introduced when more than one person col- five patients for each nurse over the 6-month
lects the data. Data collectors must be carefully period of data collection, and agreement between
 CHAPTER 13 Data-Collection Methods 291

raters was calculated with the kappa statistic. The different ways, researchers need to measure these
index of agreement in this study ranged from .64 outcomes at similar intervals and in similar ways
to .93, and on average, the degree of agreement for all participants of the study.
between rates was 86 thus, the level of inter- Physiological easure ent and biological
rater agreement between the two observers, easure ent involve the use of specialized
nurses and research assistants, was high. Inter equipment to determine the physical and biologi-
rater reliability (see Chapter 14) is the consis- cal status of participants. Frequently, such mea-
tency of observations between two or more surements also require specialized training. These
observers. It is often expressed as a percentage of measurements can be physical, such as weight or
agreement among raters or observers or as a coef- temperature chemical, such as blood glucose
ficient of agreement that considers the element of level microbiological, as with cultures or ana-
chance (coefficient kappa). tomical, as in radiological examinations. What
distinguishes these measurements from others
Evidence-Informed Practice Tip used in research is that special equipment is
It is difficult to place confidence in a study’s find- needed to make the observation. A researcher can
ings if the data-collection methods are not consistent.
say, This participant feels warm, but to deter-
mine how warm the participant is requires the use
TYPES OF DATA-COLLECTION METHODS of a sensitive instrument: a thermometer.
In general, data-collection methods can be divided Physiological or biological measurement is
into the following five types: physiological mea- particularly suited to the study of several types of
surements, observational methods, intervie s, nursing problems. Bryanton and colleagues’
uestionnaires, and records or available data (2004) example is typical of studies dealing with
Each method has a specific purpose, as well as ways to improve the performance of certain
certain advantages and disadvantages inherent nursing actions, such as the measuring and record-
in its use. In the following sections, these data- ing of patients’ physiological data. Physiological
collection methods are discussed, along with their measures may yield important criteria for deter-
respective uses and problems. mining the effectiveness of certain nursing inter-
ventions. In the study of the effect of types of
Physiological or Biological Measurements bathing and newborns’ temperatures, Bryanton
In everyday practice, nurses collect physiological and colleagues (2004) reported that tub-bathed
data about patients, such as their temperature, infants had significantly less temperature loss
pulse rate, blood pressure, blood glucose level, than did sponge-bathed infants.
urine specific gravity, and pH of bodily uids. The advantages of using physiological data-
Such data are frequently useful to nurse research- collection methods include their objectivity, pre-
ers. For example, Bryanton and colleagues (2004) cision, and sensitivity. Such methods are generally
compared the effects of tub bathing versus considered to yield objective findings because
traditional sponge bathing for healthy, full-term unless a technical malfunction occurs, two read-
newborns and their mothers’ ratings of pleasure ings of the same instrument taken at the same time
and confidence. The physiological variable of by two different nurses are likely to yield the same
newborn temperature was one of the outcome result. Because such instruments are intended
variables and was measured by one route, the to measure the variable being studied, they offer
axillary route, for standardization. Because phys- the advantage of being precise and sensitive
iological variables, such as cardiac output and enough to pick up subtle variations in the variable
blood pressure, can be measured in several of interest. Also, the deliberate distortion of
292 PART FOUR Processes Related to Research

physiological information by a participant in a Although observing the environment is a

study is highly unlikely to occur. normal part of living, scientific observation places
Physiological measurements are not without a great deal of emphasis on the objective and
inherent disadvantages, however. Some instru- systematic nature of the observation. The
ments, if not available through a hospital, may be researcher is not merely watching what is hap-
quite expensive to obtain and use. In addition, the pening but is watching with a trained eye for
accurate use of such instruments often necessi- certain specific events. Scienti c observation
tates specialized knowledge and training. Another fulfills the following four conditions:
problem with physiological measurements is that 1. The observations undertaken are consistent
simply by using them, the variable of interest may with the study’s specific objectives.
be changed. Although some researchers think of 2. A standardized and systematic plan exists
these instruments as being nonintrusive, the pres- for the observation and the recording of
ence of some types of devices might change the data.
measurement. For example, the presence of a 3. All of the observations are checked and
heart rate monitoring device might make some controlled.
patients anxious and thereby increase their heart 4. The observations are related to scientific
rate. In addition, nearly all types of measuring concepts and theories.
devices are affected in some way by the environ- Observation is particularly suitable as a data-
ment. Even a simple thermometer can be affected collection method in complex research situations
by the participant’s drinking something hot or that are best viewed as total entities and that are
cold immediately before the temperature is taken. difficult to measure in parts, such as studies
Thus, you need to consider whether the researcher dealing with the nursing process, parent-child
controlled such environmental variables in the interactions, or group processes (see the Practical
study. Finally, a physiological way to measure the Application box for an example). In addition,
variable of interest may not exist. On occasion, observational methods can be the best way to
researchers try to force a physiological parameter operationalize some variables of interest in
into a study in an effort to increase the precision nursing research studies, particularly individual
of measurement. If the device does not measure characteristics and conditions, such as traits and
the variable of interest, however, the validity of
the device’s use is suspect.

Observational Methods Practical Application

Sometimes nurse researchers are interested in Seneviratne and associates (2009; see
Appendix A) conducted an ethnographic
determining how participants behave under spe- study to understand nursing on an acute stroke
cific conditions. For example, the researcher unit in regard to perceptions of space, time, and
might be interested in how children respond to interprofessional practice. They made observations of
2 to 3 hours on 3 days of the week. The field worker
painful situations. We might ask children how made observations during every type of shift.
painful an experience was, but they may not be The field work began with 3 months of general
able to answer the question or to quantify the observation on the unit, in which the field worker
watched at the charting desk and nursing station and
amount of pain, or they may distort their responses
while walking in the hall. Notes from observations
to please the researcher. Therefore, sometimes were transcribed by means of computer. The field
observing the participant may yield a more accu- worker clarified gaps in field notes by returning to
rate description of the behaviour in question than the field site and making more focused observations
driven by the key informants.
does asking the patient.
 CHAPTER 13 Data-Collection Methods 293

symptoms verbal and nonverbal communication observations involve the use of hidden television
behaviours, activities, and skill attainment and cameras, audio recordings, or one-way mirrors.
environmental characteristics. Concealment without intervention is often used
Observational methods can also be distin- in observational studies of children. You may be
guished by the role of the observer. This role is familiar with rooms with one-way mirrors through
determined by the amount of interaction between which a researcher can observe the behaviour of
the observer and the people being observed. the occupants of the room without being observed
Each of the following four basic types of obser- by them. Such studies allow the observation of
vational roles is distinguishable by the amount of children’s natural behaviour and are often used in
concealment or intervention implemented by the developmental research.
observer: Observational studies commonly involve no
1. Concealment without intervention concealment and no intervention. In this case, the
2. Concealment with intervention researcher obtains informed consent from the par-
3. No concealment without intervention ticipant to be observed and then simply observes
4. No concealment with intervention the participant’s behaviour.
These methods are illustrated in Figure 13-1 Observing participants without their knowl-
examples are given later. Conceal ent refers to edge may violate assumptions of informed
is a study method in which participants do not consent therefore, researchers face ethical prob-
know that they are being observed and in inter lems with this type of approach. However,
vention the observer provokes actions from researchers sometimes have no other way to
those who are being observed. collect such data, and the data collected are
In the study by Seneviratne and associates unlikely to have negative consequences for the
(2009), the field worker was not concealed while participant. In these cases, the disadvantages of
observing the nurses on the acute stroke unit. the study are outweighed by the advantages. Fur-
When a researcher is concerned that the par- thermore, the problem of consent is often handled
ticipants’ behaviour will change as a result of by informing participants after the observation
being observed (reactivity), the type of observa- and allowing them the opportunity to refuse to
tion most commonly employed is that of conceal- have their data included in the study and to
ment without intervention. In this case, the discuss any questions they might have. This
researcher watches the participants without their process is called debrie ng
knowledge of the observation and does not When the observer is neither concealed nor
provoke them into action. Often, such concealed intervening, the ethical question is not a problem.

Concealment
Yes No

Researcher hidden Researcher open

Yes
Intervention

An intervention An intervention

Researcher hidden Researcher open

No
No intervention No intervention

FIGURE 13-1 Types of observational roles in research.

294 PART FOUR Processes Related to Research

Here, the observer makes no attempt to change Another type of unstructured observation is
the participants’ behaviour and informs them that the use of anecdotes. Despite popular usage of
they are to be observed. Because the observer is this term, anecdotes are not necessarily funny but
present, this type of observation allows a greater usually focus on the behaviours of interest and
depth of material to be studied than if the observer frequently add to the richness of research reports
is separated from the participants by an artificial by illustrating a particular point. In the study by
barrier, such as a one-way mirror. In a commonly Smith and associates (2006), interviews were
used observational technique, the researcher conducted, and field notes were made and incor-
functions as part of a social group to observe the porated into the data set of this very innovative
participants. The problem with this type of obser- exploration of the topic.
vation is reactivity (also referred to as the Haw- In contrast, structured observations, such as
thorne effect see Chapter 9), or the distortion the standardized tools used to evaluate mother-
created when the participants change behaviour infant interaction in Koniak-Griffin and col-
because they know they are being observed. leagues’ (2003) study, require formal training and
In their study, Seneviratne and associates the competence of the evaluators. The use of
(2009) used unconcealed observation because the structured observations without a standardized
nurses and patients had given full consent for tool involves specifying in advance what behav-
participation in the study. No concealment with iours or events are to be observed and preparing
intervention is used when the researcher is observ- forms for record keeping, such as categorization
ing the effects of an intervention introduced for systems, checklists, and rating scales. Whichever
scientific purposes. Because the participants system is employed, the observer watches the
know they are participating in a research study, participant and then marks on the recording form
few problems with ethical concerns occur, but what was seen. In both cases, the observations
reactivity is a problem with this type of study. must be similar among the observers (see the
Concealed observation with intervention earlier discussion and Chapter 14 for an explana-
involves staging a situation and observing the tion of interrater reliability). Thus, observers need
behaviours that are evoked in the participants as to be trained to be consistent in their observations
a result of the intervention. Because the partici- and ratings of behaviour.
pants are unaware of their participation in a
research study, this type of observation has fallen Evidence-Informed Practice Tip
into disfavour and is rarely used in nursing When you read a research report that uses obser-
research. vation as a data-collection method, you will want to
note evidence of consistency across data collectors
Observational methods may be structured or through use of internal consistency reliability data in
unstructured. nstructured observational methods quantitative research and credibility in qualitative
are not characterized by a total absence of struc- research. When that evidence is present, you can have
greater confidence in the results.
ture but usually involve collecting descriptive
information about the topic of interest. In unstruc-
tured observations, the observer keeps field notes Scientific observation has several advantages
that record the activities, as well as the observer’s as a data-collection method. The main advantage
interpretations of these activities. Field notes is that observation may be the only way for the
are usually not restricted to any particular type researcher to study the variable of interest. For
of action or behaviour rather, they are intended example, what people say they do is often not
to depict a social situation in a more general what they really do. Therefore, if the study is
sense. designed to obtain substantive findings about
 CHAPTER 13 Data-Collection Methods 295

human behaviour, observation may be the only a participant verbally. Interviews may be face to
way to ensure the validity of the findings. In addi- face or performed over the telephone, Skype, or
tion, no other data-collection method can match other electronic means, and may consist of open-
the depth and variety of information that can be ended or closed-ended questions. In contrast, the
collected with the techniques of scientific obser- uestionnaire is an instrument designed to gather
vation. Such techniques are also exible in that data from individuals about knowledge, attitudes,
they may be used in both experimental and non- beliefs, and feelings. Survey research relies
experimental designs and in laboratory and field almost entirely on questioning participants with
studies. either interviews or questionnaires, but these
methods of data collection can also be used in
Helpful Hint other types of research.
Sometimes researchers carefully train observers No matter what type of study is conducted, the
or data collectors, but the research report does not purpose of questioning participants is to seek
address this training. The limitations on length of
research reports often prevent the inclusion of certain information. This information may be of either
information. Readers can often assume that if reliability direct interest, such as the participant’s age, or
data are provided, then appropriate training occurred. indirect interest, such as when the researcher uses
a combination of items to estimate the degree to
As with all data-collection methods, observa- which the respondent has a particular trait or
tion also has its disadvantages. Earlier in this characteristic. An intelligence test is an example
chapter, the problems of reactivity and ethical of how individual items are combined with
concerns were mentioned with regard to conceal- several others to develop an overall scale of intel-
ment and intervention. In addition to these prob- ligence. When items of indirect interest on a
lems, data obtained by observational techniques survey or questionnaire are combined to obtain
are vulnerable to the bias of the observer. Emo- an overall score, the measurement tool is called
tions, prejudices, and values can in uence the a scale
way that behaviours and events are observed. In The investigator determines the content of an
general, the more the observer needs to make interview or questionnaire from the literature
inferences and judgements about what is being review (see Chapter 5). When evaluating inter-
observed, the more likely it is that distortion will views and questionnaires, you should consider
occur. Thus, in judging the adequacy of observa- the content of the scale, the individual items, and
tional methods, you will need to consider how the order of the items. The basic standard for
observational tools were constructed and how evaluating the individual items in an interview or
observers were trained and evaluated. questionnaire is that the item must be clearly
written so that the intention of the question and
Interviews and Questionnaires the nature of the information sought are clear
Participants in a research study often have infor- to the respondent. The only way to know whether
mation that is important to the study and that can the questions are understandable to the respon-
be obtained only by asking the participants. Such dents is to pilot test them in a similar population.
questions may be asked through the use of inter- It is also critical not to rely on only the instrument
views and questionnaires. For both, the purpose developer’s reports of reliability and validity (see
is to ask participants to report data for them- Chapter 14). A pilot test allows researchers to test
selves, but each method has unique advantages the reliability and validity for their unique sample
and disadvantages. The intervie is a method of rather than relying only on previously reported
data collection in which a data collector questions results.
296 PART FOUR Processes Related to Research

Although each questionnaire item must consist

Evidence-Informed Practice Tip
of only one question or concept, be free of sug-
Scales used in nursing research should have evi-
gestions, and be worded with correct grammar, dence of adequate reliability and validity so that readers
such items may be either open-ended or close- feel confident that the findings reflect what the
ended. An o en ended ite is used when the researcher intended to measure (see Chapter 14).
researcher wants the participants to respond in
their own words or when the researcher does not
know all of the possible alternative responses. A Figure 13-2 shows a few items from a fictional
closed ended ite is a question that the respon- survey of paediatric nurse practitioners. The first
dent may answer with only one of a fixed number items are taken from a list of similar items, and
of alternative responses. Many scales use a they are both closed-ended and of a Likert-type
fixed-response format called a Likert-type scale. format. Note that respondents are asked to choose
A Li ert ty e scale is a list of statements for how strongly they agree with each item. In using
which responses are varying degrees of agree- these questions in the survey, respondents are
ment or opinion for example, whether respon- forced to choose from only these answers because
dents strongly agree, agree, disagree, or it is thought that these will be the only responses.
strongly disagree. Sometimes finer distinctions The only possible alternative response is to skip
are given, or a neutral category (e.g., no opinion ) the item, leaving it blank.
may be provided. The use of the neutral category, Sometimes researchers have no idea or only a
however, sometimes creates problems because it limited idea of what the respondent will say, or
is often the most frequent response and is difficult researchers want the answer in the respondent’s
to interpret. Fixed-response items also can be own words, as with the second (open-ended) set
used for questions requiring a yes or no of items. In this situation, respondents may also
response or when the interview or questionnaire leave the item blank but are not forced to make a
has categories, as with income. particular response.
Structured, fixed-response items are best used Interviews and questionnaires are commonly
when the question has a finite number of responses used in nursing research. Both are strong
and the respondent is to choose the option closest approaches to gathering information for research
to the right response. Fixed-response items have because they enable the researcher to approach
the advantage of simplifying the respondent’s the task directly. In addition, both can elicit
task and the researcher’s analysis, but they may certain kinds of information, such as the partici-
miss some important information about the par- pants’ attitudes and beliefs, that would be difficult
ticipant. Sword and colleagues (2006) used self- to obtain without asking the participant directly.
reported fixed-response items via a structured All methods that involve verbal reports,
telephone interview in their study of experiences however, share a problem with accuracy. Often,
of immigrant and Canadian-born women with it is impossible to know whether what the
regard to postpartum health, service needs, and researcher is told is indeed true. For example,
access to care. nstructured response formats people are known to respond to questions in a
allow such information to be included, but a way that makes a favourable impression. This
special technique is needed to analyze the response style is known as social desirability
responses. This technique, called content analy which can be regarded as resulting from two
sis is a method for the objective, systematic, and factors: self-deception and other-deception.
quantitative description of communications and Neyerhof (2006) discussed the two main
documentary evidence. modes of coping with social desirability bias. The
 CHAPTER 13 Data-Collection Methods 297

Close-Ended (Likert-Type Scale)

A. How satisfied are you with your current position?
1. Very satisfied
2. Moderately satisfied
3. Undecided
4. Moderately satisfied
5. Very dissatisfied
B. To what extent do the following factors contribute to your current level of positive
satisfaction?

Very Moderate A great

Not at all little Somewhat amount deal

1. % of time in patient
1 2 3 4 5
care

2. Type of patients 1 2 3 4 5

3. % of time in
educational activity 1 2 3 4 5

4. % of time in
1 2 3 4 5
administration

Close-Ended
A. On average, how many patients do you see in one day?
1. 1 to 3
2. 4 to 6
3. 7 to 9
4. 10 to 12
5. 13 to 15
6. 16 to 18
7. 19 to 20
8. More than 20
B. How would you characterize your practice?
1. Too slow
2. Slow
3. About right
4. Busy
5. Too busy

Open-Ended
A. Are there incentives that the National Association of Pediatric Nurse Associates and Practitioners
ought to provide for members that are not currently being provided?

FIGURE 13-2 Examples of close-ended and open-ended questions.

298 PART FOUR Processes Related to Research

first mode is aimed at the detection and measure- of nursing practice. This use of multiple measures
ment of social desirability bias and is represented provides a more complete picture than the use of
by two methods: the use of social desirability just one measure.
scales and the rating of item desirability. The When determining whether to use interviews
second mode is aimed at preventing or reducing or questionnaires, researchers face difficult
social desirability bias and is represented by the choices. The final decision is often based on the
following methods: forced-choice items, the ran- instruments available and their relative costs and
domized response technique, the bogus pipeline, benefits.
self-administration of the questionnaire, the Both face-to-face and telephone interviews
selection of interviewers, and the use of proxy have some advantages over questionnaires. The
participants. Neyerhof found that no one method rate of response to interviews is almost always
excelled completely and suggested that a combi- better than that to questionnaires, which helps
nation of prevention and detection methods is the eliminate bias in the sample (see Chapter 12).
best strategy to reduce social desirability bias. Respondents seem to be less likely to hang up
uestionnaires and interviews also have some the telephone or to close the door in an inter-
specific purposes, advantages, and disadvantages. viewer’s face than to throw away a questionnaire.
uestionnaires are useful tools when the purpose Another advantage of the interview is that some
is to collect information. If questionnaires are too people—such as young children, people with vis-
long, however, respondents are not likely to com- ual impairments, and people who are illiterate—
plete them. uestionnaires are most useful when cannot fill out a questionnaire but can participate
the set of questions to be asked is finite and the in an interview. With an interview, the data col-
researcher can be assured of the clarity and lector knows who is giving the answers. When
specificity of the items. Face-to-face techniques questionnaires are mailed, for example, anyone in
or interviews are most appropriate when the the household could be the person who supplies
researcher may need to clarify the task for the the answers.
respondent or is interested in obtaining more per- Interviews also allow for some safeguards to
sonal information from the respondent. Telephone be built into the interview situation. Interviewers
interviews allow the researcher to reach more can clarify misunderstood questions and observe
respondents than do face-to-face interviews and the level of the respondent’s understanding and
provide more clarity than do questionnaires. cooperativeness. In addition, the researcher has
strict control over the order of the questions. With
Helpful Hint questionnaires, the respondent can answer ques-
Remember that sometimes researchers make tions in any order. Changing the order of the
trade-offs when determining the measures to be used. questions can sometimes change the response.
For example, if a researcher wants to learn about an
individual’s attitudes regarding practice, and practicali- Finally, interviews allow for richer and more
ties preclude using an interview, a questionnaire may be complex data to be collected, particularly when
used instead. open-ended responses are sought. Richard and
associates (2010) used open-ended questions in
Seneviratne and associates (2009 see Appen- their individual interviews to investigate concep-
dix A) used both unconcealed observations and tualizations of disease prevention and health pro-
interviews to understand more about work prac- motion among nurses from local public health
tices. The participants were asked grand tour organizations in Montreal. An example of such an
questions, such as Can you walk me through approach is To begin the interview, I’ll ask you
your typical day and were asked for examples to describe a disease prevention or health
 CHAPTER 13 Data-Collection Methods 299

promotion activity in which you have recently mailed. Armstrong-Stassen and Cameron (2005)
been involved within the context of your work mailed questionnaires to 3000 community health
(p. 452). nurses randomly selected from the College of
Even when closed-ended response items are Nurses of Ontario to investigate the work-related
used, interviewers can probe to understand why concerns, level of job satisfaction, and factors
a respondent answered in a particular way. Inter- in uencing the retention of community health
views can also be conducted in a group setting, nurses. The questionnaire packets contained a
called a focus group intervie , which may include cover letter from the researchers, an informed
about 6 to 8 participants. Wagner and colleagues consent form, a questionnaire booklet, and a reply
(2010) used a semistructured interview schedule envelope.
to guide focus group discussions among nursing
staff working in long-term care settings to explore Records or Available Data
patterns in communications about falls. These All of the data-collection methods discussed thus
small-group interviews allow the participants to far concern the ways that nurse researchers gather
freely explain and share information individually new data to study phenomena of interest. Not all
and collectively. Agreement and disagreement studies, however, require a researcher to acquire
among participants may be elicited, which allows new information. Existing information can some-
the researchers to obtain specific information times be examined in a new way to study a
from a number of participants efficiently and problem. The use of records and available data is
simultaneously. sometimes considered to be primarily the concern
uestionnaires are much less expensive to of historical research, but hospital records, care
administer than interviews because interviews plans, and existing data sources (e.g., the census)
may require the hiring and training of interview- are frequently used for collecting information.
ers. Thus, if a researcher has a fixed amount of What sets these studies apart from a literature
time and money, a larger and more diverse sample review is that these available data are examined
can be obtained with questionnaires. uestion- in a new way and not merely summarized they
naires also provide complete anonymity, which also answer specific research questions.
may be important if the study deals with sensitive ecords or available data then, are forms of
issues. Finally, the fact that no interviewer is information that are collected from existing mate-
present assures the researcher and the reader that rials, such as hospital records, historical docu-
no interviewer bias will occur. Interviewer bias ments, or audio or video recordings, and are used
occurs when the interviewer unwittingly leads the to answer research questions in a new manner.
respondent to answer in a certain way. This For example, the data analyzed in a study by
problem is especially pronounced in studies with Mammel and Kmet (2010) on the perioperative
unstructured interview formats. A subtle nod of administration of antibiotics in orthopaedic
the head, for example, could lead a respondent to trauma consisted of medical records to conduct
change an answer to correspond with what the the quality assurance review. The use of available
researcher wants to hear. data has certain advantages. Because the data-
uestionnaires were employed by Starzomski collection step of the research process is often
and Hilton (2000) to collect data on patient and the most difficult and time consuming, the use
family adjustment to kidney transplantation with of available records often produces a significant
and without an interim period of dialysis. To saving of time. If the records have been kept in a
reach a large sample, save time, and reduce the similar manner over time, analysis of these
labour costs of interviewing, questionnaires are records allows examination of trends over time.
300 PART FOUR Processes Related to Research

In addition, the use of available data decreases the sample records. Nonetheless, records and
problems of reactivity and response set bias. The available data constitute a rich source of data for
researcher also does not have to ask individuals study.
to participate in the study.
However, institutions are sometimes reluctant ONLINE AND COMPUTERIZED METHODS
to allow researchers access to their records. If the OF DATA COLLECTION
records are kept so that an individual cannot be With the fast-paced progression of the Internet
identified, access for research purposes is usually and computer technology, many researchers are
not a problem. Also, the Privacy Act, a federal using online data collection. The information
law, protects the rights of individuals who may obtained can be quantitative or qualitative, closed-
be identified in records, which would be a viola- ended or open-ended. This method of data collec-
tion of anonymity. tion can take the form of Web-based surveys or
One problem that affects the quality of avail- data input directly into microcomputers.
able data concerns survival of records. If the Many online survey tools, such as Survey-
records available are not representative of all of Monkey or uestionPro, are available a survey
the possible records, the researcher may have a can be downloaded quickly and the results
problem with bias. Often, because researchers obtained for a small fee. The advantages of this
have no way to tell whether the records have been method are that it is anonymous and inexpensive
saved in a biased manner, they need to make an respondents can fill out the survey in their own
intelligent guess as to their accuracy. For example, time a large number of participants can be
a researcher might be interested in studying accessed respondent time is reduced data-
socioeconomic factors associated with the suicide collection time is reduced duplicate responses
rate. These data frequently are underreported can be identified and, for the researcher, imple-
because of the stigma attached to suicide, and so mentation is time efficient. The disadvantages
the records would be biased. Recent interest in are that not everyone has access to a computer or
computerization of health records has led to an is computer literate, the response rates may be
increase in the discussion about the desirability low, and a large amount of data may be missing.
of access to such records for research. At this Computerized data collection can be accom-
time, how much of such data will continue to be plished through the use of personal digital assis-
readily available for research without consent is tants (PDAs). Researchers can input their data
unclear. directly into these handheld microcomputers. The
Another problem is related to the authenticity data can then be transferred to a larger computer
of the records. The distinction of primary and for analysis.
secondary sources is as relevant in this discussion
as it was in the discussion of the literature review Evidence-Informed Practice Tip
to determine the source of the work (see Chapter A critical evaluation of any data collection method
5). A book, for example, may have been ghost- includes evaluating the appropriateness, objectivity,
written, but all credit was accorded to the known consistency, and credibility of the method employed.
author. The researcher may have a difficult time
ferreting out these subtle types of biases.
Lastly, existing records may be missing a sig- CONSTRUCTION OF NEW INSTRUMENTS
nificant amount of data. For example, years of As already mentioned in this chapter, researchers
education may be recorded on only a portion of sometimes cannot locate an existing instrument
 CHAPTER 13 Data-Collection Methods 301

or method with acceptable reliability and validity Finally, the researcher administers or pilot
to measure the variable of interest. This situation tests the new instrument by applying it to a group
is often the case when part of a nursing theory is of people who are similar to those who will be
tested or when the effect of a clinical intervention studied in the larger investigation. The purpose of
is evaluated. For example, Gillis (1997) devel- this analysis is to determine the quality of the
oped and tested an instrument to evaluate adoles- instrument as a whole (reliability and validity)
cent lifestyle. The instrument was used to assess and the ability of each item to discriminate indi-
lifestyle patterns for identification of health vidual respondents (variance in item response).
education and counselling priorities (see Chapter The researcher also may administer a related
14). instrument to see whether the new instrument is
Instrument development is complex and time sufficiently different from the older one.
consuming, however. It consists of the following It is important that researchers who invest sig-
steps: nificant time in tool development publish their
Defining the construct to be measured results. For example, Goulet and associates
Formulating the items (questions) (2003) were interested in understanding the atti-
Assessing the items for content validity tudes of uebec’s adolescents toward breast-
Developing instructions for respondents feeding and how other people in uenced their
and users opinions. From their literature review, Goulet and
Pretesting and pilot testing the items associates determined that no suitable instrument
Estimating reliability and validity was available to measure this concept. They
Defining the construct (concepts at a higher devised their own instrument on the basis of the
level of abstraction) to be measured requires that theory of reasoned action (Fishbein Ajzen,
the researcher develop an expertise in the con- 1975). To ensure reliability and validity, they took
struct, which necessitates an extensive review of the following steps: (1) A panel composed of lac-
the literature and of all tests and measurements tation consultants and nurse researchers reviewed
that deal with related constructs. The researcher the items for content validity (2) pilot testing was
uses all this information to synthesize the performed before the full study to ensure ease of
available knowledge so that the construct can be administration, clarity, and precision of the instru-
defined. ment (3) a factor analysis was performed to
Once the construct is defined, the individual determine clusters of variables linked to form a
items for measuring the construct can be devel- construct and (4) reliability analysis produced
oped. The researcher will develop many more Cronbach’s alpha between .71 and .70 for differ-
items than are needed to address each aspect of ent parts of the instrument. This type of research
the construct or subconstruct. A panel of experts serves not only to introduce other researchers to
in the field evaluates the items so that the the tool but also to ultimately enhance the field,
researcher is assured that the items measure what inasmuch as the ability to conduct meaningful
they are intended to measure (content validity research is limited only by the ability to measure
see Chapter 14). Eventually, the number of items important phenomena.
is decreased because some items will not elicit
the intended information and will be dropped. In Helpful Hint
this phase, the researcher needs to ensure consis- Determine whether a newly developed survey or
tency both among the items and in testing and questionnaire was pilot tested to obtain preliminary evi-
dence of reliability and validity.
scoring procedures.
302 PART FOUR Processes Related to Research

APPRAISING THE EVIDENCE

Data-Collection Methods
Evaluating the adequacy of data collection methods from more inferences the observers are required to make, the
written research reports is often problematic for new more likely it is that observations will be biased. Also,
nursing research readers. Because the tool itself is not consider the problem of reactivity: In any observational
available for inspection, you may not feel comfortable situation, it is possible that the mere presence of the
judging the adequacy of the method without seeing it. observer will cause the participant to change the behav-
However, you can ask questions to judge the method iour in question. Of importance is not that reactivity could
chosen by the researcher. These questions are listed in occur but the extent to which reactivity could affect the
the Critiquing Criteria box. data. Finally, consider whether the observational proce-
In all studies, data collection methods should be dure was ethical. You need to consider whether the par-
clearly identified. The conceptual and operational defini- ticipants were informed that they were being observed,
tions of each important variable should be present in the whether any intervention was performed, and whether
report. Sometimes it is useful for the researcher to explain the participants had agreed to be observed.
why a particular method was chosen. For example, if the Interviews and questionnaires should be clearly
study dealt with young children, the researcher may described to allow the reader to decide whether the vari-
explain that a questionnaire was deemed to be an unrea- ables were adequately operationalized. Sometimes the
sonable task, and so an interview was chosen. researcher will reference the original report about the
Once you have identified the method chosen to tool, and you may wish to read this study before deciding
measure each variable of interest, you should decide whether the method was appropriate for the current
whether the method used was the best way to measure study. Also, the respondents’ task should be clear. Thus,
the variable. For example, if a questionnaire was used, the researcher should have made provisions for the par-
you might wonder why the researcher decided not to use ticipants to understand both their overall responsibilities
an interview. Also consider whether the method was and the individual items of the interview or question-
appropriate to the clinical situation. Does it make sense naire. The following questions must be considered: Who
to interview patients in the recovery room, for example? were the interviewers in the interview situation? Does the
Once you have decided whether all relevant variables researcher explain how they were trained to decrease any
are operationalized appropriately, you can begin to interviewer bias?
determine how well the method was carried out. For Available data, such as medical records, are subject to
studies involving physiological measurement, determine internal and external criticism. Internal criticism con-
whether the instrument was appropriate to the problem cerns the evaluation of the worth of the records and refers
and not forced to fit it. The rationale for selecting a par- primarily to the accuracy of the data. The researcher
ticular instrument should be given. For example, it may should present evidence that the records are genuine.
be important to know that the study was conducted External criticism is concerned with the authenticity of
under the auspices of a manufacturing firm that provided the records. Are the records really written by the first
the measuring instrument. In addition, the researcher author? The researcher may have a biased sample of all
should have made provisions to evaluate the accuracy of the possible records in the problem area, which may
of the instrument and the skill level of the people who have a profound effect on the validity of the results.
used it. Once you have decided that the data collection
Several considerations are important when you read method used was appropriate for the problem and the
studies that involve observational methods. Who were procedures were appropriate for the population studied,
the observers, and how were they trained? Is there any the reliability and validity of the instruments themselves
reason to believe that different observers perceived need to be considered. These characteristics are discussed
events or behaviours differently? Remember that the in Chapter 14.
 CHAPTER 13 Data-Collection Methods 303

CRITIQUING CRITERIA

1. Is the framework for research and the skill of the people who 4. Is any interviewer bias
clearly identified? used it? evident?

DATA-COLLECTION METHODS OBSERVATIONAL METHODS QUESTIONNAIRES

1. Are all of the data-collection 1. Who conducted the 1. Is the questionnaire described
instruments clearly identified observation? well enough for you to know
and described? 2. Were the observers trained to whether it covers the purpose of
2. Is the rationale for their selection minimize any bias? the study? Is evidence provided
given? 3. Was an observational guide that participants were able to
3. Is the method used appropriate provided? perform the task?
for the problem being studied? 4. Were the observers required to 3. Is it clear that the participants
4. Were the methods used make inferences about what understood the questionnaire?
appropriate for the clinical they saw? 4. Are the majority of the items
situation? 5. Is there any reason to believe appropriately closed- or
5. Are the data-collection that the presence of the open-ended?
procedures similar for all observers affected the behaviour
participants? of the participants? AVAILABLE DATA AND RECORDS
6. Were efforts made to ensure 6. Were the observations 1. Are the records used
intervention fidelity through the performed according to the appropriate for the problem
data-collection protocol? principles of informed consent? being studied?
2. Are the data examined in such a
PHYSIOLOGICAL MEASUREMENT INTERVIEWS way as to provide new
1. Is the instrument used 1. Is the interview schedule information and not summarize
appropriate for the research described adequately enough for the records?
problem and not forced to fit you to know whether it covers 3. Has the author addressed
it? the purpose of the study? questions of internal and
2. Is a rationale given for why a 2. Is it clear that the participants external criticism?
particular instrument was understood the task and the 4. Is there any indication of
selected? questions? selection bias in the available
3. Is there a provision for evaluating 3. Who were the interviewers, and records?
the accuracy of the instrument how were they trained?

offer to support the establishment of interrater

CRITICAL THINKING CHALLENGES reliability?
■ Observation is a data-collection method used
■ Physiological measurements are objective, frequently in nursing research. Discuss the factors
precise, and sensitive. Discuss factors that that make nurses perfect potential candidates for
might influence their validity and this role and the disadvantages of using this
feasibility. method.
■ A student in research class asks why nurses ■ Studies often use a survey to collect data. How
who participate in a clinical research study in the can researchers increase their return rate for the
role of a data collector or who perform a survey, and how do they determine whether the
“treatment intervention” need to be trained.
survey return is adequate?
What important factors or rationale would you
304 PART FOUR Processes Related to Research

can be completely anonymous and prevent

KEY POINTS interviewer bias.
• Interviews are most appropriate when a large
• Data-collection methods are described as being response rate and an unbiased sample are
both objective and systematic. The data-collection important because the refusal rate for interviews is
methods of a study provide the operational much lower than that for questionnaires.
definitions of the relevant variables. Interviews enable the participation of people who
• Types of data-collection methods include cannot use a questionnaire, such as children and
physiological measurements, observational people who are illiterate. An interviewer can clarify
methods, interviews, questionnaires, and records and maintain the order of the questions for all
or available data. Each method has advantages participants.
and disadvantages. • Records or available data are also an important
• Physiological measurements are the methods in source of research data. The use of available data
which technical instruments are used to collect may save the researcher considerable time and
data about patients’ physical, chemical, money in conducting a study. This method reduces
microbiological, or anatomical status. These problems with both reactivity and ethical concerns.
methods are suited to studying how to improve However, records and available data are subject to
the effectiveness of nursing care. Physiological problems of availability, authenticity, and accuracy.
measurements are objective, precise, and sensitive, • A critical evaluation of data-collection methods
but they may be very expensive and may distort should emphasize the appropriateness, objectivity,
the variable of interest. and consistency of the method employed.
• Observational methods are used in nursing
research when the variables of interest deal with
events or behaviours. Scientific observation
requires preplanning, systematic recording, REFERENCES
controlling the observations, and determining the
relationship to scientific theory. This method is best Armstrong-Stassen, M., Cameron, S. (2005).
suited to research problems that are difficult to Concerns, satisfaction, and retention of Canadian
view as part of a whole. Observers may be community health nurses. ournal of Community
required to perform or not perform interventions, ealth Nursing, , 181-194.
and their activity may be concealed or obvious. Bryanton, J., Walsh, D., Barrett, M., Gaudet, D.
• Observational methods have several advantages:
(2004). Tub bathing versus traditional sponge bathing
(1) they provide flexibility to measure many types
of situations, and (2) they enable a great depth for the newborn. ournal of bstetric, ynecologic,
and breadth of information to be collected. and Neonatal Nursing, , 704-712.
• Observation has disadvantages as well: (1) data Doran, D. M., Harrison, M. B., Laschinger, H. S.,
may be distorted as a result of the observer’s Hirdes, J. P., Rukholm, E., Sidani, S., . . .
presence (reactivity), (2) concealment requires the Tourangeau, A. E. (2006). Nursing-sensitive out-
consideration of ethical issues, and (3) data from comes data collection in acute care and long-term-
observations may be biased by the person who is care settings. Nursing Research, (2S), S75-S81.
doing the observing. Fishbein, M., Ajzen, I. (1975). Belief, attitude,
• Interviews are data-collection methods commonly intention, and behavior: An introduction to theory
used in nursing research. Items on interview
and research. Reading, MA: Addison-Wesley
schedules may be of direct or indirect interest.
Participants may be asked either open-ended or Gillis, A. J. (1997). The Adolescent Lifestyle uestion-
closed-ended questions. The form of the question naire: Development and psychometric testing.
should be clear to the respondent, free of Canadian ournal of Nursing Research, , 29-46.
suggestion, and grammatically correct. Goulet, C., Lampron, A., Marcil, I., Ross, L. (2003).
• Questionnaires, or surveys, are useful when the Attitudes and subjective norms of male and female
number of questions to be asked is finite. The adolescents toward breastfeeding. ournal of uman
questions need to be clear and specific. actation, , 402-410.
Questionnaires are less costly and less time Horgas, A. L., Yoon, S. L., Nichols, A. L., Marsiske,
consuming to administer to large groups of M. (2008). The relationship between pain and
participants, particularly if the participants are
functional status in black and white older adults.
geographically widespread. Questionnaires also
Research in Nursing and ealth, (94), 341-354.
 CHAPTER 13 Data-Collection Methods 305

Koniak-Griffin, D., Verzemnieks, I. L., Anderson, Smith, D., Edwards, N., Varcoe, C., Martens, P. J.,
N. L. R., Brecht, M. L., Lesser, J., Kim, S., Davies, B. (2006). Bringing safety and responsive-
Turner-Pluta, C. (2003). Nurse visitation for ness into the forefront of care for pregnant and
adolescent mothers: Two-year infant health and parenting Aboriginal people. Advances in Nursing
maternal outcomes. Nursing Research, , 127-136. Science, (2), E27-E44.
Mammel, J., Kmet, L. (2010). Perioperative antibiotic Starzomski, R., Hilton, A. (2000). Patient and family
administration in orthopaedic trauma. rthopaedic adjustment to kidney transplantation with and without
Nursing, (2), 77-83. an interim period of dialysis. Nephrology Nursing
Molzahn, A. E. (January 2007). Spirituality in later life: ournal, (1), 17-32.
Effect on quality of life. ournal of erontological Sword, W., Watt, S., Kreuger, P. (2006). Postpartum
Nursing, (1), 32-39. health, service needs, and access to care experiences
Neyerhof, A. J. (2006). Methods of coping with social of immigrant and Canadian-born women. ournal of
desirability bias: A review. European ournal of bstetric, ynecologic, and Neonatal Nursing, (6),
Psychology, , 263-280. 717-727.
Ratner, P. A., Johnson, J. L., Richardson, C. G., Van Cleve, L., Bossert, E., Beecroft, P., Adlard, K.,
Bottorff, J. L., Moffatt, B., Mackay, M., . . . Budz, B. Alvarez, O., Savedra, M. C. (2004). The pain
(2004). Efficacy of a smoking-cessation intervention experience of children with leukemia during the first
for elective-surgical patients. Research in Nursing year after diagnosis. Nursing Research, , 1-10.
ealth, , 148-161. Wagner, L. M., Damianakis, T., Mafrici, N.,
Richard, L., Gendron, S., Beaudet, N., Boisvert, N., Robinson-Holt, K. (2010). Falls communication
Sauv , M. S., Garceau-Brodeur, M. (2010). Health patterns among nursing staff working in long-term
promotion and disease prevention among nurses care settings. Clinical Nursing Research, (3),
working in local public health organizations in 311-326.
Montr al, u bec. Public ealth Nursing, (5),
450-458.
Santacroce, S. J., Maccarelli, L. M., Grey, M. (2004).
Intervention fidelity. Nursing Research, , 63-66.
FOR FURTHER STUDY
Seneviratne, C. C., Mather, C. M., Then, K. L. Go to Evolve at http://evolve.elsevier.com/
(2009). nderstanding nursing on an acute stroke Canada/LoBiondo/Research for Audio Glossary, how-to
unit: Perceptions of space, time and interprofessional instructions for Writing Proposals for Funding, and
practice. ournal of Advanced Nursing, (9), additional research articles for practice in reviewing
1872-1881. doi: 10.1111/j.1365-2648.2009.05053.x and critiquing.
C H A PTER 1 4

Rigour in Research
Geri LoBiondo-Wood | Judith Haber | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Discuss the purposes of reliability and validity.
• Define reliability.
• Discuss the concepts of stability, equivalence, and homogeneity as they relate to reliability.
• Compare the estimates of reliability.
• Define validity.
• Compare content validity, criterion-related validity, and construct validity.
• Discuss how measurement error can affect the outcomes of a research study.
• Discuss the purpose of credibility, auditability, and fittingness.
• Identify the criteria for critiquing the reliability and validity of measurement tools.
• Use the critiquing criteria to evaluate the reliability and validity of measurement tools.
• Discuss how evidence related to research rigour contributes to clinical decision making.

KEY TERMS
alpha coefficient equivalence observed test score
alternate-form reliability error variance parallel-form reliability
auditability face validity predictive validity
chance error factor analysis random error
Cohen’s kappa fittingness reliability
concurrent validity homogeneity reliability coefficient
constant error hypothesis-testing approach rigour
construct validity internal consistency split-half reliability
content validity interrater reliability stability
contrasted-groups approach item-to-total correlation systematic error
convergent validity known-groups approach test-retest reliability
credibility Kuder-Richardson (KR-20) validation sample
criterion-related validity coefficient validity
Cronbach’s alpha multitrait-multimethod
divergent validity approach

306
 CHAPTER 14 Rigour in Research 307

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

IN BOTH QUANTITATIVE AND QUALITATIVE RESEARCH, Canadian ournal of Nursing Research, Interna-
the purpose is to collect trustworthy data that can tional ournal of ualitative Methods, and other
be used for analyses to make generalizations nursing research journals. In this chapter, con-
about the population and that are transferable to cepts related to quantitative rigour are discussed
other groups. Because findings need to be gener- first, followed by factors that contribute to the
alizable and transferable, measurement of nurs- trustworthiness of qualitative research.
ing phenomena is a major concern of nursing When you read quantitative research studies
researchers, and rigour is strived for. igour and reports, you must assess the reliability and
refers to the strictness with which a study is con- validity of the instruments used in each study to
ducted to enhance the quality, believability, or determine the soundness of the selection of these
trustworthiness of the study findings. Rigour in instruments in relation to the concepts or vari-
quantitative research is determined by measure- ables under investigation. The appropriateness of
ment instruments that validly and reliably re ect the instruments and the extent to which reliability
the concepts of the theory being tested, so that and validity are demonstrated have a profound
conclusions drawn from a study will be valid and in uence on the findings and on the internal and
will advance the development of nursing theory external validity of the study. Invalid measures
and evidence-informed practice. Thus, psycho- produce invalid estimates of the relationships
metric assessments are designed to obtain evi- between variables, thus affecting internal validity.
dence of the quality of these instruments: that is, The use of invalid measures also leads to inac-
their reliability and validity. curate generalizations to the populations being
Issues of reliability and validity are of central studied, thus affecting external validity and the
concern to the researcher, as well as to you as the ability to apply or not apply research findings in
critiquer of research. From either perspective, the clinical practice. Thus, the assessment of reliabil-
measurement instruments that are used in a ity and validity is an extremely important skill to
research study must be evaluated. Many new con- develop for critiquing nursing research.
structs are relevant to nursing theory, and a Regardless of whether a new or already devel-
growing number of established measurement oped measurement tool is used in a research
instruments are available to researchers. However, study, evidence of reliability and validity is
researchers often face the challenge of develop- crucial. Box 14-1 identifies several Internet re-
ing new instruments and, as part of that process, sources that you can use to access and evaluate
establishing the reliability and validity of those the reliability and validity of the measurement
tools. instruments used in research studies.
In qualitative research, rigour is ascertained
by credibility, auditability, and fittingness. The RELIABILITY
growing importance of measurement issues, tool People are considered reliable when their behav-
development, and related issues (e.g., reliability iour is consistent and predictable. Likewise, the
and validity, qualitative rigour) is evident in reliability of a research instrument is the extent
issues of the ournal of Nursing Measurement, to which the instrument yields the same results
308 PART FOUR Processes Related to Research

BOX 14-1 tion of accuracy to inaccuracy in measurement.

INTERNET RESOURCES FOR ACCESSING AND In other words, if researchers use the same or
EVALUATING THE VALIDITY AND RELIABILITY OF comparable instruments on more than one occa-
MEASUREMENT INSTRUMENTS* sion to measure behaviours that ordinarily remain
• Mental Measurements Yearbook—Test Reviews relatively constant, the researchers would expect
Online similar results if the tools are reliable.
To subscribe, visit http://buros.unl.edu/buros/jsp/ The three main attributes of a reliable scale are
search.jsp.
Contains online abstracts of 4000 commercially stability, homogeneity, and equivalence. The sta
available tests, of which more than 2500 have been bility of an instrument refers to the instrument’s
critically appraised by the Buros Institute. These ability to produce the same results with repeated
reviews also appear in the print version of the Mental
Measurements Yearbook series. Searches can be testing. The ho ogeneity or internal consis
conducted by author or by using the 18 search tency of an instrument means that all of the items
categories. in a tool measure the same concept or character-
• WORLDviews on Evidence-Based Nursing istic. An instrument is said to exhibit e uivalence
To subscribe, visit http://www.nursingsociety.org/ if the tool produces the same results when equiva-
Publications/Journals/Pages/worldviews.aspx.
Provides full-text articles and searches, hypertext lent or parallel instruments or procedures are
navigation, links to Cumulative Index to Nursing and used. Each of these attributes and the means to
Allied Health Literature (CINAHL) and MEDLINE, and estimate them are discussed here. Before these
tables and figures.
are discussed, however, an understanding of how
• Virginia Henderson International Nursing Library
to interpret reliability is essential.
To subscribe, visit http://www.nursinglibrary.org/
vhl/?.
Includes the following databases: Registry of Nurse Interpretation of the
Researchers, Registry of Research Projects, and Registry Reliability Coefficient
of Research Results.
Because all of the attributes of reliability are con-
• Sigma Theta Tau International Nursing Honor
Society cerned with the degree of consistency between
550 West North Street Indianapolis, IN 46202; scores that is obtained at two or more independent
317-634-8171; http://www.nursingsociety.org. times of testing, these attributes often are
Supports the learning, knowledge, and professional
expressed in terms of a correlation coefficient.
development of nurses. Also promotes resources to
bridge research and clinical practice. The reliability coe cient or al ha coe cient
• ProQolid: Patient-Reported Outcome and Quality expresses the relationship between the error vari-
of Life Instruments Database ance, true variance, and the observed score, and
http://www.proqolid.org/ it ranges from 0 to 1. A correlation of 0 indicates
Identifies and describes quality-of-life instruments.
no relationship, and thus the error variance is
*See Chapter 5 for detailed information about Internet resources. high. When the error variance in a measurement
instrument is low, the reliability coefficient is
on repeated measures. Reliability, then, is con- closer to 1. The closer to 1 the coefficient is, the
cerned with consistency, accuracy, precision, sta- more reliable the tool is. For example, suppose
bility, equivalence, and homogeneity. Concurrent that a reliability coefficient of a tool is reported
with questions of validity, or after these questions to be .89. This number indicates that the error
are answered, the researcher and you, as the cri- variance is small and the tool has little measure-
tiquer, ask how reliable the instrument is. ment error. But if the reliability coefficient of a
A reliable measure can produce the same measure is reported to be .49, the error variance
results if the behaviour is measured again by the is high, and the tool has a problem with measure-
same scale. Reliability, then, refers to the propor- ment error. For a tool to be considered reliable, a
 CHAPTER 14 Rigour in Research 309

level of .70 or higher should be reported, although instrument is used on several occasions. Stability
the intended purpose of the instrument needs to is also a consideration when a researcher is con-
be considered if lower levels are accepted. ducting an intervention study that is designed to
The interpretation of the reliability coefficient effect a change in a specific variable. In this case,
depends on the proposed purpose of the measure. the instrument is administered once and then
Seven major tests of reliability can be used again after the alteration or change intervention
to calculate a reliability coefficient, depending has been completed. The tests that are used to
on the nature of the tool: test-retest reliability, estimate stability are test-retest reliability and
parallel- or alternate-form reliability, item-to- parallel- or alternate-form reliability.
total correlation, split-half reliability, uder-
Richardson coef cient, Cronbach s alpha, and Test-Retest Reliability
interrater reliability These tests are discussed as Test retest reliability is the stability of the scores
they relate to the attributes of stability, homoge- of an instrument when it is administered more
neity, and equivalence (Box 14-2). In critiquing than once to the same participants under similar
research reports, you should be aware that no conditions. Scores from repeated testing are com-
single best way exists to assess reliability in rela- pared. This comparison is expressed by a correla-
tion to these attributes and that the researcher’s tion coefficient, usually a Pearson r (see Chapter
method should be consistent with the aim of the 16). The interval between repeated administra-
research. tions varies and depends on the concept or vari-
able being measured. For example, if the variable
Stability that the test measures is related to developmental
An instrument is thought to be stable or to exhibit stages in children, the interval between test
stability when repeated administration of the administrations should be short. The amount of
instrument yields the same results. Researchers time over which the variable was measured
are concerned with an instrument’s stability should also be recorded in the report.
because they expect the instrument to measure a An example of an instrument that was assessed
concept consistently over a period of time. Mea- for test-retest reliability is Akhtar-Danesh and
surement over time is important in a longitudinal associates’ (2010) learning needs assessment
study because in that type of research, an questionnaire for community health nurses. Test-
retest reliability was assessed at approximately a
2-week interval, and a high test-retest reliability
BOX 14-2 coefficient (r = .89, p < .01) was obtained. The
MEASURES USED TO TEST RELIABILITY interval was adequate (2 weeks between testing),
STABILITY and coefficients exceeded .80 and were thus very
Test-retest reliability good (Nunnally Bernstein, 1994).
Parallel- or alternate-form reliability

HOMOGENEITY
Parallel- or Alternate-Form Reliability
Item-to-total correlation Parallel-form reliability is applicable and can be
Split-half reliability tested only if two comparable forms of the same
Kuder-Richardson (KR-20) coefficient instrument exist. Parallel or reliability or
Cronbach’s alpha
alternate or reliability is like test-retest reli-
EQUIVALENCE ability in that the same individuals are tested
Parallel- or alternate-form reliability more than once within a specific interval, but in
Interrater reliability
the assessment of parallel-form reliability, a
310 PART FOUR Processes Related to Research

different form of the same test is given to the within the scale re ect or measure the same
participants on the second testing. Parallel forms concept. In other words, the items within the scale
or tests contain the same types of items that are are correlated with, or complementary to, each
based on the same domain or concept, but the other, and the scale is unidimensional A unidi-
wording of the items is different. The develop- mensional scale measures one concept, such as
ment of parallel forms is desired if the instrument exercise self-efficacy. A total score is then used
is intended to measure a variable for which a in the analysis of data.
researcher believes that testwiseness will be a When Akhtar-Danesh and associates’ (2010)
problem, that is, respondents might recognize the learning needs assessment questionnaire was
test items and try to answer them in the same way tested for homogeneity, the reliability (alpha)
as previously, instead of spontaneously. coefficient was .99. In exceeding .70, the reli-
For example, in their randomized controlled ability coefficient provided sufficient evidence of
trial, Budin and colleagues (2008) compared the the internal consistency of the instrument. Homo-
differential effect of a phase-specific standardized geneity can be assessed with one of four methods:
educational video intervention with that of a tele- item-to-total correlations, split-half reliability,
phone counselling intervention on physical, emo- Kuder-Richardson coefficient, or Cronbach’s
tional, and social adjustment in women with alpha.
breast cancer and their partners. Because repeated
measures over the four data-collection points—
coping with the diagnosis, recovering from Helpful Hint
When the characteristics of a study sample differ
surgery, understanding adjuvant therapy, and significantly from those of the sample in the original
ongoing recovery—were used, it was appropriate study, check to see whether the researcher has reestab-
to use two alternative forms of the Partner Rela- lished the reliability of the instrument with the current
tionship Inventory (Hoskins, 1988) to measure sample.
emotional adjustment in partners. Each item on
one scale (e.g., I am able to tell my partner how
I feel ) is paired with one item on the second Item-to-Total Correlation
form (e.g., My partner tries to understand my The ite to total correlation is a measure of the
feelings ), and the responses should therefore be relationship between each scale item and the total
consistent. scale. When item-to-total correlations are calcu-
Practically speaking, developing alternative lated, a correlation for each item on the scale is
forms of an instrument is difficult because of the generated (Table 14-1). Items that do not achieve
many issues of reliability and validity. If alterna- a high correlation may be deleted from the
tive forms of a test exist, they should be highly
correlated if they are to be considered reliable.
TABLE 14-1

Helpful Hint EXAMPLES OF ITEM-TO-TOTAL CORRELATIONS

FROM COMPUTER-GENERATED DATA
When a longitudinal design with multiple data-
collection points is being conducted, look for evidence ITEM ITEM-TO-TOTAL CORRELATION
of test-retest reliability or parallel-form reliability. 1 .5069
2 .4355
3 .4479
Homogeneity, or Internal Consistency 4 .4369
Another attribute related to reliability of an instru- 5 .4213
6 .4216
ment is the homogeneity with which the items
 CHAPTER 14 Rigour in Research 311

instrument. In a research study, the lowest and self-report measure was developed to assess the
highest item-to-total correlations are typically degree to which respondents believe that thoughts
reported the other correlations are usually not about breast cancer are interfering with daily
reported unless the study is a methodological functioning. The measure is embedded within a
investigation. An example of an item-to-total cor- larger questionnaire that is also used to assess
relation report is illustrated in the study by Sidani perceived risk, intention to undergo genetic
and associates (2009), who evaluated the psycho- testing, and frequency of worry about getting
metric properties of the treatment and preferences breast cancer. The Worry Interference Scale items
measure. In that study, the item-to-total correla- concern disruptions in sleep, work, concentration,
tions ranged between .47 and .82, According to relationships, having fun, feeling sexually attrac-
Nunnally and Bernstein (1994), these results are tive, meeting family needs, and reproductive
acceptable because the minimal mandatory cor- decisions. Ibrahim (2002) computed a Spearman-
relation should be greater than .30. Brown split-half reliability and found a reliability
Heaman and Gupton (2009) conducted a study coefficient that ranged from .83 to .92 for the first
to refine a new instrument, the Perception of four items and from .75 to .83 for the other items.
Pregnancy Risk uestionnaire (PPR ). A corre- Split-half reliabilities of at least .75 are consid-
lation matrix was constructed to summarize the ered internally consistent.
interrelationships among the 11 items in the scale.
The matrix was examined to identify any items Kuder-Richardson Coefficient
to be dropped if they were either too highly cor- The uder ichardson coe cient is
related (r ≥ .80) or not correlated sufficiently the estimate of homogeneity used for instruments
(r < .30) with one another. The results indicated that have a dichotomous response format. A
that all items were adequately correlated (r ≥ .30) dichotomous response format is one in which the
with at least seven other items in the matrix. answer to a question should be either yes or
There were concerns about redundancy of items, no or either true or false. The technique
inasmuch as no correlations exceeded .74. yields a correlation that is based on the consis-
tency of responses to all items of a single form of
Split-Half Reliability a test that is administered once.
S lit hal reliability involves dividing a scale For example, in an investigation of the effec-
into halves and making a comparison. The halves tiveness of a randomized support group interven-
may be, for example, odd-numbered and even- tion for women with breast cancer, breast cancer
numbered items or a simple division of the first knowledge was assessed with a 25-item true/false
from the second half, or items may be randomly questionnaire developed for the study. Items were
grouped into halves that will be analyzed oppo- obtained from the American Cancer Society’s
site one another. Split-half reliability provides a publication Cancer acts and igures and were
measure of consistency in terms of sampling the categorized as follows: knowledge of risk factors
content. The two halves of the test or the contents for developing breast cancer (10 items e.g.,
in both halves are assumed to be comparable, and Most women diagnosed with breast cancer have
a reliability coefficient is calculated. If the scores at least one known risk factor for the disease )
for the two halves are approximately equal, the symptoms of breast cancer (5 items e.g., Women
test may be considered reliable. who have breast cancer never experience any
The Spearman-Brown formula is one method symptoms of the disease ) side effects of treat-
of calculating the reliability coefficient. In a test ment (3 items e.g., A common side effect of
of the Worry Interference Scale, a seven-item radiation is sunburn-like symptoms ) treatment
312 PART FOUR Processes Related to Research

efficacy (4 items e.g., For women with small Table 14-2. Cronbach’s alpha exceeded .70 for
tumors that may not have spread outside the breast, each option, thereby providing sufficient evi-
having either a mastectomy or lumpectomy with dence of the internal consistency of the instru-
axillary lymph node dissection results in the same ment. Examples of reported Cronbach’s alpha are
overall life expectancy ) and methods of treat- shown in Box 14-3.
ment (3 items e.g., Hormone treatment is used
only for premenopausal women ). Because the Helpful Hint
scale was a binary format (true/false), the Kuder- If a research article provides information about
the reliability of a measurement instrument but does
Richardson reliability for the entire scale was cal- not specify the type of reliability, it is probably safe to
culated at .75, which is acceptable, having exceeded assume that internal consistency reliability was assessed
the minimum acceptable score of .70 however, the with Cronbach’s alpha.
magnitude of the correlation is not robust.

Cronbach’s Alpha TABLE 14-2

The fourth and most commonly used test of inter- CRONBACH’S ALPHASCORES FOR THE FOUR
OPTIONS OF THE TREATMENT AND
nal consistency is Cronbach’s alpha. Cronbach s
PREFERENCES MEASURE
al ha is a test of internal consistency in which OPTIONS CRONBACH’S ALPHA
each item in the scale is simultaneously compared
with the others, and a total score is then used to Sleep education and hygiene .86
Stimulus control instructions .80
analyze the data. Many tools used to measure psy- Sleep restriction therapy .84
chosocial variables and attitudes have a Likert- Multicomponent .87
type scale response format, which is very suitable Adapted from Sidani, S., Epstein, D. R., Bootzin, R. R., Moritz, P., &
for testing internal consistency. In a Likert-type Miranda, J. (2009). Assessment of preferences for treatment: Validation
of a measure. Research in Nursing and Health, 32, 419–438.
scale format, the participant responds to a question
on a scale of varying degrees of intensity between
two extremes. The two extremes are anchored by BOX 14-3
responses ranging from, for example, strongly EXAMPLES OF REPORTED CRONBACH’S ALPHA
agree to strongly disagree or from most like • “Interitem correlation for the learning needs
me to least like me. The points between the two subscales ranged from r = .53 to r = .72, which
extremes may range from 1 to 5 or 1 to 7. Partici- was in the acceptable range.” (Akhtar-Danesh,
Valaitis, Schofield, Underwood, Martin-Misener,
pants are asked to circle the response that most Baumann, Kolotylo, 2010, p. 1061)
closely represents what they believe. • “The total 9-item scale had a Cronbach’s alpha of
Figure 14-1 displays examples of items from .87, the 5-item scale had an alpha of .84, and the
4-item Risk For Self-subscale had an alpha of .81.”
a tool in which a Likert-type scale format was (Heaman & Gupton, 2009, p. 500)
used to develop a career search instrument • “For this study, the Cronbach α reliabilities for the
(Roberts Ward-Smith, 2010). The psychomet- subscales ranged from .72 to .85 and a total PES
[Practice Environment Scale] reliability was found
ric properties of Sidani and colleagues’ (2009) to be .92.” (Armstrong, Laschinger, & Wong, 2009,
Treatment and Preferences measure were tested p. 58)
for internal consistency and construct validity. • “Reliability, determined using Cronbach’s alpha,
The testing revealed that there were four separate was 0.87 for the CSQ [Career Search
Questionnaire]. Subscale reliability for the 23 items
treatment options: sleep education and hygiene, purported to measure career interest was 0.77.
stimulus control instructions, sleep restriction Reliability for the 25 items which aimed to
therapy, and, lastly, a multicomponent option determine self-efficacy was 0.82.” (Roberts &
Ward-Smith, 2010, p. 8)
involving the other three options, as illustrated in
 CHAPTER 14 Rigour in Research 313

Career Search Questionnaire

No Little Some Much

interest interest Neutral interest interest

2. I am interested in learning new words and

terminology to perform a job.

3. I am interested in fine arts activities such as

design or performance.

5. I enjoy working with other people to find the

best possible solutions to their problems.

7. I enjoy thinking about things I have done and

figuring out ways to do them better.

8. I am interested in people who express opinions

very different from my own.

9. I would like to advance my career by being a

lifelong learner, even if it requires additional
classes.

22. I am interested in a job where I can empower

other people.

23. I am interested in a job where I make my own

decisions.

No
confidence Very little Moderate Much Complete
at all confidence confidence confidence confidence

2. I believe I could write speeches for people.

6. I am able to be compassionate with ill people.

7. I am able to use algebra to solve technical

problems.

8. I am capable of communicating well even in

stressful situations.

10. I believe that I can demonstrate concern for

other people’s privacy and dignity.

11. I am able to keep a healthy boundary between

my work and my own life.

12. I am capable of keeping my workplace safe for

myself and others.

22. I believe I can get people to do things my way.

FIGURE 14-1 Examples of portions of Likert-type scales.

Adapted from Roberts, C. A., & Ward-Smith, P. (2010). Choosing a career in nursing: Develop-
ment of a career search instrument. International Journal of Nursing Education Scholarship, 7(1),
Article 2.
314 PART FOUR Processes Related to Research

Resnick and colleagues noted that other tools

Equivalence available to assess activity or exercise in older
Equivalence is either the consistency or agree- adults are generally self-report instruments that
ment among observers who use the same mea- provide researchers or nurses with the individu-
surement tool or the consistency or agreement al’s perception of the amount of activity per-
between alternative forms of a tool. An instru- formed. In contrast, the SAM provides objective
ment is thought to demonstrate equivalence when information about step activity (the total number
two or more observers have a high percentage of of steps taken over a period of time).
agreement about a certain behaviour or when Participants in the study (N = 30) were recruited
alternative forms of a test yield a high correlation. from the outpatient office in a continuing care
Two methods to test equivalence are interr- retirement community. They were asked to walk
ater reliability and alternate- or parallel-form at their own speed on a carpeted level surface for
reliability. 1 minute, rest for 2 minutes, and then ambulate
over the same path for another minute. Two
Interrater Reliability observers, observing at the same time, counted
Some measurement instruments are not self- steps during both episodes of ambulation. The
administered questionnaires but instead are direct two observers, both advanced practice nurses,
measurements of observed behaviour that must had prior experience in observational gait assess-
be systematically recorded. Such instruments ment of older adults and step counting.
must be tested for interrater reliability (the con- Interrater reliability was based on intraclass
sistency of observations between two or more correlations of the two separate 1-minute obser-
observers with the same tool). To accomplish vations of the SAM performed 2 minutes apart.
interrater reliability, either two or more individu- Statistically significant correlations of .80 or
als should make an observation or one observer greater between the means of the two observers
should observe the same behaviour on several and the results of the SAM at each testing interval
occasions. The observers should score their (in which the average deviation between the
observations with regard to the definition and SAM and the mean of the two observers was
operationalization of the behaviour to be observed. calculated) produced an interrater reliability esti-
When the research method of direct observa- mate of .98, indicating a high level of agreement
tion of a behaviour is required, consistency (or between observers. According to Santacroce and
reliability) of the observations among all observ- associates (2004), consistency in observation or
ers is extremely important. Interrater reliability intervention delivery is evidence that the study
concerns the reliability (or consistency) of the findings are valid and reliable.
observer, not the reliability of the instrument. Another type of interrater reliability is Cohen’s
Interrater reliability is expressed either as a per- kappa, a coefficient of agreement between two
centage of agreement between scorers or as a raters that is considered to be a more precise
correlation coefficient of the scores assigned to estimate of interrater reliability. Cohen s a a
the observed behaviours. expresses the level of agreement that is observed
Resnick and colleagues (2001) established beyond the level that would be expected by
interrater reliability for the step activity monitor chance alone. A Cohen’s kappa of .08 or better is
(SAM), a step counter developed for a wide generally assumed to indicate good interrater reli-
variety of gait styles (ranging from a slow shuf e, ability. A Cohen’s kappa of .68 allows tentative
such as a Parkinsonian-type gait, to a fast run) conclusions to be drawn when lower levels of
and used to measure activity in older adults. reliability are acceptable (McDowell Newell,
 CHAPTER 14 Rigour in Research 315

1996). In her study on metacognitive factors that If an instrument is erratic, inconsistent, and inac-
affect student nurses’ use of point-of-care tech- curate, it cannot validly measure the attribute of
nology in clinical settings, Kuiper (2010) estab- interest.
lished interrater reliability (.70 to .90) to determine The three major kinds of validity—content,
consistency in order to re ect self-regulated criterion-related, and construct validity—vary
learning processes in journal prompts. according to the kind of information provided and
the investigator’s purpose. In critiquing research
Evidence-Informed Practice Tip articles, you will want to evaluate whether suffi-
Interrater reliability is important for minimizing
cient evidence of validity is present and whether
bias. the type of validity is appropriate to the design of
the study and instruments used in the study. The
sample that provides the initial data for determin-
Parallel- or Alternate-Form Reliability ing the reliability and validity of a measurement
Parallel- or alternate-form reliability was des- tool is termed a validation sa le
cribed in the discussion of stability (see pp. 309–
310). se of parallel forms is thus a measure of
stability and equivalence. The procedures for Evidence-Informed Practice Tip
assessing equivalence through the use of parallel Selecting measurement instruments that have
strong evidence of validity increases the reader’s confi-
forms are the same. dence in the study findings: that the researchers actually
measured what they intended to measure.
VALIDITY
alidity refers to whether a measurement instru-
ment accurately measures what it is intended to Content Validity
measure. To be valid, an instrument must first be Content validity is the degree to which the
reliable without reliability, the instrument cannot content of the measure represents the universe of
have validity. However, reliability, although nec- content: that is, the domain of a given construct.
essary, is not a sufficient condition for validity. The universe of content provides the framework
Internal and external validity of a study are dis- and basis for formulating the items that will ade-
cussed in Chapter 9. quately represent the content. When an investiga-
For example, a valid instrument that is intended tor is developing a tool and issues of content
to measure anxiety does so it does not measure validity arise, the concern is whether the mea-
another construct, such as stress. A reliable surement tool and the items it contains are repre-
measure can consistently rank participants on a sentative of the universe of content that the
given construct (e.g., anxiety), but a valid measure researcher intends to measure. The researcher
correctly measures the construct of interest. A begins by defining the concept and identifying
measure can be reliable but not valid. Suppose the dimensions that are the components of the
that a researcher wanted to measure anxiety in concept. The items that re ect the concept and its
patients by measuring their body temperatures. dimensions are formulated (see Practical Appli-
The researcher could obtain highly accurate, con- cation box for two examples).
sistent, and precise temperature recordings, but When the researcher has completed this task,
such a measure would not be a valid indicator of the items are submitted to a panel of judges con-
anxiety. Thus, the high reliability of an instrument sidered to be experts on this concept. Researchers
is not necessarily congruent with evidence of typically request that the judges indicate their
validity. A valid instrument, however, is reliable. level of agreement with the scope of the items and
316 PART FOUR Processes Related to Research

the extent to which the items re ect the concept health-protecting behaviours. To establish face
under consideration. validity, in addition to content validity, a panel of
eight nurses reviewed the items for clarity and
relevance.
Practical Application
Gélinas and colleagues (2008) reported on
the item selection process and evaluation of Evidence-Informed Practice Tip
the content validity of the Critical-Care Pain When face validity and content validity, the most
Observation Tool for nonverbal adults who were basic types of validity, are the only types of validity
critically ill. A version of this tool was developed to reported in a research article, you, as a research con-
include both behavioural and physiological sumer, cannot appraise the measurement tools as
indicators. Items were developed through the use of having strong psychometric properties; thus, you would
several sources: literature review, review of medical lack confidence in the usefulness of the study findings.
files, and consultation with critical care nurses and
physicians. Content validity was measured with 13
critical care nurses and 4 physicians through the use Criterion-Related Validity
of a content validity questionnaire. A content validity
index was calculated to determine the relevance of Criterion related validity is the degree of rela-
each indicator and clinicians’ agreement with the tionship between the participant’s performance
scales. on the measurement tool and the participant’s
Roberts and Ward-Smith (2010) sought content
validity for the Career Search Questionnaire by actual behaviour. The criterion is usually the
including local nursing school advisers and second measure, which is used to assess the same
counsellors for applicability and usability. Research concept being studied.
support personnel were consulted for readability.
Three educational psychologists reviewed the format Two types of criterion-related validity are con-
of the instrument to ensure that it was consistent current and predictive. Concurrent validity is
with that of established occupational inventories. the degree of correlation of two measures of the
Then, doctorally prepared nurse educators reviewed
the instrument for content validity. There was
same construct administered at the same time. A
consensus that the instrument was appropriate. high correlation coefficient indicates agreement
between the two measures. Predictive validity is
the degree of correlation between the measure of
A subtype of content validity is ace validity the concept and a future measure of the same
which is a rudimentary type of validity in which concept. Because of the passage of time, the cor-
the instrument intuitively gives the appearance of relation coefficients are likely to be lower for
measuring the concept. To establish face validity, predictive validity studies.
colleagues or participants are asked to read the McGilton and associates (2005) evaluated the
instrument and evaluate the content in terms of concurrent validity of the newly developed Rela-
whether it appears to re ect the concept that the tional Care Scale (RCS) and the Relational
researcher intends to measure. This procedure Behaviour Scale (RBS) because both scales are
may be useful in the tool development process in based on the empathic and reliable behaviours of
terms of determining the readability and clarity care providers and on Winnicott’s (1965) rela-
of the content. Face validity, however, should in tional theory. In addition, the RCS was assessed
no way be considered a satisfactory alternative to with the Relationship Visual Analogue Scale. The
other types of validity. In the development of the results indicated that the RCS was positively cor-
Adolescent Lifestyle uestionnaire (Gillis, 1997), related with the Relationship Visual Analogue
the concept of healthy lifestyle was derived from Scale (r = .63, p < .0001 n = 50), and the RBS
qualitative interviews with adolescents. The scale was positively moderately correlated with the
comprised two subscales: health-promoting and RCS (r = .42, p < .001 n = 72).
 CHAPTER 14 Rigour in Research 317

An example of predictive validity appears in the measure then by gathering data to test the
the study of McCarter-Spaulding and Dennis hypotheses and finally, on the basis of the find-
(2010), who assessed the psychometric properties ings, by making inferences about whether the
of the short form of the Breastfeeding Self- rationale underlying the instrument’s construc-
Efficacy Scale, a measure of the self-confidence tion is adequate to explain the findings.
of mothers in breast-feeding. The researcher For example, Barnason and colleagues (2002)
assessed the predictive validity of this scale by used a hypothesis-testing approach to establish
checking the correlation of the mothers’ scores the construct validity of the Barnason Efficacy
with their method of infant feeding at 4 months Expectation Scale (BEES). Construct validity
post partum because infant feeding was deter- was tested on the basis of the empirically sup-
mined to be a strong and objective comparison ported hypothesis that individuals with better
criterion. health status and functioning also would have
higher levels of self-efficacy. To explore this
Construct Validity hypothesis, correlations were made between the
Construct validity is the extent to which a test BEES total score and physiological behaviours of
measures a theoretical construct or trait. To estab- interest in a population of patients who had
lish this type of validity, the researcher attempts received a coronary artery bypass graft. Measures
to validate a body of theory underlying the mea- of physiological functioning used in this psycho-
surement and testing of the hypothesized relation- metric study were subscales of the Medical Out-
ships. Empirical testing confirms or fails to comes Study Short-Form 36, which has been used
confirm the relationships that would be predicted extensively in the literature as a measure of health
among concepts and, as such, provides more or status and functioning with established reliability
less support for the construct validity of the and validity. This instrument is a multidimen-
instruments measuring those concepts. Establish- sional scale on which health concepts are mea-
ing construct validity is a complex process, often sured. However, in this study, the physiological
involving several studies and approaches. The functioning subscales of physical functioning,
following approaches are discussed in this section: role-physical functioning (role limitations caused
hypothesis-testing, convergent and divergent, by physical problems), and general health were
contrasted-groups, and factor-analytical. used specifically because those aspects of func-
McCarter-Spaulding and Dennis (2010) also tioning were more closely related to the behav-
assessed construct validity by analyzing the rela- iours measured with the BEES.
tionship between network support and breast- The BEES mean score was correlated with
feeding self-efficacy. As hypothesized, network aspects of physiological functioning (physical,
support for breast-feeding was significantly cor- role-physical, and general health), with signifi-
related with the scores on the short form of the cant weak to moderate correlations ranging from
Breastfeeding Self-Efficacy Scores. .25 to .41. These findings provide support for the
hypothesis and therefore preliminary support for
Hypothesis-Testing Approach the theoretical basis, conceptual accuracy, and
When the hy othesis testing a roach is used, construct validity of the BEES in the patient pop-
the investigator uses the theory or concept under- ulation: Individuals with better health status have
lying the measurement instrument to validate the higher levels of self-efficacy. Because of the
instrument. The investigator accomplishes this homogeneous nature of the sample and the use of
task first by developing hypotheses about the a convenience sampling strategy, however, further
behaviour of individuals with varying scores on testing of the BEES is necessary with adults from
318 PART FOUR Processes Related to Research

various age, gender, socioeconomic, and cultural 31.69, p = .001—supported the hypothesis that as
groups. the mean activity in each group increased, the
mean frequency of the movements recorded by
Convergent and Divergent Approaches actigraphy would also increase, and the greatest
Two strategies for assessing construct validity are amount of activity occurred in the group with the
convergent and divergent approaches. highest scores. The researchers also found that
Convergent validity exists when two or more the mean activity of participants whose mobility
tools that are intended to measure the same con- subscale level was 4 (no mobility limitations) was
struct are administered to participants and are significantly greater than the activity of partici-
found to be positively correlated. A correlational pants in the other three levels, which supported
analysis (i.e., a test of relationship see Chapters the convergent validity of the mobility subscale
11 and 16) determines whether the measures are of the Braden Scale.
positively correlated, in which case convergent More recently, causal modelling has been used
validity is said to be supported. to develop and test a hypothesized explanation of
Powers and colleagues (2004) evaluated the causes of a phenomenon, and to establish conver-
convergent validity of one of the most widely gent validity, as in the development of the Care-
used tools for predicting the risk for pressure giver Reciprocity Scale by Carruth (1996). As
sores: the mobility subscale of the Braden Scale illustrated in Table 14-3, several indicators have
(Braden Bergstrom, 1994). Patient immobility been recommended to assess convergent validity:
had been identified as an important risk factor in standardized item loadings, item/composite reli-
the development of pressure sores, and the mobil- ability, examination of standard error, and esti-
ity subscale of the Braden Scale was designed to mated average variance, as extracted by each
quantify levels of mobility, on the basis of indi- construct. For the data in Table 14-3, these indica-
vidual clinical observations. tors were calculated by the Linear Structural
To assess convergent validity, actigraphy (a Relationships (LISREL) computer program.
noninvasive method of monitoring rest and activ- The item loadings indicate the extent to which
ity cycles) was used to measure participants’ fre- a single variable is related to the cluster of vari-
quency of movement. Powers and colleagues ables for example, in Table 14-3, under the
(2004) hypothesized that a lower mean frequency cluster of variables for armth and regard, item
of movement would be recorded by actigraphy 30 has a loading of .82, which is closer to the
for participants scoring low (i.e., a score of 1) on composite reliability of .89 and thus is more
the mobility subscale than for participants with related to the cluster than item 2, which has a
higher scores. Powers and colleagues further loading of .52. Item reliability is the reliability of
hypothesized that as the mobility subscale scores that item within that subscale for example, in
increased, the mean frequency of the movements Table 14-3, the reliability for item 4 on the sub-
recorded by actigraphy would also increase. scale love and affection is .60, whereas the reli-
Two consecutive data-collection periods, from ability for the subscale of love and affection with
6 A.M. to 6 P.M. in a 48-hour interval, were chosen the four items is .88. Average variance is the
for analysis. Powers and colleagues (2004) found amount of variance explained by a cluster of vari-
that only spontaneous patient movement was ables in that construct for example, of the four
noted in the data analysis of the actigraph record- subscales, love and affection has the highest esti-
ings. The statistically significant difference in the mate, .64. Standard error is the standard deviation
mean activity by subscale groups— (3, 15) = divided by the sample size. Table 14-3 presents
 CHAPTER 14 Rigour in Research 319

TABLE 14-3
ESTIMATED LISREL PARAMETERS OF CONSTRUCT VALIDITY
CONSTRUCT STANDARDIZED LOADINGS ITEM/COMPOSITE RELIABILITY 2 × SE VARIANCE EXTRACTED ESTIMATE

WARMTH AND REGARD

.89* .48
Item 2 .52 .27 —†
Item 10 .70 .48 .40
Item 12 .76 .58 .42
Item 21 .60 .35 .37
Item 22 .71 .51 .40
Item 25 .67 .45 .39
Item 28 .70 .49 .40
Item 30 .82 .66 .43
Item 33 .74 .54 .41

INTRINSIC REWARDS OF GIVING

.83* .49
Item 1 .62 .38 —†
Item 13 .64 .40 .29
Item 17 .74 .55 .30
Item 29 .76 .60 .30
Item 32 .74 .55 .30

LOVE AND AFFECTION

.88* .64
Item 4 .77 .60 —†
Item 5 .73 .53 .18
Item 8 .90 .80 .18
Item 15 .81 .65 .18

BALANCE WITHIN FAMILY CAREGIVING

.77* .47
Item 6 .72 .51 —†
Item 19 .69 .48 .24
Item 20 .75 .56 .24
Item 34 .56 .31 .23
From Carruth, A. K. (1996). Development and testing of the Caregiver Reciprocity Scale. Nursing Research, 45, 92–97.
LISREL, Linear Structural Relationships (computer program); SE, standard error.
*Denotes composite reliabilities.
†
Denotes that the value was not estimated (parameter constrained at 1.0 in LISREL).

data indicating significant findings between the ers search for instruments that measure the oppo-
factor structure of the Caregiver Reciprocity site of the construct. If the divergent measure is
Scale and the relevant causal modelling indica- negatively related to other measures, the mea-
tors, thereby offering support for the convergent sure’s validity is strengthened.
validity of the items in each factor of the Care- McGilton and associates (2003) assessed the
giver Reciprocity Scale. divergent validity of the RBS, which consisted of
In contrast to convergent validity, the calcula- three subscales related to the empathic and reli-
tion of divergent validity requires measurement able relational behaviours of people who pro-
approaches that differentiate one construct from vided care to older adults. More positive behaviour
others that may be similar. Sometimes research- of care providers was illustrated by higher scores
320 PART FOUR Processes Related to Research

on the RBS. The researchers hypothesized that interpersonal relationships (including social
the RBS would be negatively correlated with a support, reciprocity, and con ict), Tilden and
negative affect scale. This hypothesis was sup- associates (1990) used the multitrait-multimethod
ported through the use of the Philadelphia Center approach to validity assessment. The two traits of
Affect Rating Scale, on which the RBS was found social support and con ict listed in the Interper-
to be negatively correlated with anxiety (r = .59, sonal Relationship Inventory were each measured
p < .005), sadness (r = −.59, p < .005), and agita- with two different methods: a participant self-
tion (r = −.39, p < .05) thus, the construct validity report tool and an investigator-observation visual
of the RBS in measuring empathy and reliable analogue rating. Reciprocity was not included
relational behaviours was verified. More recently, because of its high correlation with social support.
the data from a factor analysis being conducted The use of multiple measures of a concept
for other validity purposes has been used to deter- decreases systematic error. The use of a variety
mine divergent (sometimes called discriminant) of data-collection methods (e.g., self-report,
validity. Carruth (1996) assessed the divergent observation, interview, and collection of physio-
validity of the four factors (subscales) of the logical data) also diminishes the effect of system-
Caregiver Reciprocity Scale by examining the atic error.
correlations between each factor or subscale that
appears in Table 14-3. Contrasted-Groups Approach
A specific method of assessing convergent and In the contrasted grou s a roach (sometimes
divergent validity is the ultitrait ulti ethod called the no n grou s a roach) to the
a roach Similar to the divergent validity development of construct validity, the researcher
approach just described, this method, proposed identifies two groups of individuals expected to
by Campbell and Fiske (1959), also involves score extremely high or extremely low in the
examining the relationships between instruments characteristic being measured by the instrument.
that are intended to measure the same construct The instrument is administered to both groups,
and between those that are intended to measure and the differences in scores are examined. If the
different constructs. A variety of measurement instrument is sensitive to individual differences
strategies, however, are used. In other words, this in the trait being measured, the mean performance
approach is a type of validation in which more of these two groups should differ significantly,
than one method is used to assess the accuracy of and evidence of construct validity would be sup-
an instrument. For example, anxiety could be ported. A t test or analysis of variance is used to
measured by the following: statistically measure the difference between the
Administering the State-Trait Anxiety Inv- two groups.
entory Secco (2002) sought to develop and assess the
Recording blood pressure readings psychometric properties of the Infant Care ues-
Asking the participant about anxious feel- tionnaire, a self-report scale designed to measure
ings a mother’s perception of her abilities and compe-
Observing the participant’s behaviour tence in providing care to her infant. The
The results of one of these measures should researcher used the contrasted-groups approach
then be correlated with the results of each of the to provide evidence of construct validity for the
others in a multitrait-multimethod matrix (Waltz, Mom and Baby dimension of the Infant Care
Strickland, Lenz, 1991). uestionnaire.
In their classic study designed to develop, vali- The known groups were healthy mothers of
date, and normalize a measure of dimensions of different parity (low-risk primiparous and mul-
 CHAPTER 14 Rigour in Research 321

tiparous mothers n = 164 of full-term infants)

and mothers with different amounts of past infant Factor-Analytical Approach
care experience. Multiparous mothers, mothers A final approach to assessing construct validity is
with higher ratings of past infant care experience, actor analysis This procedure gives the
and mothers with greater time in the mothering researcher information about the extent to which a
role were expected to have significantly higher set of items measures the same underlying con-
Mom and Baby scores than were primiparous struct or the same dimension of a construct. In
mothers with lower ratings of infant care factor analysis, the researcher assesses the degree
experience. to which the individual items on a scale truly clus-
As illustrated in Table 14-4, repeated-measures ter around one or more dimensions. Items designed
analysis of variance demonstrated significantly to measure the same dimension should load on the
higher Mom and Baby scores for the multipa- same factor those designed to measure differ-
rous mothers than for primiparous mothers by ing dimensions should load on different factors
time (2, 134) = 21.78, p = .000 and parity (1, (Anastasi, 1988 Nunnally Bernstein, 1994).
134) = 10.78, p = .001 . Over time, certain trends A factor analysis also indicates whether the
were observed: lower scores for primiparous items in the instrument re ect a single construct
mothers at all three measurement times and or several constructs. Several factors may be
diminished mean differences between parity identified in a set of data. The study must have a
groups. Secco (2002) proposed that these findings large sample size in order to conduct a factor
are associated with the maternal role attainment analysis. Nunnally and Bernstein (1994) recom-
theory, which is that knowledge, skill, and com- mended 10 observations for each variable. Thus,
petence perceptions are acquired during the post- to develop the factor structure and reliability of
natal period and that experience is a key factor in the Brain Impairment Behavior Scale, Cameron
attainment of competence. The significantly and associates (2008) used data from several
higher scores on the Mom and Baby dimension small studies in which this scale was used to
for the multiparous mothers suggest that the obtain a sample size of 300. Table 14-5 gives an
Infant Care uestionnaire was sensitive enough example of how factor analysis can be used, as it
to capture the differences between primiparous was in causal modelling.
and multiparous mothers, thereby providing evi- Roberts and Ward-Smith (2010) conducted a
dence of construct validity with the contrasted- factor analysis in the development of the Career
groups approach. Search uestionnaire. By performing exploratory
factor analysis and using principal component
analysis with a varimax rotation, they obtained
five factors that accounted for 57.4 of the total
TABLE 14-4 variance. The findings from the factor analysis
RESULTS OF REPEATED-MEASURES ANALYSIS OF are presented in Table 14-5.
VARIANCE FOR MEAN “MOM AND BABY”
DIMENSION SCORES BY TIME* AND PARITY†
POSTPARTUM PRIMIPAROUS MULTIPAROUS ALL MOTHERS Helpful Hint
WEEK (n = 97) (n = 43) (n = 140)
When validity data about a study’s measurement
1 3.74 4.09 3.85 instruments are not included in a research article, you
2 4.00 4.23 4.10 cannot determine whether the intended concept is
3 4.07 4.18 4.10 being captured by the measurement tool. Before you
use the results, check the instrument’s validity by
*F(2, n = 140) = 21.78, p < .001. reviewing the original source.
†
F(1, n = 140) = 10.78, p < .001.
322 PART FOUR Processes Related to Research

TABLE 14-5 Researchers may be concerned about whether

CAREER SEARCH QUESTIONNAIRE FACTORS AND the scores that were obtained for a sample of
HIGHEST FACTOR LOADINGS ON NURSING participants were consistent, true measures of the
ITEMS: POPULATIONS BEFORE NURSING STUDY behaviours, and thus an accurate re ection of the
FACTOR/ITEM LOADING differences between individuals. The extent of
FACTOR 1: ASSESSMENT AND SAFETY variability in test scores that is attributable to
Safety for self, others .713 error rather than a true measure of the behaviours
Flexibility to changes .706 is the error variance
Healthy boundaries .704
Concerns for privacy, dignity .688 An observed test score that is derived from a
Gather information .660 set of items consists of the true score plus error
Protect others, ethical treatment .588 (Figure 14-2). The error may be either chance
Understand physiology .482
Health teaching .471 (random) error or systematic error.
A chance error or a rando error is an error
FACTOR 2: ACTIVE PROBLEM SOLVING that is difficult to control (e.g., a respondent’s
Learn technical skills .752
Assign duties to others .735 anxiety at the time of testing). These errors are
Plan goals for others .690 unsystematic and are not predictable thus, they
Communicating during stress .576 cannot be corrected. However, awareness of the
Empowering others .545
Work with blood, human waste .533 sources of these errors may help the researcher
minimize their effect on measurement accuracy.
FACTOR 3: WORKING WITH PEOPLE These sources are as follows:
Enjoy working with people .797
Find solutions to problems .742
1. Transient human conditions, such as hunger,
Working with different people .694 fatigue, health, lack of motivation, and
anxiety, which are often beyond the aware-
FACTOR 4: SCIENTIFIC LEARNING ness and control of the examiner.
Lifelong learner .739
New terminology .659 2. Variations in the measurement procedure,
Read research .655 such as misplacement of the blood pressure
Ways to do things better .447 cuff, not waiting for a specific time period
FACTOR 5: COMPASSION before taking the blood pressure, or placing
Emotional and physical support .724 the arm randomly in relation to the heart
Help people at end of life .708 while measuring blood pressure changing
Compassion with ill people .567
the wording of interview questions between
Adapted from Roberts, C. A., & Ward-Smith, P. (2010). Choosing a administrations or environmental factors,
career in nursing: Development of a career search instrument.
International Journal of Nursing Education Scholarship, 7(1), Article 2. such as the presence of others while data are
being obtained, a cold room, or discomfort
Evidence-Informed Practice Tip with the researcher (who is part of the
When the tools used in a study are presented, environment).
note whether the sample used to develop the mea- 3. Errors in data processing, such as coding
surement instruments is similar to your patient errors and incorrect inputting into the
population.
computer.
Chance errors affect an individual’s observed
The Critical Thinking Decision Path will help score, so that the person’s observed score may
you assess the appropriateness of the type of be higher than his or her true score, whereas
validity and reliability selected for use in a par- another person’s observed score may be lower
ticular research study. than his or her true score. Instruments that are free
 CHAPTER 14 Rigour in Research 323

CRITICAL THINKING DECISION PATH

Determining the Appropriate Type of Validity and Reliability Selected for a Study

Assess the appropriateness of the type of

validity and reliability selected for a
particular research study

Consistency Accuracy

Reliability Validity

Determine Determine Determine Determine Determine Determine

stability homogeneity equivalence content criterion construct
domain related

Alternate or Test–retest Parallel or Interrater Content Concurrent Predictive

parallel alternate validity
forms forms

Face
validity
Item-to-total Split-half KR-20 Cronbach’s
correlation alpha

Hypothesis Convergent Divergent Contrasted Factor

testing groups analysis

Multitrait–multimethod
KR-20, Kuder-Richardson coefficient.

Observed score (XO) = True variance (XT) + Error variance (XE)

Actual score Consistent, Chance/Random Error

obtained hypothetical, ∑ Transient participant
stable, or factors
true score ∑ Instrumentation variations
∑ Transient environmental
factors
Systematic Error
∑ Consistent instrument,
participant, or environmental
factors
FIGURE 14-2 Components of observed scores.
324 PART FOUR Processes Related to Research

of chance errors are considered reliable (Nun- and communication. ualitative researchers seek
nally, 1978). to achieve two goals: (1) to account for the
A syste atic error or a constant error is a method and the data, which must be independent
measurement error that is attributable to rela- so that another researcher can analyze the same
tively stable characteristics of the study popula- data in the same way and make the same conclu-
tion that may bias their behaviour, cause incorrect sions, and (2) to produce a credible and reasoned
instrument calibration, or both. Such error has a explanation of the phenomenon under study.
systematic biasing in uence on the participants’ Thus, this rigour in qualitative methodology is
responses and thereby in uences the validity of judged by unique criteria appropriate for the
the instruments. Level of education, socioeco- research approach. Credibility, auditability, and
nomic status, social desirability, response pattern, fittingness are the scientific criteria proposed for
or other characteristics may in uence the validity qualitative research studies by Guba (1981).
of the instrument by altering the measurement of Although these criteria are not new, they still
the true responses in a systematic way. For capture the rigorous spirit of qualitative inquiry
example, a participant who wants to please the and are reasonable for evaluation. The meanings
investigator may constantly answer items in a of credibility, auditability, and fittingness are
socially desirable way, thus making the estimate brie y explained in Table 14-6.
of validity inaccurate.
Systematic error also occurs when an instru-
ment is improperly calibrated. Consider a scale
that consistently gives a person’s weight at 1 kg TABLE 14-6
less than the actual body weight. The scale could CRITERIA FOR JUDGING SCIENTIFIC RIGOUR:
be quite reliable (i.e., capable of reproducing the CREDIBILITY, AUDITABILITY, FITTINGNESS
precise measurement), but the result is consis- CRITERIA CHARACTERISTICS
tently invalid. Systematic error is considered part Credibility Truth of findings as judged by participants
of the true score. The multimethod-multitrait and others within the discipline. For
approach is one method of decreasing systematic example, you may find the researcher
returning to the participants to share
error. The validity of an instrument is the extent
interpretation of findings and query
to which it is free of both chance errors and sys- accuracy from the perspective of the
tematic errors (Nunnally, 1978). persons living the experience.
The amount of detail about reliability and valid- Auditability Accountability as judged by the adequacy
of information leading the reader from
ity varies considerably among research articles. the research question and raw data
When the focus of a study is tool development, through various steps of analysis to the
psychometric evaluation—including extensive interpretation of findings. For example,
you should be able to follow the
reliability and validity data—is carefully docu- reasoning of the researcher step by step
mented and appears throughout the article rather through explicit examples of data,
than brie y in the Instruments section, as in interpretations, and syntheses.
Fittingness Faithfulness to the everyday reality of the
other research studies. participants, described in enough detail
so that others in the discipline can
RIGOUR IN QUALITATIVE evaluate importance for their own
RESEARCH: CREDIBILITY, practice, research, and theory
AUDITABILITY, AND FITTINGNESS development. For example, you will
know enough about the human
As in quantitative research, the basic approach to experience being reported that you can
ensure rigour in qualitative research is methodical decide whether it “rings true” and is
useful for guiding your practice.
research design, data collection, interpretation,
 CHAPTER 14 Rigour in Research 325

individuals not involved in the research. The audit

Credibility trail was proposed by Guba (1981) to allow exter-
Credibility is a characteristic of qualitative nal auditors to follow the trail of qualitative data
research that refers to the accuracy, validity, and gathering and has been described by Lincoln and
soundness of data. It is similar to internal validity Guba (1985) as the most important trustworthi-
in qualitative research. The methods to ensure ness technique available to the naturalistic
credibility are prolonged engagement, persistent (p. 285). The audit trail involves reviewing all
observation, peer debriefing, and member checks documents relating to the study, such as research
(Lincoln, 1995). In prolonged engagement and protocol, memos and correspondences, research
persistent observation, the researchers spend suf- tools, and field notes. See the Practical Applica-
ficient time with the study’s participants to check tion box for an example of establishing rigour
for discrepancies in responses. Peer debriefing is through use of an audit trail.
conducted with experts in the field, whose probing
questions and review about the research can assist
the researchers in improving trustworthiness in Practical Application
the data. Member checking verifies the accuracy In an illustration of how rigour is ascertained,
of participants’ responses by asking the study par- Tourangeau and associates (2009) conducted
ticipants to review the themes and narratives to a qualitative study to identify nurse-reported
determinants of intention to remain employed and to
determine whether the researchers accurately develop a model explaining determinants of hospital
described their experiences (Lincoln Guba, nurses’ intention to remain employed. Tourangeau
1985). and associates established trustworthiness in the data
Triangulation, crystallization, and searching by engaging in a variety of methods: for credibility,
they used verbatim quotations to illustrate findings,
for disconfirming evidence through negative case member checking, presenting their preliminary
analyses are also used to ensure credibility and findings to some participants, and requesting
confirmability. In Chapter 7, triangulation—the participants’ views on the accuracy of interpretation.
“Transferability of findings was supported through
cross-checking and verification of data through descriptions of the time and context in which these
the use of different information sources, such as data were found. These allow readers to make
a variety of data sources, investigators, theoreti- decisions about transferability. Both dependability
and confirmability were strengthened through an
cal models, and research methods—and crystal- audit trail (e.g., transcripts, history of theme
lization in both qualitative and mixed method development)” (p. 26).
research are discussed. Triangulation is viewed as
offering completeness to naturalistic inquiry
(Tobin Begley, 2004). As noted by Seneviratne and associates (2009)
(see Appendix A), ethnographers keep track of
changes in their ideas, beliefs, and values as they
Auditability and Fittingness are engaged in the research. In this particu-
Engaging in an inquiry audit establishes both the lar study, re exive techniques were used to
auditability and the fittingness of the data. Audit control for biases from having an insider (emic)
ability is the characteristic of a qualitative study, perspective.
developed by the investigator’s research process, Luhanga and colleagues (2010) incorporated
that allows another researcher or a reader to follow several strategies to ascertain rigour. They
the thinking or conclusions of the investigator. explored the preceptor role support and develop-
ittingness is the degree to which study findings ment within the context of a rural northern mid-
are applicable outside the study situation and the sized Canadian community. Credibility was
degree to which the results are meaningful to established when the participant preceptors
326 PART FOUR Processes Related to Research

validated recognition of the research findings as 5. Specify the criteria built into the research-
re ective of their own experiences in follow-up er’s thinking.
debriefing sessions. Excerpts from preceptor tran- 6. Specify how and why participants in the
scripts were presented in the findings to facilitate study were selected.
auditability of the data analysis and findings. A 7. Delineate the scope of the research
comprehensive audit trail and record of the theo- 8. Describe how the literature relates to each
retical memos maintained auditability. An inde- category that emerged in the theory.
pendent reviewer read and commented on the Streubert and Carpenter (2007) commented that
fittingness of the results. Because the four criteria these criteria can provide a guide for critiquing
of credibility, fittingness, and auditability were grounded theory research.
met, confirmability was then achieved. Another concept to increase rigour that has not
Emden and Sandelowski (1999) inferred that received much attention is goodness, represented
one set of criteria cannot fit the bill for every as a means of locating situatedness, trustworthi-
research study. Thus, Chiovitti and Piran (2003, ness and authenticity (Tobin Begley, 2004,
p. 427) developed a framework of rigour that p. 391). Tobin and Begley recommended that
consisted of the following eight methods of goodness be re ected by the entire study and
research practice to enhance rigour during the discussed Arminio and Hultgren’s (2002) six ele-
research process and for critiquing published ments of goodness:
reports on research in grounded theory: 1. Foundation (epistemology and theory)
1. Let participants guide the inquiry process. 2. Approach (methodology)
2. Check the theoretical construction gener- 3. Collection of data (method)
ated against participants’ meanings of the 4. Representation of voice
phenomenon. 5. The art of meaning making
3. se participants’ actual words in the theory. 6. Implication for professional practice
4. Articulate the researcher’s personal views These six elements of goodness will help you as
and insights about the phenomenon a research consumer when you critique a study
explored. for research utilization.

APPRAISING THE EVIDENCE

Reliability and Validity
Reliability and validity are two crucial aspects in the criti- have an ethical responsibility to question the reliability
cal appraisal of a measurement instrument. The reviewer and validity of instruments used in research studies and
evaluates an instrument’s level of reliability and validity, to examine the findings in view of the quality of the
as well as how they were established. In a research instruments used and the data presented. The following
report, the reliability and validity for each measure should discussion highlights key areas related to reliability and
be presented. If these data have not been presented, the validity that should be evident in a research article.
reviewer must seriously question the merit and use of Appropriate reliability tests should have been per-
the tool and the study’s results. Criteria for critiquing formed by the developer of the measurement tool and
reliability and validity are presented in the Critiquing should then have been included by the current user in
Criteria box. the research report. If the initial standardization sample
If reliable and valid questionnaires are not used in a and the current sample have different characteristics, the
study, the results cannot be credible. As a critiquer, you reader would find either (1) that a pilot study for the
 CHAPTER 14 Rigour in Research 327

APPRAISING THE EVIDENCE

Reliability and Validity—cont’d
present sample would have been conducted to deter- weaknesses and try to find conclusive evidence of content
mine whether the reliability was maintained or (2) that a validity. In the body of a research article, however, it is
reliability estimate was calculated for the current sample. unusual to have more than a few sample items available
For example, if the standardization sample for a tool that for review. Thus, you should determine whether the
measures “satisfaction in an intimate heterosexual rela- appropriate assessment of content validity was used to
tionship” comprises undergraduate college students and meet the researcher’s goal.
if an investigator plans to use the tool with married Such procedures provide assurance that the tool is
couples, the reliability of the tool should be established psychometrically sound and that the content of the items
with the latter group. is consistent with the conceptual framework and the con-
The investigator determines which type of reliability struct definitions. Construct validity and criterion-related
procedure is used in the study, depending on the nature validity are two of the more precise statistical tests of
of the measurement tool and how it will be used. For whether the tool measures what it is intended to measure.
example, if the instrument is to be administered twice, Ideally, an instrument should provide evidence of content
you might determine that test-retest reliability should validity, as well as criterion-related or construct validity,
have been used to establish the stability of the tool. If an before a reviewer invests a high level of confidence in the
alternate form of the instrument has been developed for tool.
use in a repeated-measures design, evidence of alternate- You should also expect to see the strengths and weak-
form reliability should be presented to determine the nesses of instrument reliability and validity presented in
equivalence of the parallel forms. the “Discussion,” “Limitations,” or “Recommendations”
If the degree of internal consistency among the items section, or in all of these sections, of a research article. In
is relevant, an appropriate test of internal consistency this context, the reliability and validity might be discussed
should be presented. In some instances, more than one in relation to other tools devised to measure the same
type of reliability is presented, but you should determine variable. The relationship of the study’s findings to the
whether all are appropriate. For example, the Kuder- strengths and weaknesses in instrument reliability and
Richardson formula implies that a single right or wrong validity is another important discussion point.
answer exists, which makes use of the coefficient inap- Finally, the researcher should propose recommenda-
propriate with scales that provide a format of three or tions for improving future studies in relation to instru-
more possible responses. In such cases, another formula ment reliability and validity. For example, in the
is applied, such as Cronbach’s alpha. “Discussion” section of a report about developing and
Another important consideration is the acceptable validating the Double Duty Caregiving Scale, Ward-Griffin
level of reliability, which varies according to the type of and associates (2009) noted that nine factors were
test. Coefficients with reliability of .70 or higher are desir- extracted with five subscales. Additional analyses were
able. The validity of an instrument is limited by its reli- required to establish the psychometric properties of this
ability; in other words, less confidence can be placed in Scale, including test-retest reliability to assess construct
scores from tests with low-reliability coefficients. validity. The reliability coefficients for the five subscales
Satisfactory evidence of validity is probably the most ranged from .85 to .71. Reliability and validity in research
difficult determination for you as reviewer. This aspect of reports can be evaluated to varying degrees. As a research
measurement is most likely to fall short of meeting the consumer, you should not feel inhibited by the complex-
required criteria. Validity studies are time consuming and ity of these topics but may use the guidelines presented
complex, and researchers sometimes settle for presenting in this chapter to systematically assess the reliability and
minimal validity data. validity of a research study.
Therefore, you should closely examine the item Collegial dialogue is also an approach to evaluating
content of a tool when you evaluate its strengths and the merits and shortcomings of an existing instrument,

Continued
328 PART FOUR Processes Related to Research

APPRAISING THE EVIDENCE

Reliability and Validity—cont’d
as well as a newly developed one, that is reported in a basic knowledge of psychometrics, and encourages
the nursing literature. Such an exchange promotes the exploration of alternative methods of observation
the understanding of methodologies and techniques and the use of reliable and valid tools in clinical
of reliability and validity, stimulates the acquisition of practice.

CRITIQUING CRITERIA

1. Was an appropriate method sample, were the reliability and accurately represented
used to test the reliability of the validity recalculated to the participant’s
tool? determine whether the tool is meaning?
2. Is the reliability of the tool still adequate? 9. Have other professionals
adequate? 6. Have the strengths and confirmed the researcher’s
3. Was an appropriate method weaknesses of the reliability interpretation?
used to test the validity of the and validity of each instrument 10. Are the strengths and
instrument? been presented? weaknesses of the research
4. Is the validity of the 7. Did the researcher accurately appropriately addressed in the
measurement tool adequate? depict the participant’s “Discussion,” “Limitations,” or
5. If the sample from the reality? “Recommendations” sections of
developmental stage of the tool 8. Is evidence provided that the the report?
was different from the current researcher’s interpretation

CRITICAL THINKING CHALLENGES KEY POINTS

■ Discuss the three types of validity that must be
• Reliability and validity are crucial aspects of
established before a reviewer invests a high level conducting and critiquing research.
of confidence in the tool. Include examples of • Validity refers to whether an instrument measures
each type of validity. what it is purported to measure. It is a crucial
■ What are the major tests of reliability? Is it aspect of evaluating a tool.
necessary to establish more than one measure of • Three types of validity are content validity,
criterion-related validity, and construct validity.
reliability for each instrument used in a study? • The choice of a validation method is important and
Which do you think is the most essential measure is made by the researcher on the basis of the
of reliability? Include examples in your answer. characteristics of the measurement device in
■ Is it possible to have a valid instrument that is question and its use.
not reliable? Is the reverse possible? Support • Reliability refers to the ratio between accuracy and
inaccuracy in a measurement device.
your answer with instruments you might use in • The major tests of reliability are test-retest
the clinical setting with your patients. reliability, parallel- or alternate-form reliability,
■ What are some ways in which credibility, split-half reliability, item-to-total correlation, the
auditability, and fittingness can be evaluated? Kuder-Richardson coefficient, Cronbach’s alpha, and
■ How do you think the concept of evidence- interrater reliability.
• The selection of a method for establishing
informed practice has changed research reliability depends on the characteristics of the
utilization models? Is the review of the literature tool, the testing method that is used for collecting
the same when a research proposal is developed data from the standardization sample, and the
as it is when the steps of research utilization or kinds of data that are obtained.
an evidence-informed practice protocol is • Credibility, auditability, and fittingness are criteria
for judging the scientific rigour of a qualitative
implemented? Support your position. research study.
 CHAPTER 14 Rigour in Research 329

Doran, D. M., Harrison, M. B., Laschinger, H. S.,

REFERENCES Hirdes, J. P., Rukholm, E., Sidani, S., . . .
Akhtar-Danesh, N., Valaitis, R. K., Schofield, R., Tourangeau, A. E. (2006). Nursing-sensitive
nderwood, J., Martin-Misener, R., Baumann, A., outcomes data collection in acute care and
Kolotylo, C. (2010). A questionnaire for long-term-care settings. Nursing Research, (2S),
assessing community health nurses’ learning needs. S75-S81.
estern ournal of Nursing Research, (8), Emden, C., Sandelowski, M. (1999). The good,
1055-1072. the bad and relative, part two: Goodness and the
American Cancer Society. (2010). Cancer facts and criterion problem in qualitative research.
gures. Atlanta: Author. International ournal of Professional Nursing
Anastasi, A. (1988). Psychological testing (6th ed.). Practice, (1), 2-7.
New York: Macmillan. G linas, C., Fillion, L., Puntillo, K. A. (2008). Item
Arminio, J. L., Hultgren, F. H. (2002). Breaking out selection and content validity of the Critical-Care
from the shadow: The question of criteria in qualita- Pain Observation Tool for non-verbal adults. ournal
tive research. ournal of College Student Develop- of Advanced Nursing, (1), 203-216.
ment, , 446-456. Gillis, A. J. (1997). The Adolescent Lifestyle
Armstrong, K., Laschinger, H., Wong, C. (2009). uestionnaire: Development and psychometric
Workplace empowerment and magnet hospital testing. Canadian ournal of Nursing Research, ,
characteristics as predictors of patient safety climate. 29-46.
ournal of Nursing Care uality, (1), 55-62. Guba, E. G. (1981). Criteria for assessing the
Barnason, S., immerman, L., Atwood, J., Nieveen, J., trustworthiness of naturalistic enquiries. Educational
Schmaderer, M. (2002). Development of a Communication and Technology ournal, , 75-91.
self-efficacy instrument for coronary artery bypass Heaman, M. I., Gupton, A. L. (2009). Psychometric
graft patients. ournal of Nursing Measurement, , testing of the perception of pregnancy risk
123-133. questionnaire. Research in Nursing and ealth, ,
Braden, B., Bergstrom, N. (1994). Predictive validity 493-503.
of the Braden Scale for pressure sore risk in a nursing Hoskins, C. N. (1988). The Partner Relationship
home population. Research in Nursing and ealth, Inventory. Palo Alto, CA: Consulting Psychologists
, 459-470. Press.
Budin, W., Hoskins, C. N., Haber, J., Sherman, D. W., Ibrahim, S. E. R. (2002). Rates of adherence to
Maislin, G., Cater, J. R., . . . Shukla, S. (2008). pharmacological treatment among children and
Education, counselling, and adjustment among patient adolescents with attention deficit hyperactivity
and partners: A randomized clinical trial. Nursing disorder. uman Psychopharmacology, ,
Research, , 199-213. 225-231.
Cameron, J., Chung, A., Coyote, D., Streiner, D., Jack, S. M., DiCenso, A., Lohfeld, L. (2005). A
Singh, M., Stewart, D. (2008). Factor structure and theory of maternal engagement with public health
reliability of the Brain Impairment Behavior Scale. nurses and family visitors. ournal of Advanced
ournal of Neuroscience Nursing, (1), Nursing, , 182-190.
40-47. Kuiper, R.-A. (2010). Metacognitive factors that impact
Campbell, D., Fiske, D. (1959). Convergent and student nurse use of point of care technology in
discriminant validation by the matrix. Psychological clinical settings. International ournal of Nursing
Bulletin, , 273-302. Education Scholarship, (1), Article 5. doi:
Carruth, A. K. (1996). Development and testing of the 10.2202/1548-923 .1866
Caregiver Reciprocity Scale. Nursing Research, , Lincoln, Y. S. (1995). Emerging criteria for qualitative
92-97. and interpretive research. ualitative In uiry, ,
Chiovitti, R. F., Piran, N. (2003). Rigour and 275-289.
grounded theory research. ournal of Advanced Lincoln, Y., Guba, E. (1985). Naturalistic in uiry.
Nursing, (4), 427-435. New York: Sage.
Dennis, C. L. (2003). The Breastfeeding Self-Efficacy Luhanga, F. L., Dickeson, P., Mossey, S. D. (2010).
Scale: Psychometric assessment of the short form. Preceptor preparation: An investment in the future
ournal of bstetric, ynecologic, Neonatal generation of nurses. International ournal of
Nursing, , 734-744. Nursing Education, (1), Article 38, 1-15.
330 PART FOUR Processes Related to Research

McCarter-Spaulding, D., Dennis, C.-L. (2010). Seneviratne, C. C., Mather, C. M., Then, K. L.
Psychometric testing of the Breastfeeding Self- (2009). nderstanding nursing on an acute stroke
efficacy Scale-Short Form in a sample of Black unit: perceptions of space, time and interprofessional
women in the nited States. Research in Nursing practice. ournal of Advanced Nursing, (9),
ealth, , 111-119. 1872-1881, doi: 10.1111/j.1365-2648.2009.05053.x
McDowell, I., Newell, C. (1996). Measuring health: Sidani, S., Epstein, D. R., Bootzin, R. R., Moritz, P.,
A guide to rating scales and uestionnaires. New Miranda, J. (2009). Assessment of preferences for
York: Oxford niversity Press. treatment: Validation of a measure. Research in
McGilton, K. S., O’Brien-Pallas, L. L., Darlington, G., Nursing and ealth, , 419-438.
Evans, M., Wynn, F., Pringle, D. M. (2003). Streubert, H. S., Carpenter, D. (2007). ualitative
Effects of relationship-enhancing program of care on research in nursing: Advancing the humanistic
outcomes. ournal of Nursing Scholarship, , imperative. Philadelphia: Lippincott, Williams,
151-156. Wilkins.
McGilton, K. S., Pringle, D. M., O’Brien-Pallas, L. L., Tilden, V. P., Nelson, C. A., May, B. A. (1990). The
Wynn, F., Streiner, D. (2005). Development IPR inventory: Development and psychometric
and psychometric testing of the Relational characteristics. Nursing Research, , 337-343.
Care Scale. ournal of Nursing Measurement, , Tobin, G. A., Begley, C. M. (2004). Methodological
51-57. rigour within a qualitative framework. ournal of
Nunnally, J. C. (1978). Psychometric theory (2nd ed.). Advanced Nursing, , 388-396.
New York: McGraw-Hill. Tourangeau, A. E., Cummings, B., Cranley, L. A.,
Nunnally, J. C., Bernstein, I. H. (1994). Psychometric Ferron, E. M., Harvey, S. (2009). Determinants of
theory (3rd ed.). New York: McGraw-Hill. hospital nurse intention: Broadening our understand-
Powers, G. C., entner, T., Nelson, F., Bergstrom, N. ing. ournal of Advanced Nursing, (1), 22-32.
(2004). Validation of the mobility subscale of the Waltz, C., Strickland, O., Lenz, E. (1991). Measure-
Braden Scale for predicting pressure sore risk. ment in nursing research (3rd ed.). Philadelphia: F. A.
Nursing Research, , 340-346. Davis.
Resnick, B., Nahm, E. S., Orwig, D., immerman, Ward-Griffin, C., Keefe, J., Martin-Matthews, A.,
S. S., Magaziner, J. (2001). Measurement of Kerr, M., Brown, J.-B., Oudshoorn, A. (2009).
activity in older adults: Reliability and validity of Development and validation of the Double Duty
the step activity monitor. ournal of Nursing Caregiving Scale. Canadian ournal of Nursing
Measurement, , 275-290. Research, (3), 108-128.
Roberts, C. A., Ward-Smith, P. (2010). Choosing a Winnicott, D. W. (1965). The maturational process and
career in nursing: Development of a career search the facilitating environment. New York: International
instrument. International ournal of Nursing niversities Press.
Education Scholarship, (1), Article 2.
Santacroce, S. J., Maccarelli, L. M., Grey, M.
(2004). Intervention fidelity. Nursing Research, ,
FOR FURTHER STUDY
63-66. Go to Evolve at http://evolve.elsevier.com/
Secco, L. (2002). The Infant Care uestionnaire: Canada/LoBiondo/Research for Audio Glossary, how-to
Assessment of reliability and validity in a sample of instructions for Writing Proposals for Funding, and
healthy mothers. ournal of Nursing Measurement, additional research articles for practice in reviewing
, 97-109. and critiquing.
C H A PTER 1 5

Qualitative Data Analysis

Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Describe the processes of qualitative data analysis.
• Outline the steps common to qualitative data analysis.
• Describe how data are reduced to meaningful units (themes).
• Describe the process of identifying themes and categories and the relationships between them.
• Assess the validity of a data analysis from a study.

KEY TERMS
codes data display thematic analysis
coding data reduction themes
constant comparative method member checking

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

331
332 PART FOUR Processes Related to Research

AS DISCUSSED IN EARLIER CHAPTERS, qualitative personal biases are not juxtaposed with the par-
analysis is not an analysis of the statistical tests ticipant’s thoughts. The presence of the original
used in the study but an analysis of the qualitative words allows the reader to check the authenticity
text. The text includes transcripts of interviews, of the data. New researchers may transcribe the
narratives, documents, media such as newspapers recording into text themselves however, most
and movies, and field notes. ualitative research- researchers use a transcriptionist. It is recom-
ers collect enormous amounts of data, which must mended that the researcher spot-check interviews
be managed carefully: More than 150 pages of to ensure accuracy of the transcription (Streubert
transcript can result from 25 interviews. To add Carpenter, 2011).
to the complexity of qualitative data analysis,
many researchers take different approaches to DATA MANAGEMENT
analysis. This chapter expands on the discussion The open nature of qualitative inquiry typically
in Chapter 8, in which the analysis of data was results in the collection of more data than required.
introduced in the context of several qualitative Glesne (2011) referred to the sheer volume of
research traditions, such as phenomenology, the data collected as fat data. Consequently,
grounded theory, ethnography, and case study. researchers must be methodical in their organiza-
tion and management of the data. Fortunately,
AUDIO RECORDING INTERVIEWS computer software simplifies the storage and
As discussed in Chapter 7, qualitative researchers retrieval of data. In addition, researchers are also
gather data from a variety of sources, including required to develop a decision or audit trail, which
interviews, observations, narrative, and focus necessitates the tracking of the participants, the
groups (Streubert Carpenter, 2011). Interviews original audio recordings, and original and pho-
are the most common source and serve as the tocopied documents. Moreover, all of the data
primary source of data for many qualitative must be kept secure to maintain confidentiality.
research projects. For example, Thorne and asso- Computer software to organize and retrieve
ciates (2009) completed face-to-face interviews data is referred to as computer-assisted qualita-
with each of the participants and followed up with tive data analysis software (CA DAS). There are
bimonthly telephone interviews (see Appendix many computer programs to choose from, such
D). In addition to the field observations conducted as ATLAS.ti, Ethnograph, HyperRESEARCH,
in an ethnographic study, Seneviratne and associ- Inspiration, SR NVivo, SR Sight, and C-I-
ates (2009) conducted interviews to explore the SAID. When choosing software, researchers have
participants’ work practices (see Appendix A). to understand what their needs are to determine
Although some researchers believe that a record- the best fit. Meadows and Dodendorf (1999) cat-
ing device inhibits the free ow of discussion, egorized computer programs into the following
Seidman (1998) and other authors have found three types:
that most participants and interviewers forget 1. Code-and-retrieve programs, which assist
about the presence of the device. Consequently, in organizing and grouping data (e.g., Data
most researchers record interviews and then tran- Collector)
scribe them verbatim into written text. Some 2. Theory builders, which move to a different
researchers may consider summarizing or para- level of data organization by connecting
phrasing the spoken words (Seidman, 1998), themes and categories (e.g., SR NVivo)
but this is not commonly practised. Most resear- 3. Conceptual network builders, which incor-
chers wish to use the original words from the porate graphics with theory-building capa-
participants so that their own interpretations and bilities (e.g., Inspiration)
 CHAPTER 15 Qualitative Data Analysis 333

nlike computer programs used with quantita- what additional data need to be collected. Many
tive data, these programs do not analyze data. researchers believe that the stages of data collec-
Data analysis and interpretation remain largely tion and data analysis should be integrated
the task of the researcher. In other words, (Denzin Lincoln, 2000 Miles Huberman,
CA DAS cannot think for the researcher 1994 Streubert Carpenter, 2011), whereas
(Glesne, 2011, p. 207). However, using computer others believe that these stages should be separate
programs for orderly organization and grouping (Seidman, 1998).
of data facilitates the researcher’s job of analysis The overall goal of qualitative data analysis is
and interpretation. to make meaning out of massive amounts of text
The researcher needs to test software to deter- or data, and many methods for analysis are avail-
mine which program will be the most useful. able. Patton (2002) encouraged researchers to do
Often, this process is one of trial and error before their very best with . . . full intellectual capacity
the most appropriate computer program is found. to fairly represent the data and communicate what
Most Web sites allow the downloading of a dem- the data reveal given the purpose of study
onstration trial for a short time. Reviewing online (p. 433). As described earlier, qualitative analysis
tutorials (e.g., http://www.qsrinternational.com/ is not a linear process rather, it is cyclical, trans-
solutions multimedia.aspx) can also assist the formative, reciprocal, and iterative. Miles and
researcher in selecting the most appropriate soft- Huberman (1994, p. 9) identified some common
ware. Although learning new software is very features among different approaches to qualita-
time consuming, the benefits of using computer- tive data analysis:
ized software outweigh the time spent research- Affixing codes to a set of field notes taken
ing and learning about it (Streubert Carpenter, during observations or interviews
2011). ualitative discussion groups such as Noting re ections of other remarks in the
ALR-L LISTSERV. GA.ED , hosted at margins
the niversity of Georgia, may also provide valu- Sorting and shifting though these materials
able insight and tips. to identify similar phrases, relationships
between variables, patterns, themes, dis-
tinct differences between subgroups, and
OVERVIEW OF DATA ANALYSIS common sequences
When does data collection end and data analysis Isolating these patterns, processes, com-
begin This is a controversial area of qualitative monalities, and differences and taking them
research because not all researchers agree on out to the field in the next wave of data
whether data collection should be completed collection
before analysis begins or whether the two Gradually elaborating a small set of gener-
processes ought to take place concurrently. alizations that cover the consistencies dis-
Therefore, the researcher needs to identify the cerned in the database
process used. Confronting those generalizations with a
As mentioned previously, many researchers formalized body of knowledge in the form
begin a preliminary analysis as the material accu- of constructs or theories
mulates. Typically, the qualitative researcher Guidelines such as these are useful, but they
transcribes all of the interviews, field notes, and serve only as recommendations. Each qualita-
observations as they are collected. As each piece tive study is unique and is reliant on the crea-
of data is transcripted, researchers begin a pre- tivity, intellect, style, and experience of the
liminary analysis during which they determine researcher.
334 PART FOUR Processes Related to Research

During the data analysis phase, all researchers methods used in nursing research fit into this
fully immerse themselves in the data over a general view of qualitative analysis.
period of weeks to months. This process requires
constant reading and rereading of the text until an Data Reduction
understanding of what the data conveys is reached According to Miles and Huberman (1994), data
(Streubert Carpenter, 2011). Many researchers reduction is the process of selecting, focusing,
also listen to the recorded interviews several simplifying, abstracting, and transforming the
times to increase their understanding and to data that appear in written-up field notes or tran-
remember the emotive component. sing an elec- scriptions (p. 10). This process is ongoing as
tronic player is helpful during this intense period data are collected. Initially, the data can be orga-
of immersion. For example, during the interviews nized into meaningful clusters of data by group-
in Cameron’s (2005) study, some participants ing related or similar data. Often, these clusters
were very emotional as they relived their experi- or groups of data are labelled as the es or struc-
ences the transcribed text did not re ect the emo- tured meaning units of data that occur frequently
tions. Observations written by the researcher can in the text. The atic analysis—the process of
capture these important elements. An important recognizing and recovering the emergent
part of the data analysis is the interplay between themes—is an important aspect of organizing
data gathering or questioning and verifying what data. In the qualitative tradition, as van Manen
is heard and understood. Researchers continue to (1997) stated, grasping and formulating a the-
ask whether what they understood before is still matic understanding is not a rulebound but a free
relevant after subsequent interviews, observa- act of seeing’ meaning (p. 79).
tions, and reading of related documents. This Glesne (2011) described several methods to
cyclic nature of questioning and verifying is an help researchers discover the meanings embed-
important aspect of data collection and analysis ded in data. The first method is to write memos
(Streubert Carpenter, 2011, p. 46). or keep a re ective journal during the data-
Miles and Huberman (1994) referred to three collection stage, which allows researchers to
discrete stages of data analysis: data reduction, record thoughts about the data as these thoughts
data display, and conclusion drawing and veri- occur. Analytical files are developed to sort data
fication (Figure 15-1). Many of the common into general categories, such as interview
Data-collection period

⎩
DATA REDUCTION ⎪
⎪
Anticipatory During Past ⎪
⎪
DATA DISPLAY ⎨
⎪ –ANALYSIS
During Past ⎪
⎪
CONCLUSION DRAWING/VERIFICATION ⎪
⎧
During Past
FIGURE 15-1 Stages of data analysis.
From Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis. Thousand Oaks, CA:
Sage (Figure 1.3, p. 10). Qualitative data analysis: An expanded sourcebook by MILES, MATTHEW
B.; HUBERMAN, A. MICHAEL. Copyright 1994 Reproduced with permission of SAGE PUBLICA-
TIONS INC BOOKS in the format Other book via Copyright Clearance Center.
 CHAPTER 15 Qualitative Data Analysis 335

questions, people, and places, as well as useful example, many researchers conducting narrative
quotations from the interviews and relevant quo- inquiry do not use coding, data reduction, and
tations from the literature. These files help orga- some of the other commonly used methods of
nize researchers’ thoughts and those of others. data analysis. Moreover, remember that meanings
Next, Glesne (2011) recommended the devel- are waiting to be discovered, not imposed on
opment of rudimentary coding schemes. Coding, the data.
as Denzin and Lincoln (2000) described it, is the The next step, finding themes, is the most
heart and soul of whole-text analysis (p. 780). exciting step of the process, which occurs during
Coding is a progressive marking, sorting, resort- and after data collection. These themes or basic
ing, and defining and redefining of the collected units of analysis can be entire texts (e.g., inter-
data. Coding allows researchers to transform the view transcripts, responses to surveys), gram-
unstructured and messy data to ideas about what matical segments (words, phrases, sentences,
is going on in the data (Richards Morse, 2007, paragraphs), formatting units (rows, pages), or
p. 133). clusters of texts that re ect a single theme. Most
Last, Glesne (2011) recommended that research- researchers try to divide data into units of analysis
ers write themselves monthly field reports as a that do not overlap with others. Researchers
way of systematically reviewing the progress and approach this step in a variety of ways for
determining the next steps. Aside from helping example, experts in grounded theory recommend
researchers keep track of their progress and com- that the researcher read the text line by line.
municate progress with other members of the The coding process itself is analysis (Miles
research team, monthly summaries often result in Huberman, 1994). Glesne (2006) stated that
new insights and new ways of approaching the coding is as simple as identifying what is impor-
research. tant and giving it a name (code) (p. 154). Codes
Denzin and Lincoln (2000) provided an over- are simply tags or labels that are assigned to the
view of the fundamental steps in the coding of themes often, the code itself is only one to four
data: sampling, identifying themes, building words long. Major codes may exist along with
codebooks, and marking texts. Richards and subcodes. Codes evolve during the analysis more
Morse (2007) described three types of coding: may be added, and others may be blended together.
descriptive, topic, and analytic. Researchers use They mean something to the researcher and are
some or all of these steps when coding. Descrip- not typically included in the research study. As
tive coding helps the researcher keep track of the coding and themes are fine-tuned and final-
factual knowledge (e.g., gender). In topic coding, ized, much of the analysis is completed.
used most commonly, the data are grouped The next step is to build codebooks by orga-
together by topic to re ect on all the different nizing codes into lists, composed of either words
ways people discuss particular topics, to seek pat- or numbers that are used by the researcher. Then
terns in their responses, or to develop dimensions the text is marked, whereby the codes are assigned
of that experience (Richards Morse, 2007, to the units of text. During this process, the
p. 134). As the categories become more compli- researcher is immersed in the data, which results
cated, the topic coding becomes analytic. Ana- in new insights and interpretations.
lytic coding is more theoretical and leads to the Richards and Morse (2007) described two
development of themes. Although coding may primary steps to data analysis: categorizing and
sound complicated to you, remember that this conceptualizing: Categorizing is how we under-
process is evolutional, and it varies from project stand and come to terms with the complexity of
to project and from researcher to researcher. For data in everyday life (p. 157). Coding is one
336 PART FOUR Processes Related to Research

TABLE 15-1
DOING ABSTRACTION IN THREE DIFFERENT METHODS
WHERE DOES WHAT ANALYTICAL
WHEN DOES ABSTRACTION COME HOW IS ABSTRACTION OUTCOME IS BEING
METHOD ABSTRACTION OCCUR? FROM? DONE? SOUGHT?

Phenomenology Not until one has the Themes and meanings Deep immersion, focus, To describe the essence
data: Previous ideas in accounts, texts thorough reading of a phenomenon
and knowledge are
bracketed
Ethnography Prior knowledge of Knowledge of social Rich description; To identify themes and
site, situation; and economic combination of patterns; to explain
understanding setting; observation qualitative and and account for a
develops during and learning from quantitative patterning, social and cultural
field research the setting coding, comparing, situation
reviewing field notes
Grounded Abstraction is from Categories derived from Theoretical sensitivity; To identify a core
theory the data, but can data (observations or seeking concepts and category and theory
be informed by line-by-line analysis their dimensions; open grounded in data
previously derived of texts); constant coding,
theories comparison with dimensionalizing,
other situations or memo writing,
settings diagramming
Reproduced with permission from Richards, L. & Morse, J. M. (2007). Readme first for a user’s guide to qualitative methods (2nd ed). Thousand Oaks,
CA: Sage (p. 159). Readme first for a user’s guide to qualitative methods by MORSE, JANICE M.; RICHARDS, LYN Copyright 2007 Reproduced with
permission of SAGE PUBLICATIONS INC BOOKS in the format Textbook via Copyright Clearance Center.

method for categorizing the data: however, other TABLE 15-2

researchers in qualitative studies can think about CODE ASSIGNMENTS AS DESCRIBED BY WUEST
data without coding. Many studies do not extend AND COLLEAGUES
beyond categorizing if their research goal is to CODE TEXT EXAMPLE
describe what is going on in a specific phenom- Time for self I steal an hour during the day.
enon. Conceptualizing moves up the ladder of Social interaction I work not only for the money but to
abstraction (see Chapter 2) to build frameworks get out . . . for the social contacts.
Cultivating the So, after I get the kids to bed, I
of concepts or theory. Phenomenology, ethnogra- marital usually go downstairs and sit with
phy, and grounded theory are all methods that relationship him [partner] for an hour.
necessitate conceptualization. In Table 15-1, the
From Wuest, J., Merritt-Gray, M., Berman, H., & Ford-Gilboe, M. (2002).
differences in the methods of abstraction are Illuminating social determinants of women’s health using grounded
described by means of the following questions: theory. Health Care International, 23, 794-808.

When does abstraction occur

Where does abstraction come from by line while asking herself two questions: What
How is abstraction done is this a conceptual indicator of and What is
What analytical outcome is being sought going on here (Wuest, Merritt-Gray, Berman,
Some researchers explicitly describe the pro- Ford-Gilboe, 2002). Codes were assigned to
cess of data analysis. For example, Wuest and each grouping of data, from sentences to para-
colleagues (2002) described the data analysis of graphs to whole pages (Table 15-2). The codes
grounded theory with an example of an earlier (time for self, social interaction, and cultivating
study by Wuest (2001) on women’s caring. Wuest the marital relationship) were eventually grouped
read each field note, transcript, or document line together into a single category: replenishing.
 CHAPTER 15 Qualitative Data Analysis 337

In the study by Seneviratne and associates comparative analysis process, the emerging
(2009 see Appendix A), data were gathered from themes were clarified with each interview.
field notes that were based on observations in the Cameron (2003) analyzed the transcripts of
naturalistic setting and on interviews. The interviews with 13 participants about their experi-
researchers analyzed the field notes and discov- ences transferring from a college to a university
ered three main themes. The researchers then kept in a collaborative baccalaureate nursing program.
returning to the field for further observations to She identified an initial set of themes at three
cross-check their findings. Thorne and associates points: as they emerged while she transcribed the
(2009) used a semistructured interview guide for interviews, after she relistened to the recorded
the first face-to-face meeting to interview partici- interviews, and after a reading and a rereading of
pants. The initial analysis identified themes to the transcripts. Differences, similarities, contra-
guide the subsequent bimonthly interviews that dictions, and gaps in the data were noted as she
took place for up to 2 years. Through this constant returned to the original material to confirm the

TABLE 15-3
CAMERON’S (2005) CODING SCHEME
MAJOR THEME SUBTHEME CODE

Academic shock Workload AS-Work

Underprepared AS-Uprep
Curriculum AS-C
Unfulfilled promises AS-prom
Academic shock AS
Professional transformation Hospital to community PT-H to C
Practice to theory PT-P to T
Professionalization PT-prof
Geographic relocation Financial GR-Fin
Time and commuting GR-T/C
University structure GR-US
Campus size GR-Size
University culture GR-Cul
Guinea pigs GR-GP
Transition stress Self-doubt TS-SD
Going back TS-GB
Family and personal responsibility TS-F
Rumours TS-R
Emotions TS-E
What social life? Separation S-sep
Loneliness S-L
Bonding S-B
Adaptation Supports A-Supp
Personal attributes A-P
Learning the ropes A-L
Satisfaction A-Sat
Personal change A-C
Recommendations—college A-RC
Recommendations—university A-RU
From Cameron, C. (2005). Experiences of transfer students in a collaborative baccalaureate nursing program. Community College Review, 33(2), 22-44.
Reprinted by Permission of SAGE Publications.
338 PART FOUR Processes Related to Research

findings. As a result of highlighting the students’ the researcher understand the data and can be in
stories, 29 subthemes emerged. From these, six the form of graphs, owcharts, matrices, or any
major themes surfaced. Table 15-3 shows the other visual representation. Like the rest of the
coding scheme used to code the transcripts, and analysis, the data display changes as more is
Table 15-4 shows how one small section of the known about the phenomenon under study. For
text was coded to validate the themes after the example, Seneviratne and associates (2009) iden-
data had been collected. tified three themes (domains) and eight subthemes
(theme components) resulting from their analysis
Data Display of the nurses’ perception of the contexts of caring
The second major step in data analysis is the data for survivors of acute stroke (Table 15-5). The
display. Miles and Huberman (1994) defined themes framed how nurses understand the chal-
data dis lay as an organized, compressed lenges in organizing stroke care: (1) space con-
assembly of information that permits conclusion cerned working in exceptionally close quarters
drawing and action (p. 11). This display helps (2) time concerned how nurses had to find
novel ways to provide care when there was not
enough time and (3) interprofessional practice
TABLE 15-4
referred to the importance that nurses ascribed to
SAMPLE CODING communications and collaboration among the
TEXT CODE
interdisciplinary professionals. Ford-Gilboe and
“The first week of school?—Going out of my TS-E colleagues (2005) studied how women and their
mind—coming home and crying. It was the
children promote their health after leaving abusive
expectations, they seem to be going on and
on about these expectations that they were male partners. The researchers found that the
expecting of us. I don’t think the information families limit intrusion by the male partner
that was given to us was, I don’t think was AS-Work through a process of strengthening capacity by
clarified properly or they didn’t say it in a
way to us that made us want to do it. It was means of four subprocesses: providing, regener-
. . . I felt very intimidated by the way they ating family, renewing self, and rebuilding secu-
were talking to us by the expectations . . . rity. The structural components and how the
you are going to do this and this and this . . .
starting next week. I’m just in there—going AS-Uprep
out of my mind . . . . I don’t know if I am TABLE 15-5
going to be able to do this and it was again THEMES AND SUBTHEMES IN RESULTS FROM
the self-doubt. It was . . . I don’t think I’m APPENDIX A
going to be able to handle this. I want to go
THEMES SUBTHEMES
back to my college. And I remember me and TS-SD
my friends were just outside of our class and AS-Uprep Space Nursing in a submarine
we were just like . . . this is unbearable and TS-GB Nursing too close
a couple of my friends wanted to call the Nursing under a state of “code
college to see if they could get back in and . burgundy”
. . I knew if I could just last this week and
last through a month . . . then I will see . . . Time Lack of time
Because I wasn’t going to give up, just then. TS-E Preserving time
But it was definitely an eye opening Time with and without space
experience. It wasn’t at all what I expected. A-P Interprofessional Relationships between stroke
The professors were great but they just, it AS-Uprep practice professionals
was too much information during that one AS-Work Communication and collaboration
week of the orientation.”
From Seneviratne, C. C., Mather, C. M., & Then, K. L. (2009).
From Cameron, C. (2005). Experiences of transfer students in a Understanding nursing on an acute stroke unit: Perceptions of space,
collaborative baccalaureate nursing program. Community College time and interprofessional practice. Journal of Advanced Nursing,
Review, 33(2), 22-44. 65(9), 1872-1881. doi: 10.1111/j.1365-2648.2009.05053.x
 CHAPTER 15 Qualitative Data Analysis 339

subprocesses relate to the theoretical structure are Zion, a young minority student, immigrated to Canada
depicted in Figure 15-2. as a child. She attended college directly after graduating
Although many researchers use figures and from high school. She chose the collaborative program
because she was aware that the degree would be
charts as part of their data display, profiles or mandatory for nursing in the future and her mother felt
vignettes can also display what is to be learned that nursing, as a career choice, was perfect. During the
from the participant’s experience. Vignettes of the college portion she struggled academically as she was
participant’s experience can summarize what was commuting four hours per day. However, through
learned from each participant and can then be perseverance, she managed to achieve a “B” average in
her second year at college and was admitted to the
shared with each participant for validation
university portion of the program. Zion rated her
(Seidman, 1998). This narrative form transforms transition as fairly difficult. Her primary concern was
the text into a story—a compelling way of sharing financial and resulted in the delay of purchase of
meaning. textbooks and course materials. The ensuing stress led
For example, Cameron (2003) studied the lived to a breakdown in class. With the intervention of a
experience of transfer students in a collaborative professor, she stuck it out. Now in the final days of her
first year at the university, Zion reflects that she settled
nursing program and, as part of the mixed-methods in by February and at that point realized that she would
design, interviewed 13 students. Vignettes were not quit and could make it. Driving Zion is her
developed for each interview participant, who, in determination and to be a good example for other
turn, selected a pseudonym that had personal minority students living below the poverty line. Her
meaning. The vignette was shared with each par- motto is “I don’t let circumstances determine my
outcome.” Zion reports that this transition has made her
ticipant, who then had the opportunity to change or
stronger, more self-confident, and that overall she will
modify the description. An example of one vignette be a better nurse. (Cameron, 2003, p. 120)
is as follows:
CONDITIONS

Family ideals
Connections with family, friends, and community
Developmental stages of children
Service accessibility/quality

Providing
POSITIONING Regenerating family
SURVIVING
FOR THE FUTURE Renewing self
Rebuilding security

INTRUSION

LOW HIGH
FIGURE 15-2 Subprocesses in strengthening capacity.
Ford-Gilboe, M., Wuest, J., & Merritt-Gray, M. (2005). Strengthening capacity to limit intrusion:
Theorizing family health promotion in the aftermath of woman abuse. Qualitative Health
Research, 15(4), 477–501. doi:10.1177/1049732305274590. Reprinted by Permission of SAGE
Publications.
340 PART FOUR Processes Related to Research

Rich descriptions, such as those found in as I possibly could get, but also cognizant of the fact
vignettes or direct quotations, enliven the data that I had a limited ability to take it in. (p. 1389)
and give meaning to people’s experiences. Most Table 15-6 includes selected quotations from par-
qualitative research includes selected quotations ticipants to support the themes emerging from
to illustrate the themes and to provide readers Seneviratne and associates’ (2009) study.
with the opportunity to understand and validate When the data are presented, the most impor-
the themes chosen by the researcher. For example, tant consideration for the research is to ensure
in studying patient-provider communications at that the presentation supports the findings and
the time of cancer diagnosis from the patients’ relays what needs to be known (Streubert Car-
perspective, Thorne and associates (2009) chose penter, 2011). The purpose of the study deter-
the following quotation to exemplify the shift in mines how the story is told. If the method is
the intensity and recollection of the original descriptive phenomenology, the focus is on the
account of cancer diagnosis: description of the lived experiences, whereas in a
grounded theory study, the focus is on a more
There’s sort of a haziness of memory about the actual
sequence of events but the feelings that I felt at the time careful description of how the narrative gives rise
are still pretty clear, you know. I felt concerned and to the analysis and interpretation, which results in
worried. And you’re so thirsty for as much information theory development.

TABLE 15-6
EXAMPLES OF SELECTED QUOTATIONS TO SUPPORT THE THEMES FROM APPENDIX A
THEMES EXAMPLE OF A QUOTATION

Space: Nursing in a submarine “Our submarine . . . it’s just a more condensed unit. But the thing that most bothers me
is it’s not centred. If you have patients in the last room . . . at the other end you are
not in close proximity to anything—you’re alone. That drives me crazy because the
nursing station is so far away.”
“I’m too claustrophobic on this unit. It’s like I am closed in . . . if you look down the hall
from the nursing station you feel like the walls and curtains are closing in around you.
It is so narrow. I feel constricted because I cannot work in a cramped space. I bump
into other people all the time.”
Time: Lack of time “We are always injuring ourselves because we rush around. There is just not enough
time for us to do things properly without patients. . . . So if things get missed, so be
it.”
“It is easier to take over for patients, dressing them or brushing their teeth, rather than
helping them do the tasks. It is a matter of accomplishing what is required for patients
in a specific window of time.”
Interprofessional practice: ”We are not recognized for the mobility things we do or in any concerns we have about
Relationships between stroke our patients. So, sometimes we don’t work hard at it. The physios are only concerned
professionals that we get the patients ready for their rehab time at the gym. So we do that for them
and then concentrate on our patient’s medical needs.”
“It should be interprofessional, ideally. In general, stroke units are interdisciplinary. Only a
small part of stroke unit care is the physician roles. So during most of stroke care,
beyond the acute phase, when you have somebody settled, the physician’s role is
relatively minor. It’s all about excellent nursing care and rehabilitation. So the team, by
accident of history and hierarchy, is led by a physician but we have an NP, all the
nurses, the physiotherapist, and social work . . . Everybody is involved in care
including home planning, etc.”
Based on Seneviratne, C. C., Mather, C. M., & Then, K. L. (2009). Understanding nursing on an acute stroke unit: Perceptions of space, time and
interprofessional practice. Journal of Advanced Nursing, 65(9), 1872-1881. doi: 10.1111/j.1365-2648.2009.05053.x
 CHAPTER 15 Qualitative Data Analysis 341

4. Making metaphors (using a literary device

Conclusion Drawing and Verification in which different things are compared to
Conclusion drawing starts at the beginning of make sense of the experience)
data collection but is not finalized until the project 5. Counting (noting that something is hap-
is completed. Although qualitative research is pening a number of times)
inductive, it is tempting to draw conclusions pre- 6. Making contrasts or comparisons (com-
maturely. The challenge for the researcher is to paring sets of things)
remain amenable to new ideas, themes, and con- 7. Partitioning variables (breaking down the
cepts as they appear. themes into smaller units)
Conclusion drawing is essentially the descrip- 8. Subsuming particulars into the general
tion of the relationship between the themes. Rich- (using a higher level of abstraction)
ards and Morse (2007) described this process as 9. Factoring (generating words factors to
doing abstraction (p. 158), in which data are express common findings)
moved from categories (codes and themes) to 10. Noting relationships between variables
concepts and constructs. As discussed earlier and (depicting the relationships between the
shown in Table 15-1, the ways of abstracting vary findings)
with the type of method. Grounded theory for- 11. Finding intervening variables (discern-
malizes this stage through the development of ing other variables that may link findings
models, which lead to theory. Verification occurs together)
as the data are collected this process can vary 12. Building a logical chain of evidence
from questioning one’s own conclusion through (validating each of the relationships
the rechecking of the text to verification by col- identified)
leagues and to finding new cases and applying the 13. Making conceptual or theoretical
model to them. In grounded theory, researchers coherence (linking the findings into an
use the constant co arative ethod in which overarching how and why of the phe-
new data are compared as they emerge with data nomenon under study)
previously analyzed. Refer to Miles and Huberman (1994) for more
Miles and Huberman (1994) stated that that detail about these tactics. Richards and Morse’s
this process of making sense of the data is a skill (2007) text on qualitative methods also describes
that all nurses have. People make sense of the the step-by-step process.
world around them by organizing and interpreting To verify the emergent themes and subthemes,
it this skill is applied to drawing and verifying Seneviratne and associates (2009) returned to
conclusions. Miles and Huberman listed the fol- the field to recheck their findings. Verbatim quo-
lowing 13 tactics for generating meaning (pp. tations provided a context for the themes (see
245-246): Table 15-5). Thorne and associates (2009) used
1. Noting patterns and themes (repetitive or the themes to guide conversations in subsequent
recurring patterns among many separate interviews to further develop the emerging ana-
pieces of data) lytic themes and to clarify individual variations
2. Seeing plausibility (realizing that the across accounts (p. 1385).
finding or conclusion sounds true or makes Other researchers can use different methods
sense) to validate their themes. For example, Bottorff
3. Clustering (grouping together things that and colleagues (2004) used NVivo to search
seem to share characteristics) the data for content related to explanations of
342 PART FOUR Processes Related to Research

nicotine dependence, addiction, lack of control Trustworthiness is also important for determining
over smoking, and experiences associated with the validity of the data interpretation or analysis.
cessation. These data were then subjected to a To ensure the trustworthiness of their findings,
thematic analysis in which the participants’ qualitative researchers must ask themselves the
explanations of addiction were compared. To following questions (Hollway Jefferson, 2000):
verify the findings, the research team regularly What do you notice The researcher has
discussed the analysis and interpretation. Once captured some impressions about the data
the initial analysis was complete, secondary anal- however, information may be missing.
ysis took place with a second set of interviews. Why do you notice what you notice
As a final analysis, further interviews took place Researchers must consider their own biases
with eight selected participants to validate the and predispositions as they interpret the data
findings. to produce trustworthy interpretations.
No matter what method is used, researchers How can you interpret what you notice As
ask themselves, What have I learned How do I discussed in Chapter 14, credibility stems
understand this, make sense of it and see the con- from prolonged engagement and persistent
nections in it (Seidman, 1998). The conclusions observation. To be able to complete a full
drawn are simply to describe, make contribu- interpretation, the researcher must spend a
tions and contribute to greater understanding, or sufficient amount of time in the field to build
at least, more informed questioning (Glesne, sound relationships with the participants.
2011, p. 210). As discussed in Chapter 7, through How can you know that your interpretation
the processes of re exivity and bracketing, is the right one The quickest way to
researchers constantly compare their findings know whether the interpretation is accurate
with their own personal beliefs and knowledge to is through sharing the findings with the par-
ensure that the analysis re ects the participants’ ticipants. This sharing is an integral part of
beliefs rather than their own. participatory action research, as outlined in
Chapter 8, and is referred to in many studies
SPECIFIC ANALYTIC PROCEDURES as e ber chec ing The researcher is
The processes of data analysis vary according to also checking whether the connections
the type of qualitative research. Table 15-7 sum- between the categories or themes are logical.
marizes the methods of analysis in qualitative Inviting other experts to review the data
methods, including phenomenology, ethnogra- analysis is another option for many research-
phy, grounded theory, and case study. Excerpts ers. In addition, some researchers analyze
from Canadian studies are included to exemplify their data from several different frameworks
the methods. (a form of triangulation) to increase the
trustworthiness of the data analysis.
TRUSTWORTHINESS Finally, it is important to consider the limita-
As described in Chapter 14, rigour in qualitative tions of the study. Many researchers describe the
research is determined by credibility, auditability, issues they faced so that readers will understand
and fittingness as the criteria for evaluation. the research in the proper context (Glesne, 2006).
 CHAPTER 15 Qualitative Data Analysis 343

TABLE 15-7
METHODS OF ANALYSIS
TRADITION METHOD OF ANALYSIS EXAMPLE

Phenomenology: • Immersion in the data: listen to Charlebois and Bouchard (2007): “ ‘The Worst Experience’:
includes a variety recordings, read and reread The Experience of Grandparents Who Have a Grandchild
of traditions transcripts with Cancer”
• Identify and extract significant Eight grandparents were interviewed. Three central themes
statements emerged: living the “worst experience”; giving support,
• Determine relationships among the a crucial role for grandparents; and feeling supported to
extracted statements (themes) carry on. The authors synthesized their findings in this
• Prepare exhaustive description of way: “The essence of the phenomenon is described as
the phenomena and the follows: having a grandchild with cancer is, for all the
relationship among the themes grandparents in the study, the worst experience they
• Synthesize the themes into a can ever live, a vital duty to support members of the
consistent description or statement family, a duty that is closely related to the perception
of the phenomenon under study they have of their grandparental role, and a need to feel
(essence) supported by calling upon several strategies to better
carry on” (p. 27).
Ethnography • Immersion in the data Seneviratne, Mather, & Then (2009): “Understanding
• Identify patterns and themes Nursing on an Acute Stroke Unit: Perceptions of Space,
• Complete a cultural inventory Time and Interprofessional Practice”
• Interpret the findings The authors described their method as follows: “Analysis
• Compare the findings with those in of field notes focused on identifying central domains,
the literature specific domain components and related work
typologies. . . . Through an ongoing process of reading
field notes, transcripts, and then returning to the field
setting for further observations, we crosschecked our
findings and were assured that our study domains,
components and related subcategories were culturally
salient” (p. 1874).
Grounded theory • Examine data carefully line by line Schreiber and MacDonald (2010): “Keeping Vigil over the
• Divide data into discrete parts Patient: A Grounded Theory of Nurse Anaesthesia
• Compare data for similarities and Practice”
differences The authors described their study as follows: “We used
• Compare data with other data constant comparative method of grounded theory,
continuously in a process: constant which involves line-by-line coding to compare incident
comparative method to incident and concept to concept. Second level or
• Cluster codes to form categories theoretical coding involves grouping concepts into
• Expand and develop categories or higher level codes and exploring relationships among
collapse them into one another them. The emerging conceptualizations drive additional
• Determine relationships between sampling and data collection . . . the resulting basic
categories social process . . . presents how CRNA’s enact their daily
practice” (pp. 553-554).
Case study • Identify unit of analysis (person, Wittich and Southall (2008): “Coping with Extended
family, organization) Facedown Positioning After Macular Hole Surgery”
• Code continuously as data are In this study, all data were gathered from one patient’s
collected diary. The authors described their analysis as follows:
• Find commonalities and themes “First each section was summarized to reveal the main
• Analyze field notes themes. Each theme was designated with a meaningful
• Review description of themes to name (i.e., sleep information). This data driven approach
identify patterns and connections to thematic analysis resulted in a preliminary coding
between them schema. The two investigators then together read line
by line through the text, scrutinizing the content while
adjusting the coding theme accordingly. . . . Using the
ATLAS-ti query tool, meaningful patterns were sought in
the text” (p. 438).
344 PART FOUR Processes Related to Research

APPRAISING THE EVIDENCE

Qualitative Data Analysis
The general criteria for critiquing qualitative data analysis and reread the transcripts, and spend substantial time in
are proposed in the Critiquing Criteria box; however, the field. Although the methods may differ, the text is
remember that many different approaches to data analy- coded to search for themes and categories through a
sis exist. The data analysis is consistent with the research process of data reduction. The emergent themes are then
philosophy, the question, and the design. For example, verified through member checking. As themes emerge,
researchers using grounded theory build a case for sub- logical connections and relationships between the themes
stantive theory, explaining the phenomenon under study, are identified to form a whole picture. The results are
whereas a researcher in phenomenological studies is displayed in such a manner that the reader can under-
interested in expressing the meaning of the phenomenon stand and validate the conclusions that the researcher has
itself. drawn through the use of diagrams, tables, charts, direct
Regardless of the study’s research method, several quotations from the participants, and rich descriptions of
commonalities exist among methods used in qualitative the findings. In summary, qualitative data analysis
data analysis. For example, analysis is conducted along- involves much disparate data and transforms them into
side the data collection, and in most cases the two pro- a coherent whole or story to provide meaning about the
cesses are interrelated. Researchers become immersed in human experience.
the data; they listen over and over to the interviews, read

CRITIQUING CRITERIA

1. The method of data analysis 3. The steps of analysis should interpretation captures the
should be clearly stated. be listed for readers to phenomenon under study.
2. The strategy of data analysis follow. 5. The researcher should address
should be appropriate for the 4. The researcher should provide the credibility, auditability, and
methodology of the study. evidence that his or her fittingness of the data.

CRITICAL THINKING CHALLENGES KEY POINTS

■ How do researchers determine whether they • Qualitative data are text derived from transcripts of
have spent enough time with the data? interviews, narratives, documents, media such as
■ Is it important for the researcher to personally newspapers and movies, and field notes.
transcribe the interviews? • Computer software can simplify the storage and
retrieval of data.
■ Why do some researchers reread the literature as • Qualitative research data can be managed through
themes emerge from the data? the use of computers, but the researcher must
■ Often, data analysis takes place as data are interpret the data.
collected. How can analysis of the data change • Data analysis and data collection are parallel
the data collection? processes.
• Qualitative analysis is not a linear process; rather, it
■ Researchers validate their interpretation of the is a cyclical and iterative process.
data through a process of member checking. • The three discrete stages of data analysis are data
What happens if the participants indicate that reduction, data display, and conclusion drawing
the analysis does not reflect their experience? and verification.
 CHAPTER 15 Qualitative Data Analysis 345

Miles, M. B., Huberman, A. M. (1994). ualitative

• Data are organized into meaningful chunks of data
data analysis (2nd ed.). Thousand Oaks, CA: Sage.
through a clustering of related or similar data and
are labelled as themes. Patton, M. (2002). ualitative research evaluation
• Coding is the process of progressively marking, methods (3rd ed.). Thousand Oaks, CA: Sage.
sorting, resorting, and defining and redefining the Richards, L., Morse, J. M. (2007). Read me rst for a
collected data. user s guide to ualitative methods (2nd ed.).
• Data display involves the use of graphs, flowcharts, Thousand Oaks, CA: Sage.
matrices, or any other visual representation to Schreiber, R., MacDonald, M. (2010). Keeping vigil
assemble data and to allow for conclusion drawing. over the patient: A grounded theory of nurse anaes-
• Grounded theorists use the constant comparative thesia practice. ournal of Advanced Nursing, (3),
method, in which new data are compared with 552-561. doi: 10.1111/j.1365-2648.2009.05207.x
data previously analyzed.
Seidman, I. (1998). Intervie ing as ualitative research:
• Member checking is the process of sharing findings
with the participants in order to check whether the A guide for researchers in education and social
interpretation of the findings is accurate sciences (2nd ed.). New York: Teachers College
Press.
Seneviratne, C. C., Mather, C. M., Then, K. L.
(2009). nderstanding nursing on an acute stroke
REFERENCES unit: Perceptions of space, time and interprofessional
Bottorff, J. L., Johnson, J. L., Moffat, B., Grewal, J., practice. ournal of Advanced Nursing, (9),
Ratner, P., Kalaw, C. (2004). Adolescent construc- 1872-1881. doi: 10.1111/j.1365-2648.2009.05053.x
tions of nicotine addiction. Canadian ournal of Streubert, H. J., Carpenter, D. (2011). ualitative
Nursing Research, , 22-39. research in nursing: Advancing the humanistic
Cameron, C. (2003). The lived e perience of transfer imperative (5th ed.). Philadelphia: Wolters Kluwer.
students in a collaborative baccalaureate nursing Thorne, S., Armstrong, E., Harris, S. R., Hislop, T. G.,
program. npublished doctoral dissertation, niver- Kim-Sing, C., Oglov, V., et al. (2009). Patient
sity of Toronto. real-time and 12 month-retrospective perceptions of
Cameron, C. (2005). Experiences of transfer students in difficult communications in the cancer diagnostic
a collaborative baccalaureate nursing program. period. ualitative ealth Research, (10),
Community College Revie , (2), 22-44. 1383-1394. doi: 10.1177/1049732.309348382
Charlebois, S., Bouchard, L. (2007). The worst van Manen, M. (1997). Researching lived e perience.
experience : The experience of grandparents who London, ON: Althouse Press.
have a grandchild with cancer. Canadian ncological Wittich, W., Southall, K. (2008). Coping with
Nursing ournal, (1), 26-36. extended facedown positioning after macular hole
Denzin, N., Lincoln, Y. (2000). andboo of uali- surgery: A qualitative diary analysis. Nursing
tative research (2nd ed.). Thousand Oaks, CA: Sage. Research, (6), 436-443.
Ford-Gilboe, M., Wuest, J., Merritt-Gray, M. (2005). Wuest, J. (2001). Precarious ordering: Towards a formal
Strengthening capacity to limit intrusion: Theorizing theory of women’s caring. ealth Care for omen
family health promotion in the aftermath of woman International: Special volume, sing rounded
abuse. ualitative ealth Research, , 477-501. Theory to Study omen s ealth, (1-2), 167-193.
Glesne, C. (2006). Becoming ualitative researchers: An Wuest, J., Merritt-Gray, M., Berman, H.,
introduction (3rd ed.). Don Mills, ON: Longman. Ford-Gilboe, M. (2002). Illuminating social determi-
Glesne, C. (2011). Becoming ualitative researchers: An nants of women’s health using grounded theory.
introduction (4th ed.). Toronto: Pearson. ealth Care International, , 794-808.
Hollway, W., Jefferson, T. (2000). Doing ualitative
research differently: ree association, narrative and
the intervie method. Thousand Oaks, CA: Sage.
FOR FURTHER STUDY
Meadows, L. M., Dodendorf, D. M. (1999). Data Go to Evolve at http://evolve.elsevier.com/
management and interpretation using computers to Canada/LoBiondo/Research for Audio Glossary, how-to
assist. In B. Crabtree W. L. Miller (Eds.), Doing instructions for Writing Proposals for Funding, and
ualitative research (2nd ed., pp. 195-220). Thousand additional research articles for practice in reviewing
Oaks, CA: Sage. and critiquing.
C H A PTER 1 6

Quantitative Data Analysis

Susan Sullivan-Bolyai | Carol Bova | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Differentiate between descriptive and inferential statistics.
• State the purposes of descriptive statistics.
• Identify the levels of measurement in a research study.
• Describe a frequency distribution.
• List measures of central tendency and their use.
• List measures of variability and their use.
• Identify the purpose of inferential statistics.
• Distinguish between a parameter and a statistic.
• Explain the concept of probability as it applies to the analysis of sample data.
• Distinguish between type I and type II errors and their effects on a study’s outcome.
• Distinguish between parametric and nonparametric tests.
• List the commonly used statistical tests and their purposes.
• Critically analyze the statistics used in published research studies.

KEY TERMS
alpha level of significance (alpha nominal measurement
analysis of covariance level) nonparametric statistics
(ANCOVA) levels of measurement nonparametric tests of
analysis of variance (ANOVA) logistic regression (logit significance
chi-square (χ2) analysis) normal curve
confidence interval mean (M) null hypothesis
correlation measurement odds ratio
degree of freedom measures of central tendency ordinal measurement
descriptive statistics measures of variability P value
factor analysis median parameter
Fisher’s exact probability test modality parametric statistics
frequency distribution mode Pearson correlation coefficient
inferential statistics multiple analysis of variance (Pearson r)
interval measurement (MANOVA) percentile
kurtosis multiple regression population
346
 CHAPTER 16 Quantitative Data Analysis 347

post hoc analysis scientific hypothesis statistic

power semiquartile range (semi- systematic review
probability interquartile range) t statistic
range skew type I error
ratio measurement standard deviation type II error
sampling error standard error of the mean Z score
scatter plots

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

STATISTICS ARE USED EXTENSIVELY IN THE nursing the result of analysis with both descriptive and
and health care literature. Descriptive and infer- inferential statistics An example of what may be
ential statistics are described in the Methods found is as follows: A notable finding was the
section, the Results section, or both sections of high average age of SNLs senior nurse leaders
a research article. Before you become over- . . . current SNLs were very experienced individu-
whelmed by the complexity of the information, als inasmuch as over 80 . . . had at least 15 years
bear in mind that you do not need to be familiar of management experience (Wong, Laschinger,
with or to be able to calculate a large number of Cummings, Vincent, O’Connor, 2010, p. 127
complex statistical formulas to analyze data. An see Appendix B). The data in Table B-2 of Wong
understanding of which tests are used with which and colleagues’ article are known as descriptive
kind of design and which type of data is suffi- statistics, which are usually the first set of statisti-
cient. This basic understanding will help you to cal results in a report or published article.
appraise evidence from a research study, which is escri tive statistics are obtained through
essential for informing decisions you make in descriptive statistical techniques that reduce data
your practice. to manageable proportions by summarizing and
As a reader, you do not analyze the data your- organizing them. These techniques allow research-
self, but it is important to understand the research- ers to arrange data visually to display meaning
er’s challenge in analyzing the data. After and to help in understanding the sample charac-
carefully collecting data, the researcher is faced teristics and variables before the researchers
with the task of organizing and analyzing the engage in inferential data analyses. In some
individual pieces of information so that the studies, descriptive statistics may be the only
meaning of study results is clear. The researcher results sought from statistical analysis. Descrip-
must choose methods of organizing and analyz- tive statistical techniques include easures o
ing the raw data on the basis of the design, the central tendency which describe the average
type of data collected, and the hypothesis or ques- member of a sample, such as mode, median, and
tion that was tested. Statistical procedures are mean measures of variability, such as range and
used to organize and give meaning to the data. standard deviation and some correlation tech-
The Results section of a research article con- niques, such as a scatter lots which is a visual
tains the data generated from the testing of the representation of the strength and magnitude of
hypothesis or research questions. These data are the relationship between two variables.
348 PART FOUR Processes Related to Research

In contrast to descriptive statistics, inferential DESCRIPTIVE STATISTICS

statistics allow researchers to estimate how reli-
ably they can make predictions and generalize Levels of Measurement
findings on the basis of the data. In erential easure ent is the assignment of numbers to
statistics are statistical details that combine variables or events according to rules. Every vari-
mathematical processes and logic to test hypoth- able in a research study that is assigned a specific
eses about a population with the help of sample number must be similar to every other variable
data. Through the use of inferential statistics, assigned that number. For example, male partici-
researchers can draw conclusions that extend pants may be assigned the number 1 and female
beyond the immediate data of the study. An participants the number 2. The measurement level
example of inferential statistics was provided by is determined by the nature of the object or event
Sobieraj and colleagues (2009): The greatest being measured. Levels o easure ent—
predictors of child distress were age (β = −0.434, categorization of the precision with which an
t = −4.017, P < 0.01), with younger children event can be measured—from low to high are
being more distressed, and the presence of the nominal, ordinal, interval, and ratio. The levels of
father in the procedure room (β = −0.419, t = measurement help determine the type of statistics
−3.888, P < 0.01) (p. 75 see Appendix C). to be used in analyzing data. The higher the level
The purpose of this chapter is to demonstrate of measurement, the greater the exibility the
how researchers use descriptive and inferential researcher has in choosing statistical procedures.
statistics in nursing research studies so that you, Every attempt should be made to use the highest
as a reader, will be better able to determine the level of measurement possible so that the
appropriateness of the statistics used and to maximum amount of information will be obtained
interpret the strength and quality of the reported from the data, as highlighted in Table 16-1. The
findings, their clinical significance, and their Critical Thinking Decision Path on the facing
applicability to practice. Basic concepts and ter- page illustrates the relationship between levels of
minology common in evidence-informed practice measurement and appropriate choice of specific
publications are presented in Chapter 20. The descriptive statistics.
information in this chapter will help you begin to In no inal easure ent variables or events
make sense of the statistics used in research are classified into categories. The categories are
papers. mutually exclusive a variable or an event either

TABLE 16-1
LEVEL OF MEASUREMENT SUMMARY TABLE
MEASURES OF CENTRAL
MEASUREMENT DESCRIPTION TENDENCY MEASURES OF VARIABILITY

Nominal Classification Mode Modal percentage, range, frequency

distribution
Ordinal Relative rankings Mode, median Modal percentage, range, frequency, percentile,
semiquartile range, frequency distribution
Interval Rank ordering with equal Mode, median, mean Modal percentage, range, percentile,
intervals semiquartile range, standard deviation
Ratio Rank ordering with equal Mode, median, mean All
intervals and absolute zero
 CHAPTER 16 Quantitative Data Analysis 349

CRITICAL THINKING DECISION PATH

Descriptive Statistics

Is the study quantitative?

Yes No, qualitative

See Chapters 7 and 8

Nominal Ordinal Interval Ratio

measurement measurement measurement measurement

Frequency distribution Range Mean Mode

Range Percentile Standard deviation Median
Mode Rank order coefficients Mode Mean
of correlation Median Range
Mode Range Percentile
Median Percentile Standard deviation

has or does not have the characteristic of a par- A variable at the nominal level can also be
ticular category. The numbers assigned to each considered a dichotomous or a categorical
category are nothing more than labels such variable. A dichotomous nominal variable has
numbers do not indicate more or less of a char- only two true values, such as true/false or gender
acteristic. Nominal measurement can be used to (male/female). Nominal variables that are cate-
categorize a sample with regard to such informa- gorical still have mutually exclusive categories
tion as gender, hair colour, marital status, or reli- but have more than two true values, such as
gious affiliation. marital status (single, married, divorced, sepa-
Sobieraj and colleagues’ (2009) study regard- rated, or widowed). In both cases, the nominal
ing the effect of music on parental participation variables are mutually exclusive. The gender
during paediatric laceration repair included of the patient in the article by Wong and col-
examples of nominal measurement, including leagues (2010 see Appendix B) would be consid-
location of laceration, type of tissue repair, and ered a dichotomous nominal variable (male/
presence of a parent (see Appendix C). The female).
nominal level of measurement allows the least Ordinal easure ent reveals relative rank-
amount of mathematical manipulation. Most ings of variables or events. The numbers assigned
commonly, the frequency of each event is counted, to each category can be compared, and the
as is the percentage of the total that each category members of a higher ranked category can be said
represents. to have more of an attribute than members of a
350 PART FOUR Processes Related to Research

lower ranked category. The intervals between not absolute. For example, interval measurements
numbers on the scale are not necessarily equal, are used in measuring temperatures on the Fahr-
and zero is not absolute but arbitrary,. For enheit scale. The distances between degrees are
example, ordinal measurement is used to formu- equal, but the zero point is arbitrary and does not
late class rankings, in which one student can be represent the absence of temperature. Test scores
ranked higher or lower than another. However, also represent interval-level data. The differences
the actual grade-point averages of students may between test scores represent equal intervals, but
differ widely. Another example is ranking indi- a score of zero does not represent the total absence
viduals by their level of wellness and their ability of knowledge.
to carry out activities of daily living. Highest In many areas in the social sciences, including
level of education in the article by Wong and col- nursing, the classification of the level of measure-
leagues (2010 see Appendix B) is an example of ment of intelligence, aptitude, and personality
an ordinal variable. tests is controversial some researchers regard
The New York Heart Association’s classifica- these measurements as ordinal and others as
tion of cardiac failure adopted by the Canadian interval. You need to be aware of this controversy
Cardiovascular Society (Arnold, Liu, Demers, and to examine each study individually in terms
Dorian, Giannetti, Haddad, et al, 2006) consists of how the data are analyzed. Interval-level data
of four classifications. Classification I represents allow more manipulation of data, including the
little disease or interference with activities of addition and subtraction of numbers and the cal-
daily living, whereas classification IV represents culation of means. Because of this additional
severe disease and little ability to carry out the manipulation, many authorities argue for the
activities of daily living independently however, higher classification level. The Procedure Behav-
an individual in class IV cannot be said to be four iour Check List used by Sobieraj and colleagues
times sicker than an individual in class I. A similar (2009) is an example of ordinal measurements
scale based on an individual’s current health but used as an interval measurement (see Appen-
status is used to classify an individual’s risk for dix C).
adverse effects from anaesthesia. In ratio easure ent events or variables are
With ordinal-level data, the amount of math- ranked on scales with equal intervals and absolute
ematical manipulation possible is limited. In zeros. The number represents the actual amount
addition to what is possible with nominal-level of the property the object possesses. Ratio mea-
data, medians, percentiles, and rank-order coef- surement is the highest level of measurement but
ficients of correlation can be calculated. In most is usually achieved only in the physical sciences.
cases, ordinal variables in a scale are treated as Examples of ratio-level data are height, weight,
interval measurements when converted to numer- pulse, and blood pressure. All mathematical pro-
ical codes. For example, when patients are asked cedures can be performed with data from ratio
to rate their level of satisfaction with life as not scales. Therefore, the use of any statistical proce-
satisfied, satisfied, or very satisfied, their dure is possible as long as it is appropriate for the
responses are an ordinal measurement. When design of the study.
their ratings are treated numerically and coded as
1, 2, and 3, respectively, their responses are an Helpful Hint
interval measurement. Descriptive statistics assist in summarizing the
In interval easure ent events or variables data. The descriptive statistics calculated must be appro-
are ranked on a scale with equal intervals between priate for both the purpose of the study and the level
of measurement.
the numbers. The zero point remains arbitrary and
 CHAPTER 16 Quantitative Data Analysis 351

results are reported in two ways. The columns on

Frequency Distribution the left give the raw data tally and the frequency
One of the most basic ways of organizing data is for each grade, whereas the columns on the right
in a frequency distribution. In a re uency dis give the grouped data tally and grouped frequen-
tribution the number of times each event occurs cies. In research studies, the results are grouped
is counted, or the data are grouped and the fre- rather than reported individually for each par-
quency of each group is reported. For example, ticipant (Sobieraj, Bhatt, LeMay, Rennick,
an instructor reporting the results of an examina- Johnston, 2009 see Appendix C).
tion could report the number of students receiving When data are grouped, the researcher needs
each individual grade or could group the grades to define the size of the group or the interval
in ranges and report the number of students who width so that no score is categorized into two
received each group of grades. When reviewing groups and all groups are mutually exclusive. The
a frequency distribution, symmetry and kurtosis groupings of the data in Table 16-2 prevent
are noted. A distribution can be symmetrical overlap each score is categorized into only one
(shaped like a bell) or asymmetrical, where most group. If the grouping had been 70 to 80 and 80
of the information is to one side, either to the left to 90, scores of 80 would have been categorized
or the right. urtosis is the peakedness of the into two categories. The grouping should allow
distribution. Table 16-2 shows the results of an for a precise presentation of the data without
examination given to a class of 51 students. The serious loss of information. Very large interval

TABLE 16-2
FREQUENCY DISTRIBUTION
INDIVIDUAL GROUP
SCORE TALLY FREQUENCY SCORE TALLY FREQUENCY

90 | 1 >89 | 1
88 | 1 80–89 15
86 | 1
84 | 6
82 || 2
80 5
78 5 70–79 ||| 23
76 | 1
74 || 7
72 |||| 9
70 | 1
68 ||| 3 60–69 10
66 || 2
64 |||| 4
62 | 1
60 0
58 | 1 <59 || 2
56 0
54 | 1
52 0
50 0
Total 51 Total 51
Mean, 73.1; standard deviation, +12.1; median, 74; mode, 72; range, 36 (54–90).
352 PART FOUR Processes Related to Research

A B

Frequency

Frequency
Scores Scores
FIGURE 16-1 Frequency distributions. A, Histogram. B, Frequency polygon.

widths lead to loss of data information and may What is the average temperature of patients on
obscure patterns in the data. If the test scores in a unit These measures yield a single number
Table 16-2 had been grouped as 40 to 69 and 70 that describes the middle of the group and sum-
to 99, the pattern of the scores would have been marizes the members of a sample. In statistics,
obscured. the three measures of central tendency are the
Information about frequency distributions may mode, the median, and the mean. Depending on
be presented in the form of a table, such as Table the distribution, these measures may not all give
16-2, or in the form of a graph. Figure 16-1 illus- the same answer to the question What is the
trates the most common graph forms: the histo- average Each measure of central tendency has
gram and the frequency polygon. These two a specific use and is most appropriate for specific
methods are similar in that in both, scores or kinds of measurement and types of distributions.
percentages of occurrence are plotted against fre-
quency. The greater the number of points plotted, MODE. The ode is the most frequent score or
the smoother is the resulting graph. The shape of result and can be obtained by inspection of the
the resulting graph allows for observations that frequency distribution table or graph. Note that a
further describe the data. For example, in their sample distribution can have more than one
study of parents’ knowledge about the use of mode. The number of modes, or peaks, contained
child safety systems, Snowdon and colleagues in a distribution is called the odality of the
(2006) used histograms to illustrate the rate at distribution. The mode is the type of descriptive
which parents transitioned their infant from a statistic most appropriately used with nominal-
rear-facing seat to a forward-facing seat and the level data but can be used with all levels of mea-
timing of this transition (Figure 16-2). surement (see Table 16-1). The mode cannot be
used for any subsequent calculations and is unsta-
ble in other words, the mode can uctuate widely
Measures of Central Tendency from sample to sample from the same population.
Measures of central tendency answer questions A change in just one score in Table 16-2 would
such as What does the average nurse think and change the mode from 72.
 CHAPTER 16 Quantitative Data Analysis 353

Percentage of children using a seat belt

Percentage of children meeting the 36-kg weight
requirement for safely using a seat belt
A B
100 100 100 100
81.0
80 80

PERCENTAGE
PERCENTAGE

61.0 60
60
49.6
40 40
31.0
20 16.0 20
7.1
1.7
0 0
3 4 5 6 7 8 9 10 11 0–3 4–8 9–12
AGE (YEARS) AGE (MONTHS)
FIGURE 16-2 Histograms. A, Premature use of seat belts in children aged 3 to 11 years. B, Use
of rear-facing safety seats in infants up to 12 months (N = 296).
From Snowdon, A. W., Polgar, J., Patrick, L., & Stamler, L. (2006). Parents’ knowledge about
use of child safety systems. Canadian Journal of Nursing Research, 38(2), 107–108.

MEDIAN. The edian is the middle score: of the Because the mean is affected by every score, it is
other scores, 50 are higher and 50 are lower. affected by extreme scores however, the larger
The median is not sensitive to extremes in high the sample size, the less effect a single extreme
and low scores thus, it is a more accurate estima- score will have on the mean. For normally dis-
tor of central tendency in non-normal distribu- tributed populations, the mean is an appropriate
tions. In the series of scores in Table 16-2, the measure of central tendency and is generally con-
twenty-sixth score is always the median, regard- sidered the single best point for summarizing
less of how much the high and low scores change. data.
The median is best used when the data are skewed
(see the Normal Distribution section) and the
researcher is interested in the typical score. For Helpful Hint
example, if age is a variable, and if a wide range Of the three measures of central tendency, the
with extreme scores may affect the mean, it would mean is the most stable, the least affected by extremes,
and the most useful for other calculations. The mean
be appropriate to also report the median. The can be calculated only with interval- and ratio-level
median is easy to find either by inspection or by data.
calculation and can be used with ordinal or higher
data, as shown in Table 16-1.
Profetto-McGrath and associates (2010) used
MEAN. The ean is the arithmetical average a table and narrative to describe the sample char-
of all scores and is used with interval- or ratio- acteristics in their study examining the approaches
level data (see Table 16-1). Most statistical tests used by clinical nurse specialists to select and use
of significance refer to the mean, the most widely evidence in their daily practice.
used measure of central tendency, which is The summary statistics in Table 16-3 may also
referred to in general conversations as the average. be reported in narrative form.
354 PART FOUR Processes Related to Research

TABLE 16-3
SUMMARY STATISTICS FOR STUDY BY
PROFETTO AND ASSOCIATES
NUMBER OF
VARIABLE CHARACTERISTIC PARTICIPANTS (%)

Sex Male 2 (2.1)

Female 2 (97.9)
Age Mean/SD: 48.53 ± 7.643
Range: 30–66 years
Highest level Certificate 1 (1.1)
of nursing Diploma 22 (23.4)
education Bachelor’s 37 (39.4)
Master’s 33 (35.0) 68%
Doctorate 1 (1.1) 95%
Years as a Mean: 25.12 ± 8.521
registered Range: 2–45 years
99.7%
nurse
Adapted from Profetto-McGrath, J., Negrin, K., Hugo, K., & Bulmer FIGURE 16-3 The normal distribution and associated stan-
Smith, K. (2010) Clinical nurse specialists’ approaches in selecting and dard deviations.
using evidence to improve practice. Worldviews on Evidence-Based
Nursing, 7(1), 36-50, Table 1.

[The] sample was predominantly female (97.9%) with addition, if the means of a large number of
a mean age of 48.53 years (SD [standard deviation] = samples of the same interval- or ratio-level data
7.643). and an average experience of 10.24 years
(SD = 7.282) . . . On average, [clinical nurse specialists]
are calculated and plotted on a graph, that curve
reported 38.83% of their work responsibilities were also approximates the normal curve. This ten-
devoted to clinical practice, 22.11% to education dency of the means to approximate the normal
and training, 16.6% to consultation, and 6.32% to curve is termed the sampling distribution of the
research. (Profetto-McGrath, Negrin, Hugo, & Bulmer means The mean of the sampling distribution of
Smith, 2010, pp. 39-40)
the means is the mean of the population.
Of the measures of central tendency, the mean In visual representations of statistics, the nor
is the most stable and the median the most typical. al curve is unimodal and symmetrical about
If the distribution of a sample is symmetrical and the mean. The mean, median, and mode are equal.
unimodal, the mean, median, and mode coincide. An additional characteristic of the normal curve
is that a fixed percentage of the scores is located
Helpful Hint within a given distance of the mean. As shown in
Measures of central tendency are descriptive sta- Figure 16-3, about 68 of the scores or means
tistics that describe the characteristics of a sample.
are within one standard deviation of the mean,
95 within two standard deviations of the mean,
Normal Distribution and 99.7 within three standard deviations of the
The theoretical concept of normal distribution is mean.
based on the observation that data from repeated
interval or ratio measurements will gather at a SKEWNESS. S e is a measure of the asymmetry
midpoint in a distribution, approximating the of a set of scores. Not all samples of data approxi-
normal curve illustrated in Figure 16-3. In mate the normal curve. Some samples are
 CHAPTER 16 Quantitative Data Analysis 355

nonsymmetrical, and the peak is off centre. For

example, worldwide individual income has a Interpreting Measures of Variability
positive skew: Most individuals have incomes in Variability or dispersion is concerned with the
the low-to-moderate range and few in the upper spread of data. easures o variability—
range. In a positive skew, the peak of the distribu- statistical procedures that describe the level of
tion curve would be to the left of a normal curve, dispersion in sample data—answer questions
and the mean is to the right of the median. In such as Is the sample homogeneous or heteroge-
contrast, age at death in Canada has a negative neous and Is the sample similar or different
skew because most deaths occur at older ages. In If a researcher measures oral temperatures in two
a negative skew, the peak of the distribution curve samples, one sample drawn from a healthy popu-
would be to the right of a normal curve, and the lation and one sample from a hospitalized popula-
mean is to the left of the median. Figure 16-4 tion, it is possible that the two samples will
illustrates positive and negative skew. In each have the same mean. However, a wider range
diagram, the peak is off centre, and one tail of of temperatures is more likely to be found in
the curve is longer. the hospitalized sample than in the healthy
If the distribution is skewed, the mean will be sample. Measures of variability are used to
pulled in the direction of the long tail of the dis- describe these differences in the dispersion of
tribution. With a skewed distribution, all three data. As with measures of central tendency, the
statistics should be reported. For example, various measures of variability are appropriate
national income in Canada is skewed. The mean to specific kinds of measurement and types of
wage differs from the median wage because the distributions.
high salaries greatly outnumber the low salaries.

Helpful Hint
Evidence-Informed Practice Tip Descriptive statistics related to variability enable
The descriptive statistics for a sample indicate you to evaluate the homogeneity or heterogeneity of a
whether the sample data are skewed. sample.

A B

Positive Negative
FIGURE 16-4 Positive and negative skew. A, Positive skew. B, Negative skew.
356 PART FOUR Processes Related to Research

RANGE. The range is the simplest but most unsta- (see Figure 16-3). The standard deviation is a
ble measure of variability. ange is the distance measure of average deviation of the scores from
between the highest and lowest scores. A change the mean and, as such, should always be reported
in either of these two scores would change the with the mean. The standard deviation accounts
range. The range should always be reported with for all scores and can be used to interpret indi-
other measures of variability. For example, Sch- vidual scores. For the examination in Table 16-2,
neider and colleagues (2011), who examined dif- the mean was 73.1 and the standard deviation
ferences in psychosocial outcomes between male was 12.1 thus, a student should know that 68
and female caregivers of children with life- of the grades were between 85.1 and 61. If the
limiting illnesses, found that the range of ages student received a grade of 88, he or she would
among female caregivers was 22.99 to 68.38 know that this grade was better than those of most
years, whereas the range of ages among the male of the class, whereas a grade of 58 would indicate
caregivers was 26.98 to 57.02 years (Table 16-4). that the student did not do as well as most of the
Thus, the age range was 45 years among the class. Table C-2 in Appendix C from the study by
female caregivers and 30 years among the male Sobieraj and colleagues (2009), reports the mean
caregivers. Range affects the standard deviation, and standard deviation of the study variables’ dis-
as discussed later. The range in Table 16-4 could tress score for the both the control and interven-
easily change with an increase or decrease in the tion groups. As illustrated in this table, the mean
high scores or the low scores with a different score for the control group was 33.1 (SD = 29),
sample. whereas the mean score for the intervention group
was 28.6 (SD = 26). This means that 68 of the
SEMIQUARTILE RANGE. The se i uartile range control group scored between 2.1 and 62.1 on this
se i inter uartile range is the range of the measure and 68 of the intervention group scored
middle 50 of the scores. It is more stable between 2.6 and 54.6 on this measure. This table
than the overall range because it is less likely allows the reader to inspect the data and see the
to be changed by a single extreme score. The variation in the data.
semiquartile range lies between the upper and The standard deviation is used in the calcula-
lower quartiles the upper quartile consists of the tion of many inferential statistics. One limitation
top 25 of scores, and the lower quartile consists of the standard deviation is that it is expressed in
of the lowest 25 of the scores. In Table 16-2, terms of the units used in the measurement and
the middle 50 of the scores are between 68 and cannot be used to compare means that have dif-
78, and the semiquartile range is 10. ferent units. If researchers were interested in the
relationship between height measured in centime-
PERCENTILE. A ercentile represents the percent- tres and weight measured in kilograms, it would
age of scores that a given score exceeds. The be necessary to convert the height and weight
median is the 50th percentile, and in Table 16-2, measurements to standard units, or scores. The
it is a score of 74. A score in the 90th percentile score is used to compare measurements in stan-
is exceeded by only 10 of the scores. The zero dard units. Each of the scores is converted to a
percentile and the 100th percentile are usually not score, and then the scores are used to examine
used. the relative distance of the scores from the mean.
A score of 1.5 means that the observation is 1.5
STANDARD DEVIATION. The standard deviation is standard deviations above the mean, whereas a
the most frequently used measure of variability score of −2 means that the observation is 2 stan-
and is based on the concept of the normal curve dard deviations below the mean. By using
 TABLE 16-4
DEMOGRAPHIC INFORMATION OF CAREGIVER BY GENDER
TOTAL (N = 273) WOMEN (N = 224) MEN (N = 49)
VARIABLE M SD RANGE M SD RANGE M SD RANGE TEST STATISTIC

AGE IN YEARS 41.74 7.61 22.99–68.38 41.19 7.62 22.99–68.38 44.29 7.08 26.98–57.02 t = − 2.61, P = .009
DIFFICULTY IN MANAGING COSTS 5.71 2.56 1–10 5.90 2.56 1–10 4.92 2.42 1–10 t = 2.42, P = .016
HOURS PER WEEK SPENT PROVIDING CARE 62.16 44.72 0–126 68.00 43.88 0–126 37.64 39.97 1.50–126 t = 4.29, P = .000
IMPORTANCE OF RELIGION* 3.04 1.03 1–4 3.10 1.01 1–4 2.76 1.09 1–4 t = 2.16, P = .002

MARITAL STATUS χ2 = 12.37, P = .015

Married or lives as married 219 80.2 172 76.8 47 95.9
Widowed 2 0.7 1 0.4 1 2.0
Divorced/separated 34 12.5 33 14.7 1 2.0
Never married 13 4.8 13 5.8 0 0
Other 5 1.8 5 2.2 0 0

HIGHEST EDUCATION LEVEL χ2 = 3.83, P = .429

Elementary school completed 13 4.8 9 4.1 4 8.2
High school completed 98 36.2 85 38.3 13 26.5
College completed 39 14.4 30 13.3 9 18.4
University degree completed 64 23.6 51 23.0 13 26.5
Postgraduate degree completed 57 21.0 47 21.2 10 20.4

SIZE OF COMMUNITY χ2 = 0.61, P = .737

Metropolitan area/large city 120 44.0 97 43.3 23 46.9
Medium city/small city 91 33.3 77 34.4 14 28.6
Smaller communities 62 22.7 50 22.3 13 24.5

CURRENT EMPLOYMENT STATUS χ2 = 59.95, P = .000

Full-time 114 41.8 70 31.3 44 89.8
Part-time 38 13.9 37 16.5 1 2.0
Paid leave 5 1.8 5 2.2 0 0
Unpaid leave 3 1.1 3 1.3 0 0
Self-employed 16 5.9 13 5.8 3 6.1
Not employed 78 28.6 78 34.8 0 0
Other 19 7.0 18 8.0 1 2.0
Continued
CHAPTER 16 Quantitative Data Analysis
357
 358

TABLE 16-4
DEMOGRAPHIC INFORMATION OF CAREGIVER BY GENDER—cont’d
TOTAL (N = 273) WOMEN (N = 224) MEN (N = 49)
VARIABLE M SD RANGE M SD RANGE M SD RANGE TEST STATISTIC
2
CHANGE IN EMPLOYMENT STATUS χ = 29.05, P = .000
No 136 50.9 94 43.1 42 85.7
Yes 131 49.1 124 56.9 7 14.3

WORK ALLOWS TIME OFF χ2 = 13.11, P = .001

No 10 5.2 5 3.4 5 10.0
Yes completely 110 57.0 93 64.1 17 35.4
PART FOUR Processes Related to Research

Only partially 73 37.8 47 32.4 26 54.2

INCOME CHANGED: CHILD’S ILLNESS χ2 = 5.52, P = 0.19

No 96 35.3 72 32.1 24 50
Yes 176 64.7 152 67.9 24 50

AVERAGE HOUSEHOLD INCOME PER YEAR ($) U = 3,515.50, P = .001

<19,999 21 8.0 20 9.2 1 22
20,000–39,999 39 14.8 37 17.0 2 4.3
40,000–59,999 39 14.8 33 15.1 6 031F,
60,000–79,999 47 17.8 39 17.9 8 (]IA)
80,000–99,999 32 12.1 27 12.4 5 COSI)
100,000–119,999 38 14.4 27 12.4 11 219
120,000–139,999 18 6.8 11 5.0 7 352
140,000–199,999 22 8.3 16 7.3 6 03.0
≤200,000 8 3.0 8 3.7 0 C

CURRENT INCOME MEETS NEEDS U = 3,515.50, P = .004

Totally inadequate 10 3.7 10 4.5 0 0
Not very well 30 11.1 29 13.0 1 2.1
Adapted from Schneider, M., Steele, R., Cadell, S., & Hemsworth, D. (2011). Differences on psychosocial outcomes between male and female caregivers of children with life-limiting illnesses.
Journal of Pediatric Nursing, 26(3), 186–199. doi:10.1016/j.pedn.2010.01.007, p. 6, Copyright 2011, with permission from Elsevier.
SD, standard deviation.
*Range: 1 = not important at all to 4 = very important.
 CHAPTER 16 Quantitative Data Analysis 359

TABLE 16-5
TESTS OF DIFFERENCES BETWEEN MEANS
TWO GROUPS
MORE THAN
LEVEL OF MEASUREMENT ONE GROUP RELATED INDEPENDENT TWO GROUPS

NONPARAMETRIC
Nominal Chi-square Chi-square Chi-square Chi-square
Fisher’s exact probability test
Ordinal Kolmogorov-Smirnov test Sign test Chi-square Chi-square
Wilcoxon matched-pairs test

PARAMETRIC
Interval or ratio Correlated t test Correlated t test Independent t test ANOVA
ANOVA (repeated measures)
ANOVA ANCOVA
MANOVA
ANCOVA, Analysis of covariance; ANOVA, analysis of variance; MANOVA, multivariate analysis of variance.

scores, a researcher can compare results from TABLE 16-6

scales that use different measurement units, such TESTS OF ASSOCIATION
as height and weight. LEVEL OF MORE THAN TWO
MEASUREMENT TWO VARIABLES VARIABLES
Helpful Hint NONPARAMETRIC
Many measures of variability exist. The standard Nominal Phi coefficient Contingency
deviation is the most stable and useful because it pro- Point-biserial coefficient
vides a visual image of how the scores are dispersed correlation
around the mean. Ordinal Kendall’s tau Discriminant
Spearman’s rho function analysis

INFERENTIAL STATISTICS PARAMETRIC

Interval or ratio Pearson r Multiple regression
Inferential statistics combine mathematical pro- Path analysis
cesses with logic and allow researchers to test Canonical correlation
hypotheses about a population by using data
obtained from probability samples. Statistical the disease. A mean score of 65 for these partici-
inference is generally used for two purposes: to pants represents the sample statistic. If the
estimate the probability that statistics found in the researcher were able to study every participant
sample accurately re ect the population parame- with chronic lung disease, he or she also could
ter and to test hypotheses about a population. calculate an average knowledge score, and that
In the first purpose, a ara eter is a charac- score would be the parameter for the population.
teristic of a o ulation—a well-defined set that Researchers are rarely able to study an entire
has certain specified properties—whereas a population, but inferential statistics provide evi-
statistic is a characteristic of a sample Statistics dence that allow them to make statements about
are used to estimate population parameters. the larger population from studying the sample.
Suppose that a researcher randomly selects 100 Both parametric and nonparametric inferential
people with chronic lung disease and use an tests can be used in data analyses (see Tables 16-5
interval-level scale to study their knowledge of and 16-6). Parametric statistical models are based
360 PART FOUR Processes Related to Research

on assumptions about the distributions of sample the assumption that the sample was drawn with a
values and parameters thus, in these models, known probability. Second, the scale had to
means and variances are used to test significance. re ect the interval level of measurement. The
Nonparametric tests are used when populations mathematical operations involved in inferential
have non-normal distributions or when research- statistics require this level of measurement. Note
ers wish to explore associations among variables. that researchers who use nonprobability methods
In these tests, no assumptions about the distribu- of sampling also use inferential statistics. To
tion of the data are made. compensate for the use of nonprobability sam-
The example of the study of patients with lung pling methods, researchers use techniques such as
disease alludes to two important qualifications of sample size estimation through power analysis.
how a study must be conducted so that inferential The following two Critical Thinking Decision
statistics may be used. First, the sample was Paths provide algorithms that re ect inferential
selected randomly: that is, through the use of statistics and that researchers use for statistical
probability methods (see Chapter 12). Because decision making.
you are already familiar with the advantages of
probability sampling, you know that in order to Evidence-Informed Practice Tip
make generalizations about a population from a Try to determine whether the statistical test
sample, that sample must be representative. All chosen was appropriate for the design, the type of data
collected, and the level of measurement.
procedures for inferential statistics are based on

CRITICAL THINKING DECISION PATH

Inferential Statistics—Difference Questions

Is the research question asking for a difference?

Yes No, asking for a relationship

One group or more than one group? Go to the other algorithm

Two groups One group

Interval measure? Nominal or ordinal measure? Interval measure? Nominal or ordinal measure?

t-test Chi-square Correlated t Chi-square

ANOVA Sign test ANOVA Kolmogorov–Smirnov
Signed rank
Mann–Whitney U

ANOVA, Analysis of variance.

 CHAPTER 16 Quantitative Data Analysis 361

CRITICAL THINKING DECISION PATH

Inferential Statistics—Relationship Questions

Is the research question asking for a relationship?

Yes No, asking for a difference

Two variables or more than two variables? Go to the other algorithm

Two variables More than two variables

Interval measure? Nominal or ordinal measure? Interval measure? Nominal or ordinal measure?

Pearson product Phi coefficient Multiple regression Contingency coefficient

moment correlation Point-biserial Path analysis Discriminant function
Kendall’s tau Canonical correlation analysis
Spearman’s rho Logistic regression

outcome in the experimental group in comparison

Hypothesis Testing with that in the control group. The null hy oth
The second and most commonly used purpose esis (also called the statistical hypothesis or ),
of inferential statistics is hypothesis testing. Sta- which is the hypothesis that actually can be tested
tistical hypothesis testing allows researchers to by statistical methods, would be that no differ-
make objective decisions about the outcome of ence exists between the groups. In inferential sta-
their study and to answer questions such as How tistics, the null hypothesis is used to test the
much of this effect is a result of chance How validity of a scientific hypothesis in sample data.
strongly are these two variables associated with According to the null hypothesis, no relationship
each other and What is the effect of the exists between the variables. and any observed
intervention relationship or difference is merely a function of
The procedures used to make inferences are chance uctuations in sampling.
based on principles of negative inference. For The concept of the null hypothesis is often
example, to study the effect of a new educational confusing. An example may help clarify this
program for patients with chronic lung disease, concept. Sobieraj and colleagues (2009) used
the researcher would actually have two hypoth- music during simple laceration repair to promote
eses: the scientific hypothesis and the null hypoth- parent-led distraction they were interested in
esis. The research or scienti c hy othesis (H1) is determining whether this protocol would affect
what the researcher believes the outcome of the children’s distress during the procedure (see
study will be. In this example, the scientific Appendix C). On the basis of this hypothesis,
hypothesis would be that the educational inter- Sobieraj and colleagues wanted to determine
vention would have a marked effect on the whether the differences found in the dependent
362 PART FOUR Processes Related to Research

variables differed significantly between the inter-

vention group and the control group. The authors Probability
had to use the null hypothesis—that no difference The researcher can never prove the scientific
would exist between the intervention and control hypothesis but can show support for it by reject-
groups—to test the scientific hypothesis. They ing the null hypothesis: that is, by showing that
found no difference in distress scores, regardless the null hypothesis has a high probability of being
of parental participation. In other words, the dif- incorrect. The theory underlying all of the proce-
ferences between the control and intervention dures discussed in this chapter is probability
group scores were not large enough to conclude theory. Probability is a concept that people talk
that they were unlikely to be caused by chance. about all the time, such as the chance of rain, but
Thus, the null hypothesis was not rejected. have a difficult time defining it. The robability
On the other hand, Schneider and colleagues of an event is the event’s long-run relative fre-
(2011) were examining psychosocial outcomes of quency in repeated trials under similar condi-
meaning in caregiving, self-esteem, optimism, tions. In other words, the statistician does not
burden, depression, spirituality, and posttrau- think of the probability of obtaining a single
matic stress between male and female caregivers result from a single study but rather of the chances
of children with life-limiting illnesses. One of of obtaining the same result from an idealized
their null hypotheses was that there was no dif- study that can be carried out many times under
ference in meaningfulness of caregiving scores identical conditions. The notion of repeated trials
between women and men. They reported that allows researchers to use probability to test
women had higher average scores than did men hypotheses.
for meaningfulness in caregiving (t = 2.66, df = Statistical probability is based on the concept
270, P = .008). In other words, the null hypothesis of sampling error. The use of inferential statistics
was rejected, and the differences between the is based on random sampling. However, even
men and women were large enough to conclude when samples are randomly selected, the possi-
that they were unlikely to be caused by change. bility of errors in sampling always exists. There-
See information on the interpretation of P values fore, the characteristics of any given sample
in the later Level of Significance section. may be different from those of the entire
All statistical hypothesis testing is a process of population.
disproof or rejection. It is impossible to prove that Suppose that a large group of patients with
a scientific hypothesis is true, but it is possible to decubitus ulcers is available for study and that
demonstrate that the null hypothesis has a high researchers wish to learn the average length of
probability of being incorrect. To reject the null time for such ulcers to heal with the usual nursing
hypothesis, therefore, is to show support for the care. If the researchers studied the entire popula-
scientific hypothesis, which is the desired outcome tion, they might obtain an average healing time
of most reports of inferential statistics. of 50 days, with a standard deviation of 10 days.
Now, suppose that the researchers did not have
the money necessary to study all the patients but
Helpful Hint wished to conduct several consecutive studies of
Remember that most samples used in clinical this condition. For this study, the researchers first
research are samples of convenience, but most research-
ers use inferential statistics. Although such use violates select a sample of 25 patients, calculate the mean
one of the assumptions of such tests, the tests are and standard deviation, and then select the next
robust enough to not seriously affect the results unless sample. If this process is repeated many times in
the data are skewed in unknown ways.
different samples, a different mean for each
 CHAPTER 16 Quantitative Data Analysis 363

sample would probably result. For example, the the more accurate are those means as estimates
researchers might find that one sample’s mean of the population value.
might be 50.5 days, the next 47.5, and the next Although researchers rarely construct sam-
62.5. The tendency for statistics to uctuate from pling distributions, standard error can be esti-
one sample to another is known as sa ling mated because it bears a systematic relationship
error to the sample standard deviation and the size of
Sampling distributions are theoretical. In prac- the sample. Thus, increasing the size of the
tice, researchers do not routinely draw consecu- sample will increase the accuracy of estimates of
tive samples from the same population they population parameters. It is intuitive that an
usually compute statistics and make inferences on increase in the size of a sample will decrease the
the basis of data from one sample. However, the likelihood that one outlying score will dramati-
knowledge of the properties of the sampling cally affect the sample mean (see Chapter 12).
distribution—if these repeated samples are hypo- The other reason that the sampling distribution is
thetically obtained—enables the researcher to so important is that all statistics have sampling
draw a conclusion on the basis of data from one distributions. Researchers consult these distribu-
sample. Such a conclusion is possible because the tions when making determinations about reject-
sampling distribution of the means has certain ing the null hypothesis.
known properties.
The sampling distribution of the means is Evidence-Informed Practice Tip
shaped like a normal curve, and the mean of the Remember that the strength and quality of evi-
sampling distribution is the mean of the popula- dence are enhanced by repeated trials that have consis-
tion. As discussed in the earlier Normal Distri- tent findings, thereby increasing the generalizability of
bution section, because the sampling distribution the findings and applicability to clinical practice.
of the means is normal, several other important
characteristics are revealed. When scores are nor-
mally distributed, 68 of them are between +1 Type I and Type II Errors
standard deviation and −1 standard deviation, or The researcher’s decision to accept or fail to
the probability is 68 per 100 that any one ran- accept (reject) the null hypothesis is based on a
domly drawn sample mean is within the range of consideration of the probability that the observed
values between +1 standard deviation and −1 differences are a result of chance alone. Because
standard deviation (see Figure 16-3). In the data on the entire population are not available,
example described earlier, if only one sample the researcher cannot atly assert that the null
were selected, the chance of finding a sample hypothesis is or is not true. Thus, statistical infer-
mean between 40 and 60 would be 68 . The ence is always based on incomplete information
standard deviation of a theoretical distribution of about a population, and errors can occur when
sample means is called the standard error o the such inferences are made. These errors are clas-
ean The word error is used because the various sified as type I and type II.
means that make up the distribution contain an A ty e I error is the researcher’s incorrect
error in their estimates of the population mean. decision to reject the null hypothesis (Kline,
The error is considered to be standard because it 2005) that is, the researcher has found that results
implies the magnitude of the average error, just are statistically significant, but in fact they are
as a standard deviation implies the average varia- not, and has accepted the alternate hypothesis. If,
tion from one mean. The smaller the standard however, the researcher had found that the groups
error, the less variable are the sample means and did not differ perhaps because only a few patients
364 PART FOUR Processes Related to Research

had been studied or the design of the study was validity of the instruments used. For example, if
poor for determining differences, a type II error the instruments did not accurately and precisely
might occur. In a ty e II error—also known as measure the intervention variables, the conclu-
beta (β)—the results from the sample data lead to sion could be that the intervention made a differ-
the failure to reject the null hypothesis when it is ence, but, in reality, it did not. It is critical to
actually false that is, no statistically significant consider the reliability and validity of all of the
differences between groups were found but there measurement instruments reported (see Chapter
are indeed real differences. Po er is the condi- 14). In a practice discipline, type I errors usually
tional prior probability that the researcher will are considered more serious because if a
decide correctly to reject the null hypothesis researcher declares that differences exist where
when it is actually false (Kline, 2005). A standard none are present, then patient care can poten-
value of power of .8 is used to conduct power tially be affected adversely. Type II errors
analyses in studies to determine sample size (accepting the null hypothesis when it is false)
before the study begins. Power and beta are com- may occur if the sample in the study is too small,
plementary and sum to 1.00. When power is thereby limiting the opportunity to measure the
increased, type II error is decreased, and vice treatment effect, a true difference between two
versa. groups. A larger sample size improves the ability
In Schneider and colleagues’ (2011) study of to detect the treatment effect: that is, the differ-
male and female caregivers, one null hypothesis ences between two groups. If no significant dif-
of the study was that no differences in the psy- ference is found between two groups with a large
chological outcome of depression would exist sample, this finding provides stronger evidence
between the genders. Schneider and colleagues (than with a small sample) not to reject the null
reported a significant difference in rates of depres- hypothesis.
sion (t = 3.27, df = 271, P = .001). If the differ-
ences found were truly a function of chance Level of Significance
(because this group of participants was unusual The researcher does not know when an error in
in some way) and if the number of participants statistical decision making has occurred. It is pos-
was too small, a type I error would occur. sible to know only that the null hypothesis is
The relationship of the two types of errors is indeed true or false if data from the total popula-
shown in Figure 16-5. When you critique a study tion are available. However, the researcher can
to determine whether a type I error has occurred control the risk of making type I errors by setting
(rejecting the null hypothesis when it is actually the level of significance before the study begins
true), you should consider the reliability and (a priori). Slakter and colleagues (1991) explained

REALITY
Conclusion of test Null hypothesis Null hypothesis is
of significance is true not true
Not statistically Correct conclusion Type II error
significant
Statistically Type I error Correct conclusion
significant

FIGURE 16-5 Outcome of statistical decision making.

 CHAPTER 16 Quantitative Data Analysis 365

in detail the importance of setting the level of statistically significant at that alpha level. For
significance before the study is conducted. The example, if the alpha is set at .05 and the P value
level o signi cance al ha level is the probabil- is found to be .04, then the results are considered
ity of making a type I error: in other words, statistically significant.
the conditional probability of rejecting the null Whatever level of significance is set, the
hypothesis when it is actually true. Al ha or the researcher either rejects or accepts the null
level of significance, is considered an a priori hypothesis when comparing the statistical results
probability because it is set before the data with the preset alpha. For example, in Schneider
are collected, and it is a conditional probability and colleagues’ (2011) study, the null hypothesis
because the null hypothesis is assumed to be true regarding participants’ psychological outcomes
(Kline, 2005). The minimum level of significance was rejected because the variables of the hypoth-
acceptable for nursing research is .05. If the esis were significant at the .05 level or lower in
researcher sets alpha at .05, the researcher is other words, the P values were less than alpha.
willing to accept the fact that if the study were Sobieraj and colleagues (2009), however, failed
done 100 times, the decision to reject the null to reject the null hypothesis and found that the
hypothesis would be wrong in 5 of those 100 intervention group’s distress scores were not sig-
trials, only if the null hypothesis is true. nificantly improved over those of the control
Sometimes the researcher wants to have a group because the P values were greater than
smaller risk of rejecting a true null hypothesis in alpha (see Appendix C).
that case, the level of significance may be set at Perhaps you are thinking that researchers
.01. In this case, the researcher is willing to make should always use the lowest alpha level possible
the wrong decision only once in 100 trials. The because it makes sense that they would like to
decision as to how strictly the alpha level should keep the risk of both types of errors at a minimum.
be set depends on how important it is not to make nfortunately, decreasing the risk of making a
an error. For example, if the results of a study are type I error increases the risk of making a type II
to be used to determine whether a great deal of error that is, the stricter the researcher is in pre-
money should be spent in an area of nursing care, venting the rejection of a true null hypothesis, the
the researcher may decide that the accuracy of the more likely the researcher is to accept a false null
results is so important that an alpha level of .01 hypothesis. Therefore, researchers always have to
is chosen. In most studies, however, alpha is set accept more of a risk of one type of error when
at .05. setting the alpha level.
Another concept, the P value, is needed to Another method of determining the level of
interpret the alpha value. The value is the con- significance and whether to accept or reject the
ditional probability of obtaining, from the study null hypothesis is called the critical values method
data, a value of the test statistic that is at least as In this method, by calculating the estimates of
extreme as that obtained from the data that is like population mean and standard deviation, a range
itself given that the null hypothesis is true (Kline, of values is determined from which the researcher
2005). The P value is different from alpha because can compare the sample mean findings and decide
it is calculated from the sample data and is con- whether to reject the null hypothesis.
sidered the e act level of signi cance (Gigeren- Suppose researchers want to know the impor-
zer, 1993). Thus, if this exact level of significance tance of support groups for caregivers of older
is less than the conditional a priori probability of adults. They ask 100 caregivers to rate the impor-
making a type I error (P < alpha), then the null tance of support groups to them by using an
hypothesis is rejected, and the result is considered instrument that ranges from 0 (not important at
366 PART FOUR Processes Related to Research

all) to 100 (very important). If Figure 16-3 repre- result of chance. However, such a finding may or
sents the theoretical distribution for this study (a may not have practical significance, even though
normal distribution with a mean of 50), 68 of the finding has statistical significance. Whereas
the population would score between 40 and 60, some authorities would argue that this study
and 95 would score between 30 and 70. Thus, might have relevance to understanding the behav-
the null hypothesis would be that the mean score iour of teenagers, others would argue that the
for the population of caregivers would be 50, and study has no practical value. Thus, the findings of
the scientific hypothesis would be greater or less a study may have statistical significance, but they
than 50. After measurements with this sample are may have no practical value or significance.
completed, the researchers find that the sample Although researchers should consider the
mean score is 75. This mean is consistent with practicality of a problem in the early stages of a
the scientific hypothesis, and the researchers can research project (see Chapter 3), a distinction
be 95 sure that, most of the time, the sample between the statistical and practical significance
mean score would fall under this cut-off thus, of the findings also should be made in the discus-
they would have confidence in rejecting the null sion of the results of a study. Some authorities
hypothesis. In other words, only 5 of 100 times believe that if the findings are not statistically
would they obtain this result by chance alone. significant, they have no practical value. In Sobi-
eraj and colleagues’ (2009) study, the research
Helpful Hint hypothesis was not statistically supported, but
Decreasing the alpha level acceptable for a study nonsupported hypotheses provide as much infor-
increases the chance that a type II error will occur. When mation about the intervention as do the supported
a researcher is conducting many statistical tests, the hypotheses. The data allowed Sobieraj and col-
probability that some of the test results will be signifi-
cant increases as the number of tests increases. There-
leagues to return to the previous literature in the
fore, when a large number of tests are being conducted, area and discern from those findings both statisti-
many researchers decrease the alpha level to .01. cal and practical significance.

Practical and Statistical Significance Evidence-Informed Practice Tip

You study the results to determine the effective-
Statistical significance and practical significance ness of the new treatment and the size and clinical
are not the same. When a researcher finds a importance of the effect.
hypothesis statistically significant, this finding is
unlikely to have happened by chance. In other
words, if the level of significance has been set at Tests of Statistical Significance
.05, the odds are 95 that the researcher will Tests of statistical significance may be parametric
make the correct conclusion on the basis of the or nonparametric. In most studies in nursing
results of the statistical test performed on sample research literature, investigators use parametric
data. The researcher would reach the wrong con- tests that have the following three attributes:
clusion only 5 times in 100. 1. The estimation of at least one population
Suppose that a researcher is interested in the parameter
effect of loud rock music on the behaviour of 2. Measurement at the interval level or higher
laboratory mice. The researcher could design an 3. Assumptions about the variables being
experiment to study this question and find that studied
loud music makes the mice act strangely. A sta- One assumption is usually that the variable is
tistical test suggests that this finding is not the normally distributed in the overall population.
 CHAPTER 16 Quantitative Data Analysis 367

In contrast to parametric tests, non ara etric that this result could not have been obtained by
tests o signi cance are not based on the estima- chance.
tion of population parameters, so their assump- The likelihood of finding a statistic that is high
tions about the underlying distribution are less enough to be statistically significant is increased
restrictive. Nonparametric tests are usually as the sample size increases. This likelihood is
applied when the variables have been measured indicated by the degrees of freedom, which are
on a nominal or ordinal scale. often reported with the statistic and the probabil-
Some debate surrounds the relative merits of ity value. sually abbreviated as df, the degree
the two types of statistical tests. The moderate o reedo is the freedom of a score’s value to
position taken by most researchers and statisti- vary depending on the other scores and the sum
cians is that non ara etric statistics—also of these scores thus, df = N − 1. For example,
called distribution-free tests—are best used when imagine you have four numbers represented by
the data cannot be assumed to be at the interval letters (a, b, c, and d) that must add up to a total
level of measurement or when the sample is small of you are free to randomly choose the first
and the normality of the underlying distribution three numbers, but the fourth must be chosen to
cannot be inferred. If these assumptions can be make the total equal to , and thus your degree of
made, however, most researchers prefer to use freedom is 3.
ara etric statistics which are more powerful To make statistical inferences from data, many
and more exible than nonparametric statistics. types of tests can be conducted. Tables 16-5 and
Because stringent assumptions for parametric 16-6 list the tests most commonly used for infer-
tests makes them more powerful than nonpara- ential statistics. The test used depends on the
metric tests, researchers are able to formulate level of the measurement of the variables in ques-
simple sample statistics, such as the mean and the tion and the type of hypothesis being studied.
standard deviation, which enables them to accu- These statistics test two types of hypotheses: that
rately estimate population parameters with stan- difference exists between groups (see Table 16-5)
dard sampling distributions to obtain probabilities and that a relationship exists between two or
regarding the null hypotheses. more variables (see Table 16-6). In addition,
Researchers use many different statistical tests many types of regression analyses are available
of significance to test hypotheses however, the to predict the dependent variable. Simple regres-
procedure and the rationale for their use are sion analyses (one independent variable) and
similar from test to test. Once the researcher has multiple regression analyses (several independent
chosen a significance level and collected the data, variables) are used when the dependent variable
the data are used to compute the appropriate test is at the interval level or higher. In logistic regres
statistic. Each test has a related theoretical distri- sion logit analysis relationships between
bution that shows the probable and improbable multiple independent variables and a dependent
values for that statistic. On the basis of the statis- variable that is binary, ordinal, or polynomial are
tical result and the values in the distribution, the analyzed.
researcher either accepts or rejects the null
hypothesis and then reports both the statistical
Helpful Hint
result and its probability. Thus, a researcher
The use of nonparametric statistics in a study
may perform a t test, obtain a value of 8.98, and does not mean that the study is useless. The use of
report that it is statistically significant at the P < nonparametric statistics is appropriate when measure-
.05 level. This means that in 100 tests, the ments are not made at the interval level or the variable
under study is not normally distributed.
researcher had five chances to conclude wrongly
368 PART FOUR Processes Related to Research

2.16, df = 269, P = .032). Check for differences

Evidence-Informed Practice Tip
in means and standard deviations in Table 16-4.
Try to discern whether the test for analyzing the
data was chosen because it gave a significant P value.
A statistical test should be chosen on the basis of its Evidence-Informed Practice Tip
appropriateness for the type of data collected, not Tests of difference are most commonly used in
because it gives the answer that the researcher hoped experimental and quasi-experimental designs that
to obtain. provide level II and level III evidence.

Tests of Differences
PARAMETRIC TESTS. The statistic is commonly
The type of test used for any particular study used in nursing research. This statistic re ects
depends primarily on whether the researcher whether two group means are different. Thus, the
examines differences in one, two, or three or t statistic is used when the researcher has two
more groups and whether the data to be analyzed groups, and the question is whether the mean
are nominal, ordinal, or interval (see Table 16-5). scores on some measure are more different than
Suppose that a researcher constructs an experi- would be expected by chance. To use this test, the
mental study with an after-only design (see variables must have been measured at the interval
Chapter 10). What the researcher hopes to deter- or ratio level, and the two groups must be inde-
mine is that the two randomly assigned groups pendent, meaning that nothing in one group helps
are different after the introduction of the experi- determine what is in the other group. If the groups
mental treatment. If the measurements taken are are related in some way, as when samples are
at the interval level, the researcher would use the matched (see Chapter 12), and the researcher also
t test to analyze the data. If the t statistic was wants to determine differences between the two
found to be high enough to be unlikely to have groups, a paired, or correlated, t test would be
occurred by chance, the researcher would reject used.
the null hypothesis and conclude that the two The t statistic illustrates one of the major pur-
groups were indeed more different than would poses of research in nursing: to demonstrate that
have been expected on the basis of chance alone. differences exist between groups. Groups may be
In other words, the researcher would conclude naturally occurring collections, such as age
that the experimental treatment had the desired groups, or they may be experimentally created,
effect. such as treatment and control groups. Sometimes
Schneider and colleagues’ (2011) study illus- a study has more than two groups, or measure-
trated the use of the t statistic (see Table 16-4). In ments are taken more than once. For example,
this study, the t test was used to determine differ- Sobieraj and colleagues (2009) used three
ences in sample characteristics and psychosocial groups—mothers, fathers, and both parents—to
outcome variables by gender. As noted, the describe and compare distress scores when one or
women were slightly younger than the men (t = both parents were present in the procedure room
−2.61, df = 271, P = .009), reported more diffi- (see Table C-2 in Appendix C). These researchers
culty in managing the costs associated with care- used analysis o variance ANO A a test
giving and the child’s illness (t = 2.42, df = 252, similar to the t test, because there were three
P = .016), spent significantly more hours per groups. Like the t statistic, the ANOVA statistic
week providing care (t = 4.29, df = 237, P = .000), is used to test whether group means differ, but
and ascribed more importance to religion (t = instead of testing each pair of means separately,
 CHAPTER 16 Quantitative Data Analysis 369

ANOVA accounts for the variation between POST HOC ANALYSIS. When the decision according
groups and within groups. The ANOVA is usually to the ANOVA is to reject the null hypothesis,
performed with two or more groups by an test this indicates that at least one of the means is not
rather than multiple pairs of t tests (see Practical the same as the other means, as in Wagner and
Application box). If multiple pairs of t tests are associates’ (2009) study. To determine where the
done, the type I error rate would increase. difference in means lies, a ost hoc analysis is
conducted in this analysis, pairs of means in the
main effects and interaction effects are compared
Practical Application to determine whether they are statistically differ-
ent. Many post hoc analyses are available the
Wagner and associates (2009) described
perceptions of workplace safety culture most common include Tukey’s Honestly Signifi-
among nurses in a long-term care setting. One of cant Difference (HSD), the Scheff analysis, and
their analyses consisted of examining whether place the Bonferroni analysis. This type of analysis is
of employment (a for-profit corporation, a nonprofit
corporation, or a governmental agency) had an effect also known as paired comparisons
on perceptions of safety culture. They found that
safety culture perceptions were significantly less
positive among participants working in Helpful Hint
nonmanagement positions: F(2, 545) = 6.943, A research report may not always refer to the test
P < .001. The ANOVA indicated that there were that was done. The reader can find this information by
differences among the three groups in perceptions. looking at the tables. For example, a table with t statis-
From the post hoc analyses, Wagner and associates tics contains a column for t values, and an ANOVA table
determined that licensed nurses working in nonprofit lists F values.
settings reported more positive safety culture
perceptions than did those working in for-profit and
governmental settings. Schneider and colleagues (2011) reported that
the women were slightly younger than the men
(t = −2.61, df = 271, P = .009), reported more
When more than two groups are compared difficulty in managing the costs associated with
over time, a repeated-measures ANOVA is used caregiving and the child’s illness (t = 2.42, df =
because this variation of the ANOVA takes into 252, P = .016), spent significantly more hours per
account the fact that multiple measures at several week providing care (t = 4.29, df = 237, P = .000),
times affect the potential range of scores. and ascribed more importance to religion (t =
Leung and colleagues (2007) examined exer- 2.16, df = 269, P = .032) In other cases, particu-
cise behaviour patterns in patients with cardiac larly in experimental work, researchers use t tests
disease and investigated the sociodemographic, or ANOVA to determine whether random assign-
clinical, environmental correlates of exercise pat- ment to groups was effective in creating groups
terns. Data were collected at baseline and 9 and that are equivalent before the experimental treat-
18 months after discharge. Leung and colleagues ment is introduced. In this case, a researcher
used repeated-measures ANOVA to determine wants to show that no difference exists among the
whether changes had occurred between the base- groups.
line and 6-month collection times. Another In many cases, researchers check whether
expansion of the notion of ANOVA is ulti le groups are different at the beginning of a study
analysis o variance ANO A which is also or baseline by using the technique of analysis o
used to determine differences in group means, but covariance ANCO A ANCOVA also entails
only with more than one dependent variable. measuring differences among group means and
370 PART FOUR Processes Related to Research

helps researchers equate the groups under study scope of this chapter readers who desire further
on an important variable. information should consult a general statistics
Fox and associates (2010) investigated the book.
relationship between bed rest and sitting ortho- In nursing research studies, several different
static intolerance in adults residing in chronic statistical tests are often used. Schneider and col-
care facilities. They used the ANCOVA to leagues’ (2011) study illustrated the use of several
examine the between-group differences in self- of these statistical tests. They were interested in
reported orthostatic intolerance, while controlling comparing the psychological outcomes between
for drugs with known orthostatic effects. two groups, and although the patients could not
be randomly assigned to the group because of the
NONPARAMETRIC TESTS. As mentioned previously, attribute of gender, the researchers needed to
Schneider and colleagues (2011) tested whether determine whether the convenience sampling
differences existed between men and women in procedure succeeded in creating equivalent
marital status, employment, change in employ- groups. For data measured at the nominal level,
ment status, and whether their employer allowed such as marital status, the chi-square statistic was
them to take time off work. These variables are used, as mentioned previously. For data measured
not interval-level data but categorical, and so this at the interval level, such as the age, hours per
difference could not be analyzed with any of the week spent providing care, self-esteem, and spiri-
tests discussed thus far. tual involvement and beliefs scales, the t test was
When data are at the nominal or ordinal level used. Finally, to test the differences between the
and the researcher wants to determine whether two groups, the chi-square method was used for
groups are different, the chi-square, another com- nominal variables, such as change in employment
monly used statistic, is helpful. The chi s uare status and community (see Table 16-4).
χ is a nonparametric statistic used to determine
whether the frequency in each category is differ- Tests of Relationships
ent from what would be expected by chance. As Researchers often are interested in exploring the
with the t test and ANOVA, if the calculated chi- relationship between two or more variables. In
square is high enough, the researcher would con- such studies, they use statistics that determine the
clude that the frequencies found would not be correlation or the degree of association, between
expected on the basis of chance alone, and the two or more variables. Tests of the relationships
null hypothesis would be rejected. Although this between variables are sometimes considered to be
test is robust and can be used in many different descriptive statistics when they are used to
situations, it cannot be used to compare frequen- describe the magnitude and direction of a rela-
cies when samples are small and expected fre- tionship of two variables in a sample and when
quencies are less than six in each cell. In those the researcher does not wish to make statements
instances, isher s e act robability test is about the larger population. Such statistics also
used. can be inferential when they are used to test
When the data are ranks, or are at the ordinal hypotheses about the correlations that exist in the
level, several other nonparametric tests may be target population.
used: the Kolmogorov-Smirnov test, the sign In tests of the null hypothesis, no relationship
test, the Wilcoxon matched-pairs test, the signed- is assumed to exist between the variables. Thus,
rank test for related groups, the median test, when a researcher rejects this type of null hypoth-
and the Mann-Whitney test for independent esis, the conclusion is that the variables are, in
groups. Explanation of these tests is beyond the fact, related. Suppose that a researcher is
 CHAPTER 16 Quantitative Data Analysis 371

interested in the relationship between the age of recovery time is. A negative coefficient would
patients and the length of time it takes them to imply that the younger the patient is, the longer
recover from surgery. As with other statistics dis- the recovery time is. Figure 16-6 illustrates a
cussed, the researcher would design a study to perfect positive correlation, a perfect negative
collect the appropriate data and then analyze the correlation, and a zero correlation. A correlation
data by using measures of association. In this value of 0 to .2 is considered extremely weak, a
example, age and length of time until recovery value of .2 to .4 is weak, a value of .4 to .6 is
can be considered interval measurements. The moderate, a value of .6 to .8 is strong, and a value
researcher would use the Pearson correlation of .8 to 1.0 is very strong (Bluman, 2009).
coe cient (Pearson r also called the Pearson Of course, relationships are rarely perfect. The
product-moment correlation coef cient) in which magnitude of the relationship is indicated by how
the calculation re ects the degree of relationship close the correlation is to the absolute value of 1
between two interval variables. The distribution (see the Practical Application box). Thus, a cor-
of the Pearson r enables the researcher to deter- relation of −.76 is just as strong as a correlation
mine whether the value obtained is likely to have of +.76, but the direction of the relationship is
occurred by chance. Again, the research reports opposite. In addition, a correlation of .76 is stron-
both the value of the correlation and its probabil- ger than a correlation of .32. In testing hypotheses
ity of occurring by chance. about the relationships between two variables, the
Correlation coefficients can range in value researcher considers whether the magnitude of
from −1.0 to +1.0 and also can be zero. A zero the correlation is large enough not to have
coefficient means that no relationship exists occurred by chance. This is the meaning of the
between the variables. A perfect positive correla- probability value, or the P value, reported with
tion is indicated by a coefficient of +1.0, and a correlation coefficients. As with other statistical
perfect negative correlation by a coefficient of tests of significance, the larger the sample is, the
−1.0. The meaning of these coefficients is illus- greater is the likelihood of finding a significant
trated by the example from the previous para- correlation. Therefore, researchers also report the
graph. If no relationship exists between the age degrees of freedom associated with the test
of the patient and the time required for the patient performed.
to recover from surgery, the correlation would be Nominal- and ordinal-level data also can be
zero. However, a correlation of +1.0 would mean tested for relationships by nonparametric statis-
that the older the patient is, the longer the tics. When two variables being tested are only

0 Perfect + 0 Perfect – 0 No
correlation correlation correlation
FIGURE 16-6 Scatter plots illustrating the different types of correlations.
372 PART FOUR Processes Related to Research

For example, a researcher may be interested in

Practical Application determining what factors help women decide to
An example of a descriptive, correlational breast-feed their infants. A number of variables—
study is that of Bailey and colleagues (2010),
such as the mother’s age, previous experience with
who described family member perception of
informational support, anxiety, satisfaction with care, breast-feeding, number of other children, and
and their interrelationships in the intensive care knowledge of the advantages of breast-feeding—
setting. Bailey and colleagues found a significant might be measured and then analyzed to deter-
positive correlation between informational support
and satisfaction with care (r = 0.741, P < .001). mine whether they, separately and together, are
Schneider and colleagues (2011) also conducted predictive of the length of breast-feeding. Such a
correlations between two interval variables and study would require the use of multiple regres-
found positive correlations for women between
meaningfulness in caregiving and self-esteem sion. The results of such a study might help nurses
(r = .34, P = .000), optimism (r = .19, P = .006), know that a younger mother with only one other
spirituality (r = .48, P = .000), and post-trauma child might be more likely to benefit from a
growth (r = .66, P = .000). All of these variables
teaching program about breast-feeding than would
were measured with validated scales.
an older mother with several other children.
In reading research reports, you will often see
multiple regression techniques described as
dichotomous (e.g., male/female yes/no), the phi for ard solution, bac ard solution, or step ise
coefficient can be used to express relationships. solution These techniques are used in multiple
When the researcher is interested in the relation- regression to find the smallest group of variables
ship between a nominal variable and an interval that will account for the greatest proportion of
variable, the point-biserial correlation is used. variance in the dependent variable. In the forward
Spearman’s rho is used to determine the degree solution, the independent variable that has the
of association between two sets of ranks, as is highest correlation with the dependent variables
Kendall’s tau. All these correlation coefficients is entered first, and the next variable is the one
may range in value from −1.0 to +1.0. These tests that will increase the explained variance the most.
are listed in Table 16-6. In the backward solution, all variables are entered
Nursing problems are rarely so simple that into the solution, and each variable is deleted to
they can be explained by only two variables. determine whether the explained variance drops
When researchers are interested in studying significantly. The stepwise solution is a combina-
complex relationships among more than two vari- tion of the two approaches. In general, all of the
ables, they use techniques other than those dis- approaches yield similar, although not identical,
cussed thus far. When researchers are interested results.
in understanding more about a problem than Suppose that the individual who was research-
just the relationship between two variables, they ing breast-feeding was interested in not just
often use ulti le regression in which the breast-feeding but also maternal satisfaction.
relationship between one dependent variable Canonical correlation is used with more than one
at the interval level and several independent dependent variable. If the data are nominal or
variables is measured. Multiple regression is the ordinal, the contingency coefficient or discrimi-
expansion of correlation to include more than nant function analyses are used. These three tests
two variables and is used when the researcher are beyond the scope of this text further informa-
wants to determine what variables contribute to tion can be found in statistical texts.
the explanation of the dependent variable and to Janzen and Hadjistavropoulos (2008) were
what degree. interested in understanding to what extent anxiety
 CHAPTER 16 Quantitative Data Analysis 373

sensitivity, health anxiety, depression, and anxiety of precision or uncertainty about the sample find-
were predictive of coping, concern for waiting, ings. In other words, the confidence interval is an
and anxiety about surgery in individuals waiting estimated range of values, which is likely to
for surgery. Three regression analyses were con- include an unknown population parameter calcu-
ducted. The results indicated that health anxiety lated from a given set of sample data.
(β = .43, P > .05) and anxiety sensitivity (β = .50, Typically, investigators record their confidence
P > .05) were statistically significant predictors interval results as a 95 degree of certainty
of emotional coping. These data allowed them to sometimes, the degree of certainty is recorded as
build on the past research that they had reviewed 99 . Today, professional journals often require
and to suggest both future descriptive and inter- investigators to report confidence intervals as
vention research, thus moving the data toward one of the statistical methods used to interpret
evidence-informed practice. study findings. Even when confidence intervals
are not reported, they can be easily calculated
Evidence-Informed Practice Tip from study data. The method for performing
Tests of relationship are usually associated with these calculations is widely available in statistical
nonexperimental designs that provide level IV evidence. texts.
A strong, statistically significant relationship between
variables often provides support for replicating the
Ostry and associates (2010) explored whether
study, in order to increase the consistency of the findings differences in mental health outcomes were
and provide a foundation for developing an intervention observable between a cohort of sawmill workers
study. living in rural areas and a cohort living in urban
places in British Columbia (Table 16-7). The con-
fidence interval helps place the results in context
The Use of Confidence Intervals for all patients in the study. The results shown in
A con dence interval is a range of values, based Table 16-7 demonstrate, for example, that workers
on a random sample, that is often described with who remain at an urban mill have higher odds for
measures of central tendency and measures of neurotic disorder (value greater than 1.0), adjust-
association and provides the nurse with a measure ment reaction, and acute reaction to stress.

TABLE 16-7
UNIVARIATE ANALYSES: ODDS RATIOS FOR FOUR MENTAL HEALTH DIAGNOSES AMONG SAWMILL
WORKERS, 1994 TO 2001
MENTAL HEALTH DIAGNOSIS*
NEUROTIC DISORDER: ACUTE REACTION TO ADJUSTMENT REACTION: ANXIETY/DEPRESSION:
ICD-9 CODE 300 STRESS: ICD-9 CODE ICD-9 CODE 309 ICD-9 CODE 311
LOCATION (N = 6306) 308 (N = 4104) (N = 2133) (N = 7816)

Urban stay 1.14 (1.02–1.27) 1.04 (0.84–1.29) 1.42 (1.08–1.87) 0.99 (0.85–1.15)
Migrate from urban 0.67 (0.48–0.93) 1.19 (0.82–1.72) 0.82 (0.48–1.38) 0.94 (0.73–1.21)
Rural stay 0.94 (0.79–1.11) 0.68 (0.54–0.86) 0.74 (0.55–0.99) 1.04 (0.89–1.21)
Migrate from rural to urban 1.58 (1.28–1.94) 1.69 (1.30–2.19) 1.54 (1.11–2.13) 1.30 (1.09–1.56)
Migrate from urban to rural 0.95 (0.60–0.94) 0.86 (0.66–1.11) 0.63 (0.44–0.99) 0.77 (0.63–0.92)
Adapted from Ostry, A., Maggi, S., Hershler, R., Chen, L., & Hertzman, C. (2010). Mental health differences among middle-aged sawmill workers in
rural compared to urban British Columbia. Canadian Journal of Nursing Research, 42(3), 84-100.
ICD-9, International Classification of Diseases and Related Health Problems, Ninth Revision (World Health Organization, 1977).
*Numbers in parentheses are 95% confidence intervals.
374 PART FOUR Processes Related to Research

The use of the odds ratio to describe the prob-

Helpful Hint ability of an event is illustrated by a study in
When evaluating whether you should spend time which investigators sought to determine risk
reviewing an article, examine the article’s tables. The
information you need to answer your clinical question factors for the postneonatal death of full-term,
should be contained in one or more of the tables. healthy infants born to young mothers (Phipps,
Blume, DeMonner, 2002). The authors used a
large data set (N = 1,830,350) of mothers 12 to
Harm Studies
29 years of age who had delivered healthy babies.
The odds ratio in logistic regression can be used The investigators wanted to determine whether
in exploring clinical questions of harm, when the death within the first year after birth was more
investigators want to determine whether an indi- likely to occur among infants of young mothers
vidual has been harmed by being exposed to a than among infants of older mothers. The main
particular event. In this type of study, investiga- finding of the study, stratified by age, is described
tors select the outcome of interest (e.g., pressure in Table 16-9. For both unadjusted and adjusted
ulcers) and try to determine whether any one odds, as the mother’s age declined, the probability
factor explains why a patient has or does not have of death of an infant in the postneonatal period
the outcome of interest. The measure of associa- increased (the odds ratio became larger). Healthy,
tion that best describes the analyzed data is the full-term infants born to mothers younger than 15
odds ratio which communicates that one event years of age were three to four times more likely
is likelier to occur than other events. An odds to die within their first year than were infants born
ratio is calculated by dividing the odds in the to mothers 23 to 29 years of age. The study was
treated or exposed group by the odds in the not designed to determine which factors are asso-
control group. Investigators present an odds ratio ciated with death. However, Phipps and associates
of factors in study tables thus, calculation of the (2002) indicated that unmeasured social factors
odds ratio is rarely necessary. The interpretation could have a potential association further studies
of the odds ratio is straightforward and presented are needed to test the validity of this association.
in Table 16-8. Note that the null value for the With this evidence, a nurse could justify the need
odds ratio is equal to 1. for increased home care visits to younger mothers

TABLE 16-8
TABLE 16-9
INTERPRETATION OF ODDS RATIOS ODDS RATIOS* FOR POSTNEONATAL
BENEFICIAL MORTALITY ASSOCIATED WITH MATERNAL AGE
OUTCOME (E.G.,
GROUPS
ADVERSE OUTCOME ADHERENCE TO
(E.G., MYOCARDIAL MEDICATION MATERNAL CRUDE ODDS ADJUSTED
ODDS RATIO INFARCTION) REGIMEN) AGE (YEARS) RATIO ODDS RATIO†

<1 (e.g., 0.375) Intervention Intervention ≤15 4.1 (3.4–4.8) 3.0 (2.5–3.6)
produced better produced 16–17 3.1 (2.8–3.5) 2.4 (2.1–2.7)
results worse results 18–19 2.5 (2.3–2.8) 2.0 (1.8–2.3)
1 Intervention Intervention 20–22 1.8 (1.6–2.0) 1.5 (1.4–1.7)
produced no produced no 23–29 1.0 1.0
better/worse better/worse Modified from Phipps, M., Blume, J., & DeMonner, S. (2002). Young
results results maternal age associated with increased risk of postneonatal death.
>1 (e.g., 4.0) Intervention Intervention Obstetrics and Gynecology, 100, 481–486.
produced worse produced *Numbers in parentheses are 95% confidence intervals.
†
results better results Adjusted for maternal race/ethnicity, adequacy of prenatal care
utilization, and marital status.
 CHAPTER 16 Quantitative Data Analysis 375

in the immediate postpartum period. Another experimental studies by using their measures of
nurse can use this evidence to start a nurse- association (see Table 16-6). An odds ratio is the
managed support group for younger mothers. statistic of choice for use in a meta-analysis. The
Harm data, with their measure of probabilities, same interpretation of odds ratio described in
help nurses identify factors that may or may not Table 16-8 applies to the odds ratios obtained in
contribute to an adverse or beneficial outcome. a meta-analysis.
The usual manner of displaying data from a
Meta-analysis meta-analysis is by a pictorial representation
Meta-analysis is not a type of study design but a known as a blobbogram, accompanied by a
research method in which the results of multiple summary measure of effect size in odds ratios. In
studies (usually randomized controlled trials) are the meta-analysis depicted in Figure 16-7, the
statistically combined to answer focused clinical investigators were interested in comparing the
questions through an objective appraisal of care- efficacy of a beta-agonist given by a metered-dose
fully synthesized research evidence (see Chapter inhaler with a chamber versus a nebulizer on hos-
11). People sometimes use the terms meta- pital admission in children younger than 5 years
analysis and systematic revie interchangeably (Castro-Rodriquez Rodrigo, 2004). The inves-
however, a meta-analysis is a quantitative analy- tigators searched the literature for randomized
sis used in a systematic review. controlled trials with children younger than 5
Syste atic revie is the process whereby the years with acute asthma who were treated in the
investigators evaluate all relevant studies, pub- emergency department and were randomly
lished and unpublished, on the topic or question assigned to receive treatment either a metered-
(Higgins Green, 2009). At least two members dose inhaler with a chamber or a nebulizer. The
of the review team independently assess the investigators found six trials that met this criterion
quality of each study, include or exclude studies (see Figure 16-7). The study groups are repre-
on the basis of preestablished criteria, statistically sented by a fraction for example, in the first trial
combine the results of individual studies, and listed, of 17 children who received metered-dose
present a balanced and impartial summary of the inhalers with a chamber, 4 were admitted to the
findings that represents a state-of-the-science hospital, and of 17 children who received nebuliz-
conclusion about the evidence supporting the ers, 4 were admitted to the hospital. In the centre
benefits and risks of a given health care of the figure, each trial in the analysis is repre-
practice. sented by a horizontal line. The findings from
In the evidence-informed hierarchy, the find- each study are represented as a blob or square (the
ings of a systematic review are considered to measured effect) on the vertical line. The size of
provide the strongest evidence available to the the blob or square (sometimes just a small vertical
clinician because they summarize large amounts line) re ects the amount of information in that
of information derived from multiple experimen- study. The width of the horizontal line represents
tal studies of the effect of the same intervention. the 95 confidence interval. A vertical line is the
A methodologically sound systematic review line of no effect (odds ratio = 1). When the confi-
with a rigorous meta-analysis is more likely than dence interval of the result (horizontal line)
an individual study to identify the true effect of crosses the line of no effect (vertical line), then the
an intervention because the meta-analysis limits differences in the effect of the treatment are not
bias. statistically significant. If the confidence interval
In a systematic review, the researcher quanti- does not cross the vertical line, then the study
tatively combines the data from the selected results are statistically significant.
376 PART FOUR Processes Related to Research

MDI + VHC Nebulization OR Weight OR

Study n/N n/N (95% CI Random) % (95% CI Random)
Closa 4/17   4/17 12.2 1.00 (0.20–4.88)
Delgado 5/83 20/85 28.7 0.21 (0.07–0.59)
Leversha 10/30 18/30 27.6 0.33 (0.12–0.96)
Mandelberg 6/23   7/19 17.7 0.61 (0.16–2.26)
Ploin 3/32 3/32 10.8 1.00 (0.19–5.37)
Rubilar 0/62 1/61 3.0 0.32 (0.01–8.08)
Total (95% CI) 28/247 53/244 100.0 0.42 (0.24–0.72)
Test for heterogeneity chi-square = 4.46, df = 55, p = .49
Test for overall effect z = –3.10, p = .002
.01 .1 1 10 100
Favours MDI + VHC Favours nebulizer
FIGURE 16-7 Tabular display of systematic review with meta-analysis data. These data reflect
the efficacy of administering a β-agonist by metered-dose inhaler (MDI) with a valved holding
chamber (VHC) versus a nebulizer in children younger than 5 years who had acute exacerbation
of wheezing or asthma in the emergency department on hospitalization.
CI, confidence interval; OR: odds ratio.
From Castro-Rodriquez, J., & Rodrigo, G. (2004). Beta-agonists through metered-dose inhaler
with valved holding chamber versus nebulizer for acute exacerbation of wheezing or asthma
in children under 5 years of age: A systematic review with meta-analysis. Journal of Pediatrics,
145(2), 172–177. doi:10.1016/j.jpeds.2004.04.007, Copyright 2004, with permission from
Elsevier.

In the blobbograms in Figure 16-7, only two it would support the clinical practice of providing
of the six studies (the second and third listed) metered-dose inhalers with a chamber to children
yielded analysis lines that did not cross the line younger than 5 years of age with asthma exacer-
of no effect. Because the analysis lines did not bation in order to prevent hospitalization.
cross the line of no effect, these studies have
statistically significant findings. In the two right Advanced Statistics
columns of the figure, Castro-Rodriquez and Sometimes, researchers are interested in even
Rodrigo (2004) also provided the numerical more complex problems. For example, Linton
equivalent of each blobbogram. Other important and colleagues (2009) conducted a cross-sectional
information and additional statistical analysis study to identify predictors of human immunode-
may accompany the blobbogram, such as a test ficiency virus (HIV) status in homeless youth.
to determine the degree to which the results of Linton and colleagues had a sample size of nearly
each of the individual trials are mathematically 140 patients who accessed preventive health and
compatible (heterogeneity). For more informa- social services. The results of the logistic regres-
tion, refer to a book of advanced research methods. sion revealed that the variables of age and ethnic-
In Figure 16-7, the summary odds ratio for all ity were significant predictors of self-reported
of the studies combined is represented by a HIV status. In comparison with 18- to 25-year-
diamond. In this case, after the results of each of old participants, participants who were 26 to 30
the controlled trials are statistically pooled, these years of age were at increased odds (11 times
statistically combined results favour the metered- higher) to self-report their HIV status as being
dose inhaler with a chamber for preventing hos- positive (β = 2.413, P < .001, odds ratio = 11.17,
pitalization of children younger than 5 years of 95 confidence interval = 3.344 to 37.303). On
age, and this option is statistically significant. If the basis of a proposed model, the relationships
this meta-analysis were methodologically sound, between the independent and dependent variables
 CHAPTER 16 Quantitative Data Analysis 377

were tested through logistic regression analysis.

Logistic regression is a form of advanced statis- The Use of Statistics
tics used when a researcher wishes to confirm the Statistics are used in nursing research to describe
relationship of a set of categorical data (data that the samples of research studies and to test for
have a discrete value). hypothesized differences or associations in the
This notion of testing specific relationships in sample. Knowing the characteristics of the sample
a specific order can be extended further to test of a research study allows the researcher to deter-
hypothesized variables that are made up of mine the population for whom the results will
several measures. In structural equation model- be generalized. For example, if a study sample
ling, path models made up of variables that are was primarily white with a mean age of 42 years
not actually measured are tested. For example, a (SD = 2.5), the findings may not be applicable to
researcher might study the concept of self-esteem elderly Punjabi men and women. The cultural,
and use three different measures to determine demographic, or clinical factors of an elderly
participants’ levels of self-esteem. The researcher population of a different ethnic group may con-
would test how carefully these three measures tribute to different results. Thus, understanding
gauge self-esteem by testing a measurement the descriptive statistics of a study assists in
model for example, Samuels-Dennis and associ- determining the applicability of findings to differ-
ates (2010) tested the validity of the intersection- ent practice settings.
ality model of trauma and posttraumatic stress Statistics are also used to test hypotheses pro-
disorder. posed by the researchers. Inferential statistics
Another advanced technique often used in used to analyze data (e.g., t test, test, r coeffi-
nursing research is factor analysis. Factor analy- cient) and the associated significance level (P
sis helps researchers understand concepts more value) indicate the likelihood that the association
fully and contributes to their ability to measure or difference found in any study results from
concepts reliably and validly (see Chapter 14). In chance or because of a true difference between
actor analysis a large number of variables are groups. The closer the P value is to zero, the
grouped into a smaller number of factors to less likely it is for the association or difference
reduce a set of data so that it may be easily of a study to result from chance. Thus, inferential
described and used this statistical procedure is statistics provide an objective way to determine
used to determine the underlying dimensions or whether the results of the study are likely to be a
components of a variable. Factor analysis is also true representation of reality (Munro, 2004).
used for instrument development and theory
development.
In instrument development, factor analysis is Evidence-Informed Practice Tip
used to group individual items on a scale into A basic understanding of statistics will improve
your ability to assess the effect of the independent vari-
meaningful factors or subscales. Heaman and able on the dependent variable and related patient out-
Gupton (2009) used factor analysis to refine a comes for your patient population and practice setting.
new instrument, the Perception of Pregnancy
Risk uestionnaire. The original scale contained
11 items, and psychometric analyses were per-
formed on the final 9-item version. Factor analy- EXAMPLE OF THE USE AND CRITIQUE
sis was used to determine whether the scale OF STATISTICS
measured the concepts that it was intended to The purpose of the study by Sobieraj and col-
measure. Many other statistical techniques are leagues’ (2009) was to determine the effect of
available to nurse researchers. music in distracting parents during laceration
378 PART FOUR Processes Related to Research

repair (see Appendix C). The statement of purpose the transitional discharge intervention. Various
implies that the investigators were interested in statistical tests were used to examine differences,
differences between groups thus, an experimen- depending on the level of measurement. Depen-
tal design that provides level II evidence was dent variables calculated at the interval level were
appropriate. Therefore, you should expect the compared in repeated-measures ANOVA.
analysis to consist of statistical tests in which These tests are appropriate for the study design
differences between means were examined, such and the hypotheses because Sobieraj and col-
as t tests or ANOVA. leagues (2009) were interested in differences
Sobieraj and colleagues (2009 see Appendix between the two groups. The results for each of
C) adequately described sample characteristics. If the hypotheses were suggestive of differences in
the participants who did not complete the study some of the outcomes between the two groups.
differed from those who completed the study, the The tables agreed with the text, and the results
findings would be difficult to interpret (i.e., those were understandable to readers. The discussion
who completed the program had fewer problems). pointed out limitations to the study. Clear impli-
Dependent variables consisted of quality of life, cations for practice were described, and they sup-
health care utilization, and levels of functioning ported the practical significance of the study. The
and were measured over time at discharge and 1 statistical level of significance was set at .05 and
month and 12 months after discharge. Sobieraj and was consistent throughout the article. Therefore,
colleagues were interested in looking at differ- the researchers’ statistics were appropriate to the
ences between the participants who received stan- study’s purpose, design, method, sample, and
dard care (control group) and those who received levels of measurement.

APPRAISING THE EVIDENCE

Descriptive and Inferential Statistics
Many students who have not had a course in statistics hypotheses or research questions suggest the type of
think they cannot critique the statistics of research. analysis that follows. The hypotheses or the research
However, students should be able to critically analyze the questions should indicate the major variables that are
use of statistics even if they do not understand how the expected to be presented in summary form. Each of the
numbers presented were derived. What is most impor- variables in the hypotheses or research questions should
tant in critiquing this aspect of a research study is that the be presented in the “Results” section along with appropri-
procedures for summarizing and analyzing the data make ate descriptive information.
sense in view of the purpose of the study (see the Critiqu- After you study the hypotheses or research questions,
ing Criteria box). proceed to the “Methods” section. Using the operational
Before you decide whether the statistics used make definition provided, identify the levels of measurement
sense, return to the beginning of the study and determine used to measure each of the variables listed in the
the purpose. Although descriptive statistics are used in all hypotheses or research questions. From this information,
studies to summarize the data obtained, many investiga- you should be able to determine the measures of central
tors use inferential statistics to test specific hypotheses. In tendency and variability that should be used to summa-
a report of an exploratory study, it is possible that only rize the data. For example, you would not expect to see
descriptive statistics are presented because the purpose a mean used as a summary statistic for the nominal vari-
is to describe the characteristics of a population. able of gender; gender would probably be reported as a
Just as the hypotheses or research questions should frequency distribution. The means and standard devia-
follow from the purpose of a study, so should the tions should be provided for measurements performed at
 CHAPTER 16 Quantitative Data Analysis 379

APPRAISING THE EVIDENCE

Descriptive and Inferential Statistics—cont’d
the interval level. The sample size is another feature nominal or ordinal level, however, the statistics used
described in the “Methods” section that is helpful for should be nonparametric. Also, consider the sample
evaluating the researcher’s use of descriptive statistics. size and remember that samples need to be large enough
The larger the sample is, the less chance there is that one to enable the assumption of normality. If the sample is
outlying score will affect the summary statistics. quite small—for example, 5 to 10 participants—the
If tables or graphs are used, they should agree with researcher may have violated the assumptions necessary
the information presented in the text. The tables and for inferential statistics to be used (see Chapter 12). Thus,
charts should be clearly and completely labelled. If the the important question is whether the researcher has
researcher presents grouped frequency data, the groups provided enough justification to use the statistics
should be logical and mutually exclusive. The size of the presented.
interval in grouped data should not obscure the pattern Finally, consider the results as they are presented.
of the data, nor should it create an artificial pattern. Each Enough data should be presented for each hypothesis or
table and chart should be referred to in the text, but research question for you to determine whether the
each should add to the text, not merely repeat it. Each researcher actually examined each one. The tables
table or graph should have an obvious connection to the should accurately reflect the procedure performed and be
study. in harmony with the text. For example, the text should
In reading a table such as Table 16-4, first look at the not say that a test result reached statistical significance,
table title. The title should give an indication of the infor- but the tables show that the probability value of the test
mation in the table. Next, review the column headings. was higher than .05. If the researcher used analyses that
Do these headings follow from the title? Is each heading are not discussed in this text, you may want to refer to a
clear, and are any nonstandard abbreviations explained? statistics text to decide whether the analysis was appro-
Are the statistics contained in the table appropriate to the priate for the hypothesis or research question and the
level of measurement used? In Table 16-4, the column level of measurement.
headings follow from the title. Each study variable is You should critique two other aspects of the data
listed, along with its mean and standard deviation. Mean analysis. The study should not read as if it were a statisti-
and standard deviation are appropriate statistics because cal textbook. The results should be presented clearly
these data were regarded as interval-level data. enough that the reader can determine what was done
After you evaluate the descriptive statistics, evaluate and what the results were. In addition, the author should
the inferential statistical analysis of a research report, distinguish between the practical and the statistical sig-
beginning with the hypothesis or research question. If the nificance of the evidence in relation to the findings. Some
hypothesis or research question indicates that a relation- results may be statistically significant, but their practical
ship will be found, you should expect to find indices importance may be doubtful in terms of applicability to
of correlation. If the study is experimental or quasiex- a patient population or clinical setting. In this case, the
perimental, the hypothesis should indicate that the author should note the deficiency. Alternatively, a
author is looking for differences between the groups research report may be elegantly presented, but the find-
studied, and you would expect to find statistical tests of ings do not impress you. Such a feeling may indicate that
differences between means that test the effect of the the practical significance of the study and its findings have
intervention. not been adequately explained in the report. From an
As you read the “Methods” section of the article, evidence-informed practice perspective, a significant
again consider the level of measurement used to hypothesis or research question should contribute to
measure the important variables. If the level of measure- improving patient care and clinical outcomes.
ment is interval or ratio, the statistics will probably Note that the critical analysis of a research article’s
be parametric. If the variables are measured at the statistical analysis is not conducted in a vacuum. The

Continued
380 PART FOUR Processes Related to Research

APPRAISING THE EVIDENCE

Descriptive and Inferential Statistics—cont’d
adequacy of the analysis can only be judged in relation use a nonparametric statistic when a parametric statistic
to the other important aspects of the article: the problem, is appropriate. Because parametric statistics are more
the hypotheses, the research question, the design, the powerful than nonparametric statistics, the result of the
data-collection methods, and the sample. If these aspects parametric analysis may not have been what the
of the research process are not considered, the statistics researcher expected. However, the nonparametric result
themselves have very little meaning. Statistics can be might be in the expected direction, and so the researcher
misleading; thus, the researcher must use the appropriate reports only that result.
statistic for the problem. For example, a researcher may

CRITIQUING CRITERIA

1. Were appropriate descriptive 7. Does the hypothesis indicate the hypothesis, the method,
statistics used? that the researcher tested for the sample, and the level of
2. What level of measurement is differences between groups or measurement?
used for each major variable? tested for relationships? What is 12. Are the results for each of the
3. Is the sample size large enough the level of significance? hypotheses presented clearly
to prevent one extreme score 8. Does the level of measurement and appropriately?
from affecting the summary enable the use of parametric 13. If tables and graphs are used,
statistics used? statistics? do they agree with the text and
4. What descriptive statistics are 9. Is the sample size large enough extend it, or do they merely
reported? to use parametric statistics? repeat it?
5. Were these descriptive statistics 10. Has the researcher provided 14. Are the results clear?
appropriate to the level of enough information for you to 15. Is a distinction made between
measurement for each variable? decide whether the appropriate practical significance and
6. Are appropriate summary statistics were used? statistical significance? How is it
statistics provided for each 11. Are the statistics used made?
major variable? appropriate for the problem,

students has taken a statistics course. Would you

CRITICAL THINKING CHALLENGES agree or disagree? Defend your position.
■ What assumptions are violated when a clinical
■ Discuss the ways a researcher might use a
research study uses a convenience sample and
computer to analyze data and present the
applies inferential statistics?
descriptive statistical results of a study.
■ What are the advantages and disadvantages of
■ What is the relationship between the level of
decreasing the alpha level for a study? What is
measurement used and the choice of a statistical
the relationship between setting an alpha level
procedure? How is the level of measurement
and type I and type II errors?
related to the level of evidence in the study
■ Discuss the parameters for using nonparametric
design?
statistics in a study and their effect on the
■ What type of visual depiction can be used to
usefulness of applying the evidence provided by
show the use of correlations? Use examples from
the findings in practice.
clinical practice to illustrate the difference
■ A research study’s findings are not considered
between positive and negative correlations.
significant at the .05 level; are they deemed to
■ A classmate states that it is ridiculous for the
provide evidence that is applicable to practice?
instructor to have students critique the descrip-
Justify your answer.
tive statistics used in a study when none of the
 CHAPTER 16 Quantitative Data Analysis 381

• A type I error is the researchers’ incorrect decision

KEY POINTS to reject the null hypothesis.
• A type II error occurs when the results from the
• Descriptive statistics are a means of describing and sample data lead to the acceptance of the null
organizing data gathered in research. hypothesis when it is actually false; this error is
• The four levels of measurement are nominal, ordinal, also known as beta (β).
interval, and ratio. Measurement at each level is • The researcher controls the risk of making a type I
performed with appropriate descriptive techniques. error by setting the alpha level, or level of
• Measures of central tendency describe the average significance. Unfortunately, reducing the risk of a
member of a sample. The mode is the score that type I error by reducing the level of significance
occurs most frequently, the median is the middle increases the risk of making a type II error.
score, and the mean is the arithmetical average of • The results of statistical tests are reported to be
the scores. The mean is the most stable and useful significant or nonsignificant. For a result to be
of the measures of central tendency and, with the statistically significant, the probability of occurring
standard deviation, forms the basis for many must be less than .05 or .01, depending on the
inferential statistics. level of significance set by the researcher.
• The frequency distribution is depicted in tabular or • Commonly used parametric and nonparametric
graphic form and allows calculation or observation statistical tests include tests for differences between
of characteristics of the data distribution, including means, such as the t test and ANOVA, and tests for
skewness, symmetry, modality, and kurtosis. differences in proportions, such as the chi-square
• In nonsymmetrical distributions, the degree and test.
direction of the pull of the off-centre peak are • Tests in which data are examined for the presence
described in terms of skew. of relationships include the Pearson r, the sign test,
• The ranges reflect differences between high and the Wilcoxon matched-pairs test, the signed-rank
low scores. test, and multiple regression.
• The standard deviation is the most stable and most • Advanced statistical procedures include path
useful measure of variability. It is derived from the analysis and factor analysis.
concept of the normal curve. In the normal curve, • The most important aspect of critiquing statistical
sample scores and the means of large numbers of analyses is the relationship between the statistics
samples cluster around the midpoint in the distri- used and the problem, design, and method used
bution, and a fixed percentage of the scores is in the study. Clues to the appropriate statistical test
within given distances of the mean. This tendency to be used by the researcher should stem from the
of means to approximate the normal curve is called researcher’s hypotheses. You also should
the sampling distribution of the means. A Z score is determine whether all of the hypotheses have
the standard deviation converted to standard units. been presented in the article.
• Because the sampling distribution of the means • A basic understanding of statistics will improve
follows a normal curve, researchers are able to your ability to think about the level of evidence
estimate the probability that a certain sample will provided by the study design and findings and
have the same properties as the total population of their relevance to patient outcomes for your
interest. Sampling distributions provide the basis patient population and practice setting.
for all inferential statistics.
• Inferential statistics allow researchers to estimate
population parameters and to test hypotheses
about populations from sample data. The use of
these statistics allows researchers to make objective REFERENCES
decisions about the outcome of the study. Such Arnold, J. M., Liu, P., Demers, C., Dorian, P., Giannetti,
decisions are based on the rejection or acceptance
N., Haddad, H., et al.(2006). Canadian Cardiovascu-
of the null hypothesis, which is that no relationship
exists between the variables. lar Society consensus conference recommendations
• If the null hypothesis is supported, then the on heart failure 2006: Diagnosis and management.
findings are likely to have occurred by chance. If Canadian ournal of Cardiology, (1), 23-45.
the null hypothesis is rejected, then a relationship Bailey, J. J., Sabbagh, M., Loiselle, C. G., Boileau, J.,
does exist between the variables and is unlikely to McVey, L. (2010). Supporting families in the IC :
have occurred by chance. A descriptive correlational study of informational
• Statistical hypothesis testing is subject to two types support, anxiety, and satisfaction with care. Intensive
of errors: type I and type II. and Critical Care Nursing, , 114-122.
382 PART FOUR Processes Related to Research

Bluman, A. J. (2009). A brief version elementary compared to urban British Columbia. Canadian
statistics: A step by step approach. New York: ournal of Nursing Research, (3), 84-100.
McGraw-Hill. Phipps, M., Blume, J., DeMonner, S. (2002). Young
Castro-Rodriquez, J., Rodrigo, G. (2004). Beta- maternal age associated with increased risk of post-
agonists through metered-dose inhaler with valved neonatal death. bstetrics and ynecology, ,
holding chamber versus nebulizer for acute exacerba- 481-486.
tion of wheezing or asthma in children under 5 years Profetto-McGrath, J., Negrin, K., Hugo, K.,
of age: A systematic review with meta-analysis. Bulmer Smith, K. (2010). Clinical nurse specialists’
ournal of Pediatrics, (2), 172-177. approaches in selecting and using evidence to
Fox, M. T., Sidani, S., Brooks, D. (2010). The improve practice. orldvie s on Evidence-Based
relationship between bed rest and sitting orthostatic Nursing, (1), 36-50.
intolerance in adults residing in chronic care Samuels-Dennis, J., Ford-Gilboe, M., Ray, S. (2010).
facilities. ournal of Nursing and ealthcare of Single mother’s adverse and traumatic experiences
Chronic Illness, , 187-196. and post-traumatic stress symptoms. ournal of
doi:10.1111/j/1752-9824.2010.01058.x amily Violence, (1), 9-20.
Gigerenzer, G. (1993). The superego, the ego, and the id Schneider, M., Steele, R., Cadell, S., Hemsworth, D.
in statistical reasoning. In G. Keren C. Lewis (2011). Differences on psychosocial outcomes
(Eds.), A handboo for data analysis in the behav- between male and female caregivers of children with
ioural sciences: Vol Methodological issues (pp. life-limiting illnesses. ournal of Pediatric Nursing,
311-339). Hillsdale, NJ: Erlbaum. (3), 186-199. doi: 10.1016/j.pedn.2010.01.007
Heaman, M. I., Gupton, A. L. (2009). Psychometric Slakter, M. J., Wu, Y. W. B., Suzuki-Slakter, N. S.
testing of the Perception of Pregnancy Risk uestion- (1991). Statistical nonsense at the .00000 level.
naire. Research in Nursing and ealth, , 493-503. Nursing Research, , 248-249.
Higgins, J. T., Green, S. (Eds.), (2009). Cochrane Snowdon, A. W., Polgar, J., Patrick, L., Stamler, L.
handboo for systematic revie s of interventions. (2006). Parents’ knowledge about use of child safety
New York: The Cochrane Collaboration, John Wiley systems. Canadian ournal of Nursing Research,
Sons. (2), 98-114.
Janzen, J., Hadjistavropoulos, H. D. (2008). Exami- Sobieraj, G., Bhatt, M., LeMay, S., Rennick, J.,
nation of negative affective responses to waiting for Johnston, C. (2009). The effect of music on parental
surgery. Canadian ournal of Nursing Research, participation during pediatric laceration repair.
(4), 72-91. Canadian ournal of Nursing Research, (4), 68-82.
Kline, R. B. (2005). Beyond signi cance testing: Wagner, L. M., Capezuti, E., Rice, J. C. (2009).
Reforming data analysis methods in behavioral Nurses’ perceptions of safety culture in long-term
research. Washington, DC: American Psychological care settings. ournal of Nursing Scholarship, (2),
Association. 184-192.
Leung, Y. W., Ceccato, N., Stewart, D. E., Grace, S. Wong, C. A., Laschinger, H., Cummings, G. G.,
(2007). A prospective examination of patterns and Vincent, L., O’Connor, P. (2010). Decisional
correlates of exercise maintenance in coronary artery involvement of senior nurse leaders in Canadian
disease patients. ournal of Cardiopulmonary acute care hospitals. ournal of Nursing Management,
Rehabilitation and Prevention, (5), 347-357. , 122-133. doi: 10.1111/j.1365-2834.2010.01053.x
Linton, A., Singh, M., Turbow, D., Legg, T. J. (2009). World Health Organization. (1977). International
Street youth in Toronto: An investigation of demo- classi cation of diseases and related health problems,
graphic predictors of high risk behaviors and HIV ninth revision ICD-9 . Geneva, Switzerland: Author.
status. ournal of IV AIDS Social Services, ,
375-396.
Munro, B. H. (2004). Statistical methods for health care
FOR FURTHER STUDY
research (5th ed.). Philadelphia: Lippincott Williams Go to Evolve at http://evolve.elsevier.com/
Wilkins. Canada/LoBiondo/Research for Audio Glossary, how-to
Ostry, A., Maggi, S., Hershler, R., Chen, L., instructions for Writing Proposals for Funding, and
Hertzman, C. (2010). Mental health differences additional research articles for practice in reviewing
among middle-aged sawmill workers in rural and critiquing.
C H A PTER 1 7

Presenting the Findings

Geri LoBiondo-Wood | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Discuss the difference between a study’s “Results” section and the “Discussion” section.
• Identify the format of the “Results” section.
• Determine whether both statistically supported and statistically unsupported findings are
discussed.
• Determine whether the results are objectively reported.
• Describe how tables and figures are used in a research report.
• List the criteria of a meaningful table.
• Identify the format and components of the “Discussion of the Results” section.
• Determine the purpose of the “Discussion” section.
• Discuss the importance of including the generalizations and limitations of a study in the report.
• Determine the purpose of including recommendations in the study report.
• Discuss how the strength, quality, and consistency of evidence provided by the findings are related
to a study’s limitations, generalizability, transferability, and applicability to practice.

KEY TERMS
confidence interval generalizability recommendations
findings limitations transferability

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

383
384 PART FOUR Processes Related to Research

THE ULTIMATE GOALS OF NURSING RESEARCH are to presents in these sections will help you critically
develop nursing knowledge and to promote analyze the findings.
evidence-informed nursing practice, thereby sup-
porting the scientific basis of nursing. From the FINDINGS
viewpoint of the research consumer, the analysis The ndings of a study are the results, conclu-
of the results, interpretations, and the conclusions sions, interpretations, recommendations, general-
that a researcher makes from a study becomes a izations, and implications for future research and
highly important piece of the research report. nursing practice, which are separated into two
After the analysis of the data, the researcher con- major areas: the results and the discussion of the
structs an overall view of the findings, like putting results. The Results section focuses on the
the pieces of a jigsaw puzzle together to view the thematic results or statistical findings of a study,
total picture. This process is analogous to evalu- and the Discussion section focuses on the
ation, the last step in the nursing process. In the remaining topics. For both sections, as well as all
final sections of the report, after the statistical other sections of a report, the same rule applies:
procedures have been applied, the statistical or The content must be presented clearly, concisely,
numerical findings are described in relation to and logically.
the theoretical framework, literature, methods,
hypotheses, and problem statements. In qualita-
tive research, after the content analyses have been Evidence-Informed Practice Tip
concluded, the themes are discussed in relation to Evidence-informed practice is an active process
that requires you to consider how, and whether, research
the literature, problem statements, and a theoreti- findings are applicable to your patient population and
cal framework, as appropriate. practice setting.
The final sections of published research reports
are generally titled Results and Discussion,
but other topics, such as limitations of findings, Results
implications for future research and nursing prac- In the Results section of a research report, the
tice, recommendations, and conclusions, may be researcher presents the quantitative data or
addressed separately or subsumed within these numbers generated by the descriptive and infer-
sections. The format of the Results and Dis- ential statistical tests or the themes from narra-
cussion is contingent on the stylistic consider- tives generated from a content or coding analysis.
ations of the author and the journal. The function The results of the data analysis are the foundation
of these final sections is to depict all aspects of for the interpretations or Discussion section
the research process, as well as to discuss, inter- that follows the results. The Results section
pret, and identify the limitations, generalizations, should then re ect the question being posed or
and applicability relevant to the investigation, hypothesis tested. The information from each
thereby furthering research-based practice. hypothesis or research question should be pre-
The process that both the investigator and you sented sequentially. The tests used to analyze the
as the research consumer use to assess the results data should be identified. If the author does not
of a study is depicted in the Critical Thinking explicitly state the exact test that was used,
Decision Path. The goal of this chapter is to intro- then the values obtained should be noted. The
duce the purpose and content of the final sections researcher typically provides the numerical
of a research investigation, in which the data are values of the statistics and states the specific test
presented, interpreted, discussed, and general- value and probability level achieved (see Chapter
ized. An understanding of what an investigator 16). Examples of statistical tests and the corre-
 CHAPTER 17 Presenting the Findings 385

CRITICAL THINKING DECISION PATH

Assessing Study Results

Results

Descriptive Inferential
analysis analysis

Discussion
of analysis

Interpretation
of analysis

Literature
review

Theoretical Hypothesis or
framework research question

Decision: utility
of results

sponding statistical values can be found in Table TABLE 17-1

17-1. An example of a qualitative analysis with EXAMPLES OF REPORTED STATISTICAL RESULTS
themes and subthemes is given later in this STATISTICAL TEST EXAMPLES OF REPORTED RESULTS
chapter (see Table 17-4). Mean M = 118.28
You should not be intimidated by numbers and Standard deviation SD = 62.5
Pearson correlation r = .39, P < .01
symbols. Although these numbers are important,
Analysis of variance F = 3.59; df = 2, 48; P < .05
they are only one piece of the whole the research (ANOVA)
process is much more important. Whether you t test t = 2.65, P < .01
superficially understand statistics or have an Chi-square χ2 = 2.52, df = 1, P < .05
in-depth knowledge of statistics, you can expect df , degrees of freedom.
386 PART FOUR Processes Related to Research

to find the study results clearly stated. Thus, you section contains not only the results but also
should note the presence or absence of any statis- the researcher’s interpretations, which are more
tically significant results. For the conceptual commonly found in the Discussion section.
meanings of the numbers found in studies, refer Integrating the results with the discussion in a
to the discussion in Chapter 16. report is the decision of the author or the journal
editor. The two sections may be integrated when
a study contains several segments that may be
Helpful Hint
viewed as separate subproblems of a major overall
In the “Results” section of a research report, the
descriptive statistics are generally presented first, fol- problem.
lowed by the results of each hypothesis or research When presenting the results, the investigator
question tested. should show objectivity. The following quotation
gives the appropriate way to express results:
The researcher must present the data for all of Analysis of the effect of time was statistically
the hypotheses posed or research questions asked significant for intensity and unpleasantness related
(e.g., whether the hypotheses were accepted or to pain (F = 160.395, P < 0.0001).
rejected, supported or not supported). If the data
support the hypotheses, you might assume that Investigators would be accused of lacking objec-
the hypotheses were proven, but this is not neces- tivity if they stated the results as follows:
sarily true. It only means that the hypotheses were The results were not surprising as we found a
supported, and the results suggest that the rela- significant relationship between effect of time and
tionships or differences tested, which were intensity and unpleasantness, as we expected.
derived from the theoretical framework, were
probably logical in that study’s sample. Opinions or reactionary statements to the data in
As a novice research consumer, you might also the Results section are therefore avoided. Box
think that if a researcher’s hypotheses are not 17-1 gives examples of objectively stated results.
supported statistically or are only partially sup- You should consider the following points when
ported, the study is irrelevant or possibly should you read a Results section:
not have been published. This is also not true. If The investigators responded objectively to
the hypotheses are not supported, you should not the results in the discussion of the results.
expect the researcher to bury the work in a file. In the discussion of the results, the investi-
Reviewing and understanding unsupported hypo- gators interpreted the results, with careful
theses is as important for a research consumer as re ection on all aspects of the study that
it is for the researcher. Information obtained from preceded the results.
such studies can often be as useful as data obtained The data presented are summarized. Many
from supported studies. data are generated, but only the critical
nsupported hypotheses can be used to suggest summary numbers for each test are pre-
li itations (weaknesses) of particular aspects of sented. Examples of summarized data are
a study’s design and procedures. Data from such the means and standard deviations of age,
studies may suggest that current modes of practice education, and income. Including all data is
or current theory in an area may not be supported too cumbersome. The Results section can
by research and so should be reexamined and be viewed as a summary.
researched further. Data help generate new knowl- The data are condensed both in the written
edge, as well as prevent knowledge stagnation. text and through the use of tables and
In general, the results are interpreted in a sepa- figures. Tables and figures facilitate the pre-
rate section of the report. Sometimes the Results sentation of large amounts of data.
 CHAPTER 17 Presenting the Findings 387

BOX 17-1
EXAMPLES OF OBJECTIVE STATEMENTS IN THE RESULTS SECTION
“There were statistically significant differences within satisfied). RNs reported less satisfaction (mean = 6.60;
each group in mean of pain-related unpleasantness, SD = 2.42) than RPNs (mean = 7.37; SD = 1.93), and this
between triage and 24 hours post-discharge from the difference was statistically significant (t = 2.38; P <
ED [emergency department] (Exp: t = 4.541, P < 0.0001; 0.02).” (Kaasalainen, Agarwal, Dolovich, Papaioannou,
Ctrl: t = 3.847, P < 0.0001).” (LeMay, Johnston, Brazil, Akhtar-Danesh, 2010, p. 65)
Choiniére, Fortin, Hubert, Fréchette, … Murray, 2010, “Learning-centered characteristics emerged as an
p. 2446) indicator that the teacher valued students, and
“Health board chairperson reported that seniors empowered them in a partnership that increased
accounted for over half of the residents in rural self-awareness through shared assessment.” (Greer,
communities and that mortality, in combination with Pokorny, Clay, Brown, & Steele, 2010, p. 5)
out-migration, had resulted in an overall decrease in the “Some survey respondents indicated that bringing
size of the population.” (Martin-Misener, Reilly, & family members to the bedside during resuscitation
Vollman, 2010, p. 35) helped families to better understand and thus accept
“The participants were moderately satisfied with the the resuscitation team’s decision to discontinue attempts
current system (mean = 7.0; SD = 2.2) as measured on at reviving the patient.” (McClement, Fallis, & Pereira,
a scale of 1 (completely dissatisfied) to 10 (completely 2009, p. 236)

TABLE 17-2
DEMOGRAPHICS OF STUDY PARTICIPANTS
SENIOR NURSE LEADER CEO
M SD M SD

Age 50.4 4.8 51.97 6.01

Years of management experience 20.8 6.7
Years in role 3.7 3.3 5.32 4.10
n % n %
Gender
Male 0 0 37 75.5
Female 63 100 11 22.4
Highest (educational) level
Diploma 2 3.4 1 2.1
Baccalaureate 19 32.2 5 10.4
Masters 34 57.6 40 8
PhD 4 6.8 3.3
Other 2 4.2
From Wong, C. A., Laschinger, H. A., Cummings, G. G., Vincent, L., & O’Connor, P. (2010). Decisional involvement of senior nurse leaders in Canadian
acute care hospitals. Journal of Nursing Management, 18, 122–133.
CEO, chief executive officer; M, mean; SD, standard deviation.

Results for the descriptive and inferential variables. Tables allow researchers to provide a
statistics for each hypothesis or research more visually thorough explanation and discus-
question are presented. No data should be sion of the results. If tables and figures are used,
omitted even if insignificant. they must be concise. Although the text is the
In the study in Appendix B, Wong and col- major mode of communicating the results, the
leagues (2010) developed tables to present the tables and figures serve a supplementary but inde-
results visually. Table 17-2 lists demographic pendent role. The role of tables and figures is to
descriptive results about the study’s participants report results with details that the investigator
Table 17-3 lists the correlations among the study does not enter into the text. This does not mean
388 PART FOUR Processes Related to Research

TABLE 17-3
CORRELATIONS AMONG VARIABLES REGARDING SENIOR NURSE LEADERS
VARIABLE N M (SD) α 1 2 3 4

Timing of involvement 63 1.93 (0.95) 0.84 —

Breadth of content expertise 63 4.08 (0.98) 0.89 0.133 —
Number of decision activities 57 3.88 (1.19) 0.90 0.064 −0.022 —
Decision-making influence 63 4.23 (0.76) 0.83 0.022 0.324* 0.356* —
Quality of decisions 62 3.80 (0.59) 0.68 0.133 0.219† 0.242† 0.195
From Wong, C. A., Laschinger, H. A., Cummings, G. G., Vincent, L., & O’Connor, P. (2010). Decisional involvement of senior nurse leaders in Canadian
acute care hospitals. Journal of Nursing Management, 18, 122–133.
M, mean; SD, standard deviation.
*P < 0.01, one-tailed.
†P < 0.05, one-tailed.

TABLE 17-4 results, as this table does. Descriptions of each

THEMES AND SUBTHEMES IN RESULTS FROM theme in the text of the article would have taken a
A STUDY ON PEER LEARNING lot of space, and the results would have been diffi-
THEMES SUBTHEMES cult to visualize. The table developed by the
Novelty of peer Initial reaction researchers allows you not only to visualize the
learning Adjustment to peer learning concepts quickly but also to assess the results (see
“You’re not alone” Emotional support Table 17-4).
Physical support
Need for support vs.
independence Helpful Hint
Peer-to-peer Positive exchange A well-written “Results” section is systematic,
communication Communication discomforts logical, concise, and drawn from all of the analyzed
Facilitating learning “Bouncing ideas off each other” data. All that is written in the “Results” section should
Sharing tasks be geared to letting the data reflect the testing of the
From Chojecki, P., Lamarre, J., Buck, M., St-Sauveur, I., Eldaoud, N., &
problems and hypotheses. The length of this section
Purden, M. (2010). Perceptions of peer learning approach to pediatric depends on the scope and breadth of the analysis.
clinical education. International Journal of Nursing Education
Scholarship, 7(1), Article 39, 1–12. International Journal of Nursing
Education Scholarship by Berkeley Electronic Press Copyright 2010
Reproduced with permission of BERKELEY ELECTRONIC PRESS in the
Evidence-Informed Practice Tip
format Textbook via Copyright Clearance Center. As you reflect on the results of a study, think
about how the results fit with previous research on the
that the content of tables and figures should not topic and the strength and quality of available evidence
on which to base clinical practice decisions.
be mentioned in the text. The amount of detail
that the author uses in the text to describe the
specific tabular data varies with the needs of the Discussion of the Results
researcher. In the final section of the report, the investigator
A good table meets the following criteria: interprets and discusses the results of the study.
It supplements and economizes the text. In the discussion, a skilled researcher makes the
It has precise titles and headings. data come alive. The researcher interprets and
It does not repeat the text. gives meaning to the numbers in quantitative
Table 17-4 is an example of a table that meets these studies or the concepts in qualitative studies. You
criteria. This table, which is from the article by may ask how the investigator extracted the
Chojecki and associates (2010), lists the study’s meaning that is applied in this section. If the
themes and subthemes. Visualizing the findings of researcher reports properly, the discussion will
a study is easier if a table clearly summarizes the refer to the beginning of the study, in which a
 CHAPTER 17 Presenting the Findings 389

problem statement was identified and indepen- As indicated throughout this text, many impor-
dent and dependent variables were related on the tant pieces in the research puzzle must fit together
basis of a theoretical framework (see Chapter 2) for a study to be evaluated as a well-done project.
and literature review (see Chapter 5). In this Therefore, researchers and reviewers should
section, the researcher discusses the following: accept statistical significance with prudence. Sta-
The supported and the nonsupported hypo- tistically significant findings are not the sole
theses means of establishing the study’s merit. Remem-
The limitations, or weaknesses, of a study ber that accepting statistical significance only
in view of the design and the sample or data- means acceptance that the sample mean is the
collection procedures same as the population mean, which may not be
How the theoretical framework was sup- true (see Chapter 16).
ported Another way to assess whether the findings
Additional or previously unrealized rela- from one study can be generalized is to calculate
tionships suggested by the data a confidence interval. A con dence interval is an
Even if the hypotheses are supported, the estimated range of values that quantifies the uncer-
reviewer should not believe the conclusions to be tainty of a statistic that is, it is the probable value
the final word. Statistical significance is not the range within which a population parameter—for
endpoint of a researcher’s thinking, and low P example, the mean—is expected to lie. The width
values may not be indicative of research break- of the confidence interval gives the researcher
throughs. Thus, statistical significance in a some idea about the uncertainty surrounding the
research study does not always mean that the unknown parameter. A very wide interval may
results of a study are clinically significant. As the indicate that more data should be collected before
body of nursing research grows, so does the pro- definite assertions can be made about the param-
fession’s ability to critically analyze beyond the eter. Confidence intervals are more informative
test of significance and assess a research study’s than the simple results of hypothesis tests (in
applicability to practice. Chapter 20 reviews which a researcher rejects the null hypothesis or
methods for analyzing the usefulness of research fails to reject the null hypothesis) because they
findings. Within the nursing literature, discussion provide a range of plausible values for the unknown
of clinical significance and evidence-informed parameter. For example, Kling and colleagues
practice has also emerged (Goode, 2000 (2009) used confidence intervals around risk ratios
Ingersoll, 2000 Melnyk Fineout-Overholt, to present risk of violence in health care in British
2002). Columbia (Table 17-5).
TABLE 17-5
RISK FACTORS ASSOCIATED WITH WORK-RELATED VIOLENT INCIDENTS AMONG
HEALTH CARE WORKERS IN BRITISH COLUMBIA
ADJUSTED RESULTS
UNADJUSTED RESULTS (MODEL 2)
VARIABLE: NUMBER OF RATE (INCIDENTS/100,000
OCCUPATION VIOLENT INCIDENTS PRODUCTIVE HOURS) RISK RATIO 95% CI RISK RATIO 95% CI

RN 347 (40%) 1.97 6.62 4.63–9.46 6.45 4.37–9.52

LPN 79 (9%) 1.97 10.85 7.23–16.29 8.64 5.56–13.42
Care aide 320 (37%) 3.74 12.55 8.77–17.96 10.05 6.72–15.05
Adapted from Kling, R. N., Yassi, A., Smailes, E., Lovato, C. Y., & Koehoorn, M. (2009). Characterizing violence in health care in British Columbia. Journal
of Advanced Nursing, 65(8), 1655–1663.
CI, confidence interval; LPN, licensed practical nurse; RN, registered nurse.
390 PART FOUR Processes Related to Research

The process used to calculate a confidence sample. Reviewers of research are cautioned not
interval is beyond the scope of this text, but refer- to generalize beyond the population on which a
ences are provided for further explanation study is based. Rarely, if ever, can one study be
(Bluman, 2009). Other aspects of the study, such a recommendation for action. Beware of research
as theory, sample, instrumentation, and methods, studies that may overgeneralize. An example of
should also be considered. making a sweeping generalization is concluding
When the results do not statistically support the that all patients waiting for cardiac bypass can
hypothesis, the researcher refers to the theoretical benefit from preoperative teaching and support
framework and analyzes the earlier thinking when the study sample consisted of white men,
process. The results of nonsupported hypotheses 50 to 70 years of age. Attention must be paid to
do not require that the investigator find fault with the limitations section of an article to note what
each piece of the project. Such a course can become the researchers have considered to affect the gen-
an overdone process. All research has weaknesses. eralizability of their study findings. Generaliza-
This analysis is an attempt to identify the weak- tions that draw conclusions and make inferences
nesses and to suggest the possible or actual prob- within a particular situation and at a particular
lems in the study. At times, the theoretical thinking time are appropriate.
is correct, but the researcher finds problems or An example of an appropriate generalization
limitations that could be attributed to the tools (see is drawn from the study conducted by Luhanga
Chapter 14), the sampling methods (see Chapter and colleagues (2010), who explored and
12), the design (see Chapters 10 and 11), or the described preceptor role support and develop-
analysis (see Chapters 15 and 16). Therefore, ment within the context of a rural and northern
when the hypotheses are not supported, the inves- midsized Canadian community. When discussing
tigator attempts to find facts rather than fault. The the sample in light of the results, Luhanga and
purpose of the discussion, then, is not to show colleagues appropriately noted the following:
humility or one’s technical competence but rather
to enable reviewers to judge the validity of the The main limitation of this study was that the sample
interpretations drawn from the data and the general was limited to one hospital and one university in
worth of the study. northeastern Canada, which limits its generalizability
to other settings and geographical locations.
In the Discussion section, the researcher Additionally, the specificity of the formal education of
summarizes all the aspects of the study and refers the majority of the participants limits transference of
to the beginning to assess whether the findings findings to other settings. (p. 15)
support, extend, or counter the theoretical frame-
work of the study. From this point, you can begin This type of statement is important for reviewers
to think about clinical relevance, the need for of research. It helps guide thinking in terms of a
replication, or the germination of an idea for study’s clinical relevance and suggests areas for
further research study. Finally, you should find further research (see Chapter 20).
the results discussion either in a separate section In a qualitative study, the limitations may be
or subsumed within the Discussion section, and stated differently, as Chiovitti (2008) wrote:
it should include generalizability, applicability,
and recommendations for future research, as well One of the purposes of this qualitative study was to
specify, not generalize, the conditions and actions of
as a summary or a conclusion.
caring. Consequently, as conditions change, it is
enerali ability is the extent to which data expected that the theoretical formulation presented
can be inferred to be representative of similar will also change to meet new conditions, different
phenomena in a population beyond the study’s settings and samples. Therefore, what cannot be
 CHAPTER 17 Presenting the Findings 391

found in the actual data, at the time of the study, is

one of the limitations. (p. 208) Helpful Hint
Trans erability is the extent to which findings It has been said that a good study is one that
raises more questions than it answers. Thus, you should
from one qualitative research study have meaning not view a study’s limitations, generalizations, and
in other studies with similar situations. Authors implications of the findings for practice as an investiga-
must note the issues of a qualitative study to tor’s lack of research skills but as the beginning of the
next step in the research process.
prevent a sweeping transferability of findings,
which would lead to misinterpretations of the
results. In an example of how the limitations in a BOX 17-2
qualitative study can affect transferability, Cohen EXAMPLES OF RESEARCH RECOMMENDATIONS
and Gregory (2009) conducted a qualitative, AND PRACTICE IMPLICATIONS
descriptive study to explore how baccalaureate RESEARCH RECOMMENDATIONS
nursing programs in Canada address the develop- • “Future studies are needed to support and verify
the findings supporting the relationship between
ment of competencies to promote social justice, orthostatic intolerance and bed rest.” (Fox, Sidani,
equity, and the social determinants of health in & Brooks, 2010, p. 194)
their community health clinical courses. Cohen • “In the current study, we included some
demographic variables to enable description of the
and Gregory noted several limitations of the sample, but the emphasis was on psychometrics of
study and that these may affect the transferability the scale. Accordingly, we did not perform detailed
of findings. The focus groups were clinical course analysis of the relationships among those variables
leaders from programs that submitted course and scale responses. Future researchers could
explore those relationships.” (Watson, Oberle, &
syllabi and supporting documentation. Also, close Deutscher, 2008, p. 592)
to 25 of the English language programs did not • “An important consideration for future research is
participate in the study, and no French language the use of total number of cues as a self-
awareness measure. Many participants lost
programs participated. It is possible that some particular cues for hypo- or hyperglycemia over the
course leaders may have been using nontradi- study period.” (Hernandez, Hume, & Rodger, 2008,
tional placement sites that were not represented p. 51)
in the study. One study does not provide all of the PRACTICE IMPLICATIONS
answers, nor should it. The final steps of evalua- • “Our findings suggest that significant predictors of
tion are critical links to the refinement of practice higher levels of distress during laceration repair are
and the generation of future research. Evaluation younger age and parental accompaniment.”
(Sobieraj, Bhatt, LeMay, Rennick, & Johnston,
of research, like evaluation of the nursing process, 2009, p. 79; see Appendix C)
is not the last link in the chain but a connection • “The higher levels of assaultive and psychological
between findings that may serve to improve traumas found in this study suggest the need for a
gender sensitive approach to violence prevention.”
nursing theory and nursing practice. (Samuels-Dennis, Ford-Gilboe, & Ray, 2010, p. 18)
Hall and Irvine (2008) conducted a qualitative • “Learning about the dominant and distinct values,
study to explain how mothers used a community- normative practices, and expectations of the critical
nursing subculture can help newcomers to
based, electronic communication system to build understand differences between the critical care
a local community, request and provide emotional nursing subculture and the nursing culture as a
support, share information and facilitate learning, whole.” (Evans, Bell, Sweeney, Morgan, & Kelly,
and provide validation for mothering experi- 2010, p. 339)
• “Further study identifying the relationships
ences. Hall and Irvine noted that an important between staffing, care-planning practices, and
limitation was the lack of transferability of the quality improvement initiatives are warranted.”
study design and findings to mothers who did not (Wagner, Damianakis, Mafrici, & Robinson-Holt,
2010, p. 322)
have access to electronic communication.
392 PART FOUR Processes Related to Research

The final topic that the investigator integrates implications for nursing practice. This evaluation
into the Discussion section is that of the recom- places the study in the realm of what is known and
mendations. The reco endations are the inves- what needs to be known before being used. Nursing
tigator’s suggestions for the study’s application has grown tremendously over the last century
to practice, theory, and further research. These through the efforts of many nursing researchers
suggestions require the investigator to re ect and scholars. This thought is critical and has been
on the question What contribution to nursing reaffirmed by many nurse researchers in the past
does this study make Box 17-2 provides exam- decade, such as Gortner (2000) and Hinshaw
ples of recommendations for future research and (2000).

APPRAISING THE EVIDENCE

Results and Discussion
The results and the discussion of the results are the For qualitative studies, the richness of the data should
researcher’s opportunity to examine the logic of the be described. The consumer must also have sufficient
hypothesis or question posed, the theoretical framework, detail about the analysis, the coding, the categories of
the methods, and the analysis (see the Critiquing Criteria coding or themes, and the level of coding agreement.
box). This final section requires as much logic, concise- The “Discussion” section should interpret the data,
ness, and specificity as employed in the preceding steps gaps, limitations, and conclusions of the study, as well as
of the research process. provide recommendations for further research. Drawing
For quantitative studies, the research consumer these aspects into the study should give the research
should be able to identify statements on the type of consumer an understanding of the relationship between
analysis that was used and whether the data statistically the findings and the theoretical framework. Statements
supported the hypothesis. These statements should be reflecting the underlying theory are necessary, whether
straightforward and not reflect bias (see Tables 17-3 and or not the hypotheses were supported.
17-4 and Box 18-1, p. 402). Auxiliary data or serendipitous If the findings were not supported, the consumer
findings also may be presented. If such auxiliary findings should—as the researcher did—attempt to identify,
are presented, they should be stated as dispassionately without fault finding, possible methodological problems.
as were the hypothesis data. The statistical test used also Finally, a concise presentation of the study’s generaliz-
should be noted, as well as the numerical value of the ability and the implications of the findings for practice and
data (see Tables 17-1, 17-3, and 17-5). The presentation of research should be evident. The last presentation can
the tests, the numerical values found, and the statements help the research consumer begin to rethink clinical prac-
of support or nonsupport should be clear, concise, and tice, provoke discussion in clinical settings (see Chapter
systematically reported. For illustrative purposes that 20), and find similar studies that may support or refute
facilitate readability, the researchers should present the phenomena being studied to more fully understand
extensive findings in tables rather than in the text. the problem.

CRITIQUING CRITERIA

1. Are the results of each 4. Are the results presented • They have precise titles and
hypothesis presented? objectively? headings.
2. Is the information regarding the 5. If tables or figures are used, do • They do not repeat the text.
results concisely and they meet the following 6. Are the results interpreted in
sequentially presented? standards? light of the hypotheses and
3. Are the tests that were used • They supplement and theoretical framework and all
to analyze the data economize the text. of the other steps that preceded
presented? the results?
 CHAPTER 17 Presenting the Findings 393

CRITIQUING CRITERIA

7. If the data are supported, does and strengths, as well as scope of the findings or beyond
the investigator provide a suggest possible solutions for the findings?
discussion of how the the research area? 11. Are any recommendations for
theoretical framework was 9. Does the researcher discuss the future research stated or
supported? study’s clinical relevance? implied?
8. If the data are not supported, 10. Are any generalizations made, 12. What is the study’s strength of
does the investigator attempt to and, if so, are they within the evidence?
identify the study’s weaknesses

CRITICAL THINKING CHALLENGES used to illustrate and condense data for

presentation.
• Once the results are reported, the researcher
■ Defend or refute the following statement: “All interprets the results. In this presentation, usually
results should be reported and interpreted titled “Discussion,” the consumer should be able to
whether or not they support the hypothesis identify the key topics being discussed. The key
(hypotheses).” topics, which include an interpretation of the
■ What type of knowledge does the researcher results, are the limitations, generalizations,
implications, and recommendations for future
draw on to interpret the results of a study? research.
■ What new knowledge is contributed from the • The researcher draws together the theoretical
research findings? Are they clinically significant framework and makes interpretations based on the
and do they have practice implications? findings and theory in the section on the
■ Do you agree or disagree with the statement that interpretation of the results. Both statistically
supported and unsupported results should be
a good study raises more questions than it interpreted. If the results are not supported, the
answers? Support your view with examples. researcher should discuss the results reflecting on
■ How is it possible for readers of research to the theory, as well as possible problems with the
critique the findings and recommendations of a methods, procedures, design, and analysis.
reported study? How could you use the Internet • The researcher should present the limitations or
weaknesses of the study. This presentation is
for critiquing the findings of a study? important because it affects the study’s
■ Now that nursing students and nurses have generalizability. The generalizations or inferences
access to reports of clinical problems (i.e., about similar findings in other samples also are
critiques of multiple studies available on a presented in light of the findings.
clinical topic) or critiques of individual studies of • The research consumer should be alert for
sweeping claims or overgeneralizations that a
a clinical topic published in Evidence-Based researcher may state. An overextension of the data
Nursing, as well as published meta-analyses and can alert the consumer to possible researcher bias.
meta-syntheses on clinical topics, why is it • The recommendations provide the consumer with
necessary for them to read and critique research suggestions regarding the study’s application to
studies on their own? Justify your response. practice, theory, and future research. These
recommendations furnish the reader with a final
perspective of the utility of the investigation’s
findings in practice.
KEY POINTS
• The analysis of the findings is the final step of a REFERENCES
research investigation. In this section, the research Bluman, A. J. (2009). Elementary statistics: A brief
consumer will find the results presented in a version. New York: McGraw-Hill.
straightforward manner. Chiovitti, R. (2008) Nurses’ meaning of caring with
• All results should be reported whether or not they
patients in acute psychiatric hospital settings: A
support the hypothesis. Tables and figures may be
grounded theory study. International ournal of
394 PART FOUR Processes Related to Research

Nursing Studies, (2), 203-223. doi: dx.doi. health care in British Columbia. ournal of Advanced
org/10.1016/j.ijnurstu.2006.08.018 Nursing, (8), 1655-1663.
Chojecki, P., Lamarre, J., Buck, M., St-Sauveur, I., LeMay, S., Johnston, C., Choini re, M., Fortin, C.,
Eldaoud, N., Purden, M. (2010). Perceptions of Hubert, I., Fr chette, G., . . . Murray, L. (2010). Pain
peer learning approach to pediatric clinical education. management interventions with parents in the
International ournal of Nursing Education Scholar- emergency department: A randomized trial. ournal
ship, (1), Article 39, 1-12. of Advanced Nursing, , 2442-2449.
Cohen, B. E., Gregory, D. (2009). Community health Luhanga, F. L., Dickeson P., Mossey, S. D. (2010).
clinical education in Canada: Part 2—developing Preceptor preparation: An investment in the future
competencies to address social justice, equity, and the generation of nurses. International ournal of
social determinants of health. International ournal of Nursing Education, (1), Article 38, 1-15.
Nursing Education Scholarship, (1), Article 2. Martin-Misener, R., Reilly, S. M., Vollman, A. R.
Evans, J., Bell, J. L., Sweeney, A. E., Morgan, J. I., (2010). Defining the role of primary health care nurse
Kelly, H. M. (2010). Confidence in critical care practitioners in rural Nova Scotia. Canadian ournal
nursing. Nursing Science uarterly, , 334-340. of Nursing Research, (2), 30-47.
Fox, M. T., Sidani, S., Brooks, D. (2010). The McClement, S. E., Fallis, W. M., Pereira, A. (2009).
relationship between bed rest and sitting orthostatic Family presence during resuscitation: Canadian
intolerance in adults residing in chronic care critical care nurses’ perspectives. ournal of Nursing
facilities. ournal of Nursing and ealthcare of Scholarship, , 233-240.
Chronic Illness, (3), 187-196. doi: Melnyk, B. M., Fineout-Overholt, E. (2002). Key
10.1111/j/1752-9824.2010.01058 steps in evidence based practice: Asking compelling
Goode, C. J. (2000). What constitutes the evidence in questions and searching for the best evidence.
evidence-based practice. Applied Nursing Research, Pediatric Nursing, , 262-263, 266.
, 222-225. Samuels-Dennis, J., Ford-Gilboe, M., Ray, S. (2010).
Gortner, S. (2000). Knowledge development in nursing: Single mother’s adverse and traumatic experiences
Our historical roots and future opportunities. Nursing and post-traumatic stress symptoms. ournal of
utloo , , 60-67. amily Violence, (1), 9-20.
Greer, A. G., Pokorny, M., Clay, M. C., Brown, S., Sobieraj, G., Bhatt, M., LeMay, S., Rennick, J.,
Steele, L. L. (2010). Learner-centered characteristics Johnston, C. (2009). The effect of music on parental
of nurse educators. International ournal of Nursing participation during pediatric laceration repair.
Education Scholarship, (1), Article 6. Canadian ournal of Nursing Research, (4), 68-82.
Hall, W., Irvine, V. (2008). E-communication among Wagner, L. M., Damianakis, T., Mafrici, N.,
mothers of infants and toddlers in a community-based Robinson-Holt, K. (2010). Falls communication
cohort: A content analysis. ournal of Advanced patterns among nursing staff working in long-term
Nursing, (1), 175-183. care settings. Clinical Nursing Research, (3),
Hernandez, C. A., Hume, M. R., Rodger, N. W. 311-326.
(2008). Evaluation of a self-awareness intervention Watson, L., Oberle, K., Deutscher, D. (2008).
for adults with type 1 diabetes and hypoglycemia Development and psychometric testing of the Nurses’
awareness. Canadian ournal of Nursing Research, Attitudes Toward Obesity and Obese Patients
(3), 38-56. (NATOOPS) scale. Research in Nursing and ealth,
Hinshaw, A. S. (2000). Nursing knowledge for the 21st , 586-593.
century: Opportunities and challenges. ournal of Wong, C. A., Laschinger, H. A., Cummings, G. G.,
Nursing Scholarship, , 117-123. Vincent, L., O’Connor, P. (2010). Decisional
Ingersoll, G. L. (2000). Evidence-based nursing: What it involvement of senior nurse leaders in Canadian
is and what it isn’t. Nursing utloo , , 151. acute care hospitals. ournal of Nursing Management,
Kaasalainen, S., Agarwal, G., Dolovich, L., , 122-133.
Papaioannou, A., Brazil, K., Akhtar-Danesh, N.
(2010), Nurses’ perceptions of and satisfaction with
FOR FURTHER STUDY
the medication administration system in long-term- Go to Evolve at http://evolve.elsevier.com/
care homes. Canadian ournal of Nursing Research, Canada/LoBiondo/Research for Audio Glossary, how-to
(4), 58-79. instructions for Writing Proposals for Funding, and
Kling, R. N., Yassi, A., Smailes, E., Lovato, C. Y., additional research articles for practice in reviewing and
Koehoorn, M. (2009). Characterizing violence in critiquing.
395

PA RT FI VE
PART ONE Part Title

Critiquing Research

18 Critiquing Qualitative
Research
19 Critiquing Quantitative
Research
RESEARCH VIGNETTE engaged in a participatory action
study to determine the sociocul-
Evolution in a Program of Research tural in uences on the meanings
of and responses to woman abuse.
Focusing on Family Violence, This participatory study helped
Caregiving, and Women’s Health enhance the visibility of domestic
violence and resulted in commu-
nity actions to help women and
change common understandings
Judith Wuest RN, BScN, MN, PhD disease. Throughout this research, or attitudes that allowed abuse to
Professor Emerita, Faculty of Nursing I became very conscious that continue.
University of New Brunswick
family caregivers were largely This work took place over
Fredericton, New Brunswick
women who carried the brunt of several years, when I was also
Programs of research evolve in caring, sometimes at great per- spending summers in doctoral
unpredictable ways, depending on sonal cost, including disruptions work at Wayne State niversity.
the findings of previous studies, of family relationships. I entered There I met Jacquelyn Campbell,
opportunities for collaboration, the summer doctoral program at an international expert in domes-
and successes and failures in Wayne State niversity with this tic violence, who introduced me to
applications for funding. My in mind. a wide range of feminist thought
current program of research At this same time, my col- and emphasized IPV as a nursing
began with my first research study, league Marilyn Merritt-Gray and I and health issue. In my doctoral
the thesis of my master’s degree began a grounded theory study of research, I combined feminist
in nursing: a grounded theory women who left abusive partners. theory and grounded theory and
study of family caregiving for Our goal was to fill the gap in our studied women’s caring across the
children with chronic middle theoretical knowledge about the life span, developing the theory of
ear disease. I had a generous process of leaving. As a commu- precarious ordering An impor-
mentor in my supervisor, Phyllis nity mental health nurse, Marilyn tant finding in this study was that
Noerager Stern, who was an felt the need for a framework to women who were caring for
expert in grounded theory. This guide her practice and to help family members who in the past
first research experience was women understand what stage in had abused them, or with whom
exciting and interesting because life they were in, where they were they had very strained relation-
the research approach enabled me headed, and how to get there. ships, had more difficulty gaining
to learn about family issues from Our initial findings, funded by control of their health and had
the perspective of the families a local grant, allowed us to obtain poorer health outcomes. My
themselves, and developing a a larger local grant to develop the two areas of interest, IPV and
beginning theory was useful for theory of reclaiming self We women’s caring, had begun to
both nurses and families. became affiliated with the Muriel intersect
As a new faculty member who Mc ueen Fergusson Centre for After graduation, I grappled
was expected to develop a program Family Violence Research and with the challenge of getting
of research and to obtain funding began to collaborate with profes- national funding. By combining
locally, I built on my thesis sionals and survivors of abusive my and Marilyn Merritt-Gray’s
research by conducting a similar relationships, who were concerned interest in IPV and my colleague
study with Aboriginal families. that, despite their best efforts, the Marilyn Ford-Gilboe’s interest in
This research was followed by a incidence of intimate partner vio- single motherhood, I began a
small study of family caregiving lence (IPV) in their rural area program of research focusing on
for relatives with Alzheimer’s remained high. Together, we single-parent families after they

396
left abusive relationships. This intervention for women in the collection to capture changes in
work has been well funded by the early years after leaving (iHEAL). health and health promotion
Medical Research Council, the Currently, feasibility and efficacy over time.
National Health Research and studies of the iHEAL are under The final portion of my program
Development Program, and, cur- way in both New Brunswick and of research also stems from the
rently, the Canadian Institutes of Ontario. The New Brunswick study of women who left abusive
Health Research (CIHR). Our first study is being conducted in part- partners. Although the theory of
study, built on the study of leaving, nership with the provincial domes- reclaiming self was useful for these
focused on health promotion tic violence outreach program, women, it did not help us under-
in single-parent families after the New Brunswick Women’s stand the experiences of some
leaving, on the basis of a feminist Issues Branch, and with the women who remain with their part-
grounded theory perspective. We New Brunswick Department of ners and whose relationships
developed the theory of strength- Health and Liberty Lane Inc. it become nonviolent. With funding
ening capacity to limit intrusion is funded under the Partnerships from the Social Sciences and
and noted that women in the study for Health System Improvement Humanities Research Council
had many health issues related to program by CIHR and the New (SSHRC), our research team con-
IPV that had persisted for as long Brunswick Health Research Foun- ducted a grounded theory study of
as 20 years after they left abusive dation. Our policy forums, held in how women achieve nonviolence,
partners. However, little was New Brunswick and Ontario to and we developed a theory: shifting
known about the long-term health share findings from the WHES the pattern of abusive control
consequences of IPV. with policymakers and front-line Although violence cessation was
This finding led to the develop- workers, were key to forming the important for women staying in
ment of a CIHR New Emerging partnerships necessary to apply relationships, equally important
Team focusing on the long-term for these funds. for some women was men reinvest-
health effects of IPV. With funding Similarly, my research on care- ing in the relationship. We realized
from CIHR operating grants, we giving has evolved from a qualita- that to fully understand how rela-
recruited 309 women who had tive to a quantitative focus. With tionships shift when they become
been out of abusive relationships further CIHR funding, Marilyn nonviolent, we needed to under-
no longer than 3 years to partici- Hodgins and I assembled a care- stand men’s perspective on how
pate in structured interviews and giving team to develop an instru- their relationships changed when
health assessments annually for 5 ment measuring selected concepts they ceased to be abusive toward
years. This study, the Women’s in the grounded theory of precari- their partners. This study was also
Health Effects Study (WHES), ous ordering, with the goal of funded by SSHRC, and the find-
demonstrated that physical and testing how past relationships, ings will soon be available.
mental health problems persist obligation, health, and health pro- What is most interesting to me
long after leaving an abusive motion in women caregivers of is how my program of research
partner, are linked to patterns of adult family members are related. has evolved methodologically and
cumulative lifetime abuse, and We surveyed more than 250 substantively in response to each
lead to higher use of health service female caregivers and found that study’s findings and the collabora-
with little symptom relief. Four past relationships and obligation tive opportunities that have arisen.
nurse researchers—Marilyn Ford- accounted for significant variance I have an affinity for grounded
Gilboe, Marilyn Merritt-Gray, in health outcomes. With funding theory research and probably
Colleen Varcoe, and I—have used from the Alzheimer Society of approach most research with the
the findings of the WHES, com- Canada, we completed a second mindset of a grounded theorist.
bined with the intrusion theory we study that replicated the first However, each study raises essen-
developed, to design a health study but included repeated data tial questions that cannot always

397
be answered by the method methodological versatility needed for nursing practice to respond
that the researcher likes best. Res- for research to evolve from induc- to the enormous challenges of
earch partnerships with experts in tive to deductive to interventional. the effects of family violence
diverse approaches are essential These partnerships are needed to on individuals, families, and
for building teams with the provide the necessary knowledge communities. ■

398
C H A PTER 1 8

Critiquing Qualitative Research

Helen J. Streubert | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Identify the influence of stylistic considerations on the presentation of a qualitative research report.
• Identify the criteria for critiquing a qualitative research report.
• Evaluate the strengths and weaknesses of a qualitative research report.
• Describe the applicability of the findings of a qualitative research report.
• Construct a critique of a qualitative research report.

KEY TERMS
auditability phenomena theoretical sampling
credibility saturation trustworthiness
fittingness

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

399
400 PART FIVE Critiquing Research

NURSE SCIENTISTS CONTRIBUTE SIGNIFICANTLY TO THE

body of health care research. These contributions STYLISTIC CONSIDERATIONS
are evident in nursing, medical, health care, ualitative research differs from quantitative
and business journals. Nurse researchers are research in some very fundamental ways. In qual-
partnering at an ever-increasing rate with other itative research, investigators seek to discover
health care professionals to develop, implement, and understand concepts, phenomena, or cultures.
and evaluate a variety of evidence-informed Creswell (2009) stated that qualitative research
interventions to improve patient outcomes. The is exploratory and researchers use it to explore a
methods used to develop evidence-informed topic when the variables and theory base are
practice include quantitative, qualitative, and unknown (p. 98). Jackson and associates (2007)
mixed research approaches. In addition to the noted that the primary focus of qualitative
increase in the number of research studies research is to understand human beings’ experi-
and publications, there is also a significant ences in a humanistic and interpretive way
record of externally funded research by nurses (p. 21). In a qualitative study, you should not
that adds to the credibility—accuracy, validity, expect to find hypotheses theoretical frame-
and soundness—of the work. The willingness works dependent and independent variables
of private and publicly funded organizations large, random samples complex statistical proce-
to invest in nursing research attests to its dures scaled instruments or definitive conclu-
quality and potential for affecting health care sions about how to use the findings. Because the
outcomes of individuals, families, groups, and purpose of qualitative research is to describe or
communities. explain concepts, phenomena, or cultures, the
Because the expansion in nursing research and report is generally written in a way that allows
related interventions affect increasing numbers of the researcher to convey the full meaning and
patients, the evaluation of this research is crucial. complexity of the phenomena or cultures being
The focus of this chapter is on assessing the studied. This narrative includes subjective com-
quality of qualitative research studies. This ments that are intended to depict the depth and
chapter demonstrates a set of criteria that can be richness of the phenomena under study. The goal
used to determine the quality of a qualitative of the qualitative research report is to describe in
research report. Nurses must fully understand as much detail as possible the insider’s, or
how to assess the value of qualitative research, emic, view of the phenomenon being studied. The
particularly in view of the requirement that emic view is the view of the person experiencing
nursing practice be evidence based. According to the phenomenon that re ects his or her culture,
Straus and colleagues (2011), evidence-informed values, beliefs, and experiences. In reports of
practice requires the integration of critical qualitative research, the investigator hopes to
appraisal with the clinical expertise and the convey an understanding of what it is like to
patient’s unique biology, values, and circum- experience a particular phenomenon or to be part
stances (p. 3). of a specific culture.
As a framework for understanding the appraisal One of the most effective ways to convey
of qualitative research as a basis for evidence- the emic view is to use quotations that re ect the
informed practice, a published research report, phenomenon as experienced. For this reason, the
as well as critiquing criteria, is presented. qualitative research report has a more conversa-
The criteria are then used to demonstrate tional tone than a quantitative report. In addition,
the process of appraising a qualitative research data are frequently articulated in concepts or
report. phrases, which the researcher calls themes (see
 CHAPTER 18 Critiquing Qualitative Research 401

Chapter 8), as a way of describing large quantities requirements within the page limit imposed by
of data in a condensed format. the journal of interest.
The richness of the narrative provided in a Nursing journals do not generally offer their
qualitative research study cannot be shared in its reviewers specific guidelines for evaluating qual-
entirety in a journal publication. Page require- itative and quantitative research reports. The
ments imposed by journals frequently limit editors try to ensure that reviewers are knowl-
research reports to 15 pages. Despite this con- edgeable in the method and subject matter of the
straint, investigators in qualitative research need study. This determination is often made, however,
to illustrate the richness of the data and convey on the basis of the reviewer’s self-identified area
to the audience the relationship between the of interest. Research reports are often evaluated
themes identified and the quotes shared. This is in accordance with the ideas or philosophical
essential in order to document the rigour of the viewpoints held by the reviewer. The reviewer
research, which is called trust orthiness in may have strong feelings about particular types
a qualitative research study. Of importance is of qualitative or quantitative research methods.
that conveying the depth and richness of the Therefore, it is important to clearly state the qual-
findings of a qualitative study is challenging itative approach used and, if appropriate, its phil-
in a published research report. However, regard- osophical base.
less of the page limit, Jackson and associates The principles for evaluating different qualita-
(2007) suggested that it is the researcher’s tive research approaches are very similar funda-
responsibility to ensure objectivity (use of facts mentally. Research consumers are concerned with
without distortion by personal feeling or bias), the plausibility and trustworthiness of the research-
ethical diligence (see Chapter 6), and rigour er’s account of the research and its relevance
regardless of the method selected to conduct to current or future theory and practice, or both
the study. Fully sharing the depth and richness (Horsburgh, 2003). Box 18-1 provides general
of the data will also help practitioners decide on guidelines for evaluating qualitative research, and
the appropriateness of applying the findings to Box 18-2 provides guidelines for evaluating
their practice. grounded theory. For information on specific
Some journals, such as ualitative ealth guidelines for the evaluation of phenomenology,
Research, are committed to publication of more ethnography, grounded theory, and historical
lengthy reports. Guidelines for publication of and action research, see Streubert and Carpenter
research reports are generally listed in each (2011). You should review Chapters 7 and 8 in this
nursing journal or are available from the journal text before completing this chapter.
editor. Of importance is that criteria for publica-
tion of research reports are not based on a specific APPLICATION OF QUALITATIVE
type of research method (i.e., quantitative or qual- RESEARCH FINDINGS IN PRACTICE
itative). The primary goal of journal editors is to As already stated, one of the purposes of qualita-
provide their readers with high-quality, informa- tive research is to describe, understand, or explain
tive, timely, and interesting articles. To meet this phenomena. Pheno ena are events perceived by
goal, regardless of the type of research report, the senses and may be experienced emotionally,
editors prefer to publish manuscripts that have such as pain and losing a loved one. In addition
scientific merit, present new knowledge, support to clarifying phenomena, qualitative research
the current state of the science, and engage can give voice to people who have been disen-
their readers. As stated earlier, the challenge in franchised and whose experiences would have
qualitative research is to meet these editorial otherwise not been documented (Barbour
402 PART FIVE Critiquing Research

BOX 18-1
CRITIQUING GUIDELINES FOR QUALITATIVE RESEARCH
STATEMENT OF THE PHENOMENON OF INTEREST 4. Does the researcher address the credibility,
(CHAPTER 4) auditability, and fittingness of the data? (See Chapter
1. What is the phenomenon of interest, and is it clearly 14 for a complete discussion.)
stated for the reader?
2. What is the justification for using a qualitative Credibility
method? • Do the participants recognize the experience as
3. What are the philosophical underpinnings of the their own?
research method? • Has adequate time been allowed to fully
understand the phenomenon?
PURPOSE (CHAPTER 4)
1. What is the purpose of the study? Auditability
2. What is the projected significance of the work for • Can the reader follow the researcher’s thinking?
nursing? • Does the researcher document the research
process?
METHOD (CHAPTER 8)
1. Is the method used to collect the data compatible Fittingness
with the purpose of the research? • Are the findings applicable outside of the study
2. Is the method adequate for addressing the situation?
phenomenon of interest? • Are the results meaningful to individuals not
3. If a particular approach is used to guide the inquiry, involved in the research?
does the researcher complete the study according to • Is the strategy used for analysis compatible with
the processes described? the purpose of the study?

SAMPLING (CHAPTER 12) FINDINGS (CHAPTERS 7 AND 17)

1. What type of sampling is used? Is it appropriate for 1. Are the findings presented within a context?
the particular method? 2. Is the reader able to apprehend the essence of the
2. Are the participants who were chosen appropriate for experience from the report of the findings?
informing the research? 3. Do the researcher’s conceptualizations accurately
reflect the data?
DATA COLLECTION (CHAPTERS 7, 8, AND 13) 4. Does the researcher place the report in the context of
1. Are the data to be collected focused on human what is already known about the phenomenon? Was
experience? the existing literature on the topic related to the
2. Does the researcher describe the data-collection findings?
strategies (e.g., interview, observation, field
notes)? CONCLUSION, IMPLICATIONS, AND
3. What are the procedures for collecting the data? RECOMMENDATIONS (CHAPTER 17)
4. Is the protection of human participants addressed? 1. Do the conclusions, implications, and
5. Is saturation of the data described? recommendations give the reader a context in which
to use the findings?
DATA ANALYSIS (CHAPTERS 14 AND 15) 2. Do the conclusions reflect the study findings?
1. What strategies are used to analyze the data? 3. What are the recommendations for future study? Do
2. Has the researcher reported the data truthfully? they reflect the findings?
3. Does the researcher describe the steps used for the 4. How has the researcher made explicit the significance
data analysis? of the study for nursing theory, research, or practice?
 CHAPTER 18 Critiquing Qualitative Research 403

BOX 18-2
CRITIQUING GUIDELINES FOR RESEARCH CONDUCTED WITH THE GROUNDED THEORY METHOD
FOCUS/TOPIC (CHAPTERS 4 AND 8) 3. How did theoretical formulations guide the data
1. What is the focus or the topic of the study? What is it collection?
that the researcher is studying? Is the topic
researchable? Is it focused enough to be meaningful DATA ANALYSIS (CHAPTERS 8 AND 15)
but not too limited so as to be trivial? 1. Does the researcher describe the strategies used to
2. Has the researcher identified why the phenomenon analyze the data?
requires a qualitative format? What is the rationale • Has the theoretical construction been checked
for selecting the grounded theory approach as the against the participants’ descriptions of the
qualitative approach for the investigation? phenomenon?
• Are the researcher’s views and insights about the
PURPOSE (CHAPTER 4) phenomenon articulated?
1. Has the researcher made explicit the purpose for • Has each category that emerged in the theory
conducting the research? been described previously in the literature?
2. How does the researcher address the credibility,
SIGNIFICANCE (CHAPTER 4) auditability, and fittingness of the data?
3. Does the researcher clearly describe how and why
1. Has the researcher described the projected the core category was selected?
significance of the work for nursing?
2. What is the relevance of the study to what is already EMPIRICAL GROUNDING OF THE STUDY FINDINGS
known about the topic? (CHAPTERS 8 AND 15)
1. Are the concepts grounded in the data?
METHOD (CHAPTER 8)
2. How are the concepts systematically related?
1. In view of the topic of study and the researcher’s 3. Are conceptual linkages described, and are the
stated purpose, how does grounded theory categories well developed? Do they have conceptual
methodology help to achieve the stated purpose? density (in-depth conceptual discussion)?
2. Is the method adequate for addressing the research 4. Are the theoretical findings significant? If so, to what
topic? extent?
3. What approach is used to guide the inquiry? Does 5. Were the data-collection strategies comprehensive,
the researcher complete the study according to the and were analytical interpretations conceptual and
processes described? broad?
6. Is variation in the interpretations sufficient to allow
SAMPLING (CHAPTERS 8 AND 12) for applicability in a variety of contexts related to the
1. Does the researcher describe the selection process phenomenon investigated?
and protection of human participants?
2. What major categories emerged? CONCLUSIONS, IMPLICATIONS, AND
3. What were some of the events, incidents, or actions RECOMMENDATIONS (CHAPTERS 8 AND 17)
on which these major categories were based? 1. How do the conclusions, implications, and
4. What categories led to theoretical sampling? recommendations provide context in which to use
the findings?
DATA GENERATION (CHAPTERS 8 AND 13) 2. Are the conclusions drawn from the study
1. Does the researcher describe the data-collection appropriate? Explain.
strategies? 3. What are the recommendations for future research?
2. Have participants been allowed to guide the direction 4. Are the recommendations, conclusions, and
of the inquiry? implications clearly related to the findings?
From Streubert, H. J., & Carpenter, D. R. (2011). Qualitative research in nursing: Advancing the humanistic imperative. Philadelphia: Wolters Kluwer.
Adapted from Chiovitti, R., & Prian, N. (2003). Rigour and grounded theory research. Journal of Advanced Nursing Practice, 44(4), 427-435; and
from Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage.
404 PART FIVE Critiquing Research

Barbour, 2003 Schepner-Hughes, 1992). In qual-

Evidence-Informed Practice Tip
itative inquiry, unlike quantitative research, pre-
Nurses using qualitative research findings should
diction and control of phenomena are not the aim. ask whether the evidence provided in the study
Therefore, qualitative results are applied differ- enhances their understanding of particular patient care
ently than more traditional quantitative research situations.
findings. As Barbour and Barbour (2003) stated,
. . . rather than seeking to import and impose It is important to view research findings within
templates and methods devised for another purpose, their context, whether quantitative or qualitative.
qualitative researchers and reviewers should look . . . For instance, a study of survivorship among
for inspiration from their own modes of working and
patients with cancer should not be viewed as
collaborating and seek to incorporate these, forging
new and creative solutions to perennial problems, applicable to survivors of another illness situa-
rather than hoping that these will simply disappear in tion, such as an epidemic. The findings must be
the face of application of pre-existing sets of used within their given context, or additional
procedures. (p. 185) studies must be conducted to validate the appli-
Therefore, findings may be applicable only in cability of the findings across situations and
certain circumstances. As Lincoln and Guba patient populations. This is true in qualitative
(1985) stated, the trouble with generalizations is research as well. Nurses who wish to use the find-
that they don’t apply to particulars (p. 110). ings of qualitative research in their practice must
Thus, in qualitative research, if the investigator validate them, through thorough examination and
studies the pain experience of individuals under- synthesis of the literature on the topic, through
going bone marrow biopsy, for example, the find- their own observations, or through interaction
ings will be applicable only to individuals who with groups similar to the study participants, to
are similar to those in the study. determine whether the findings accurately re ect
In another example, a study of depression in the experience.
children with chronic disease should not be In qualitative outcome analysis ( OA),
viewed as having direct application to adults suf- researchers can use the findings of a qualitative
fering from chronic disease. The findings must be study to develop interventions and then to test
used within a specific context, or additional them with selected patients. Morse and colleagues
studies must be conducted to validate the appli- (2000) described OA as a systematic means to
cability of the findings across contexts. Hence, confirm the applicability of clinical strategies
nurses who wish to use the findings of qualitative developed from a single qualitative project, to
research in their practices must first validate extend the repertoire of clinical interventions, and
them, either through their own observations or evaluate clinical outcomes (p. 125). Application
through interaction with groups similar to the of knowledge discovered during OA adds to
original study participants, to determine whether clinicians’ understanding of clinical phenomena
the findings accurately re ect wider experiences. because they can select interventions that are
In spite of the limitations associated with qualita- based on the patient’s expressed experience of a
tive research findings, Glesne (2011) stated that particular clinical phenomenon. OA is consid-
qualitative research findings can be used to create ered a form of evaluation research and, as such,
solutions to practical problems. ualitative has the potential to add to the literature on evi-
research also has the ability to contribute to the dence-informed practice at either level V or level
evidence-informed practice literature (Cesario, VI, depending on how the study was designed.
Morin, Santa-Donato, 2002 Gibson Martin, Another use of qualitative research findings is
2003). to initiate examination of important concepts in
 CHAPTER 18 Critiquing Qualitative Research 405

nursing practice, education, or administration.

Evidence-Informed Practice Tip
Caring, for example, is considered a significant
Qualitative research studies can be used to guide
concept in nursing therefore, studying its multi- practice when they are applied within a context. The
ple dimensions is important. sing a qualitative nurse should ask the following question: “Does this
approach, Wilkin and Slevin (2004) explored the study provide me with a direction for caring for a par-
ticular patient group?”
meaning of caring for intensive care nurses.
Wilkin and Slevin posited that although caring
had been studied extensively, little research had Finally, qualitative research can be used to
been conducted in the highly technological area discover information about phenomena of inter-
of critical care. The authors identified caring in est that can lead to instrument development.
an intensive care setting as a process of compe- sually, qualitative methods are used to direct
tent physical and technical action imbued with the development of structured research instru-
affective skills and confirmed that to care is ments as part of a larger empirical research
human and the capacity to care is affirmed and project. Instrument development from qualitative
actualised in caring for the critically ill patient research studies is useful to practising nurses
and their relatives (p. 50). Wilkin and Slevin’s because it is grounded in the reality of human
study adds to the existing body of knowledge on experience with a particular phenomenon. For
caring and extends the current state of nursing example, after an initial qualitative exploration
science because it was an examination of a spe- of the phenomenon, the researcher may develop
cific area of nursing practice and the experience a survey to collect the data related to specific
of caring by critical care nurses. variables.
CRITIQUING A QUALITATIVE RESEARCH STUDY
The study “Immunizing Children Who Fear and Resist Needles: Is it a Problem for Nurses?” by
Mary Ives and Sherri Melrose (2010) is critiqued here. The article is presented in its entirety and
CRITIQUE 1 is followed by the critique on pp. 416–418. (From Nursing Forum, 45[1], 29–39, reprinted with
permission. © 2010 Wiley Periodicals, Inc.)

Immunizing Children Who Fear and Resist Needles:

Is It a Problem for Nurses?
Mary Ives, RN, MHS
Chilliwack Health Unit INTRODUCTION
Chilliwack, British Columbia
This article describes findings from a qualitative
Sherri Melrose, RN, PhD research project that investigated the experiences,
Assistant Professor re ections, and feelings of public health nurses
Centre for Nursing and Health Studies who immunize fearful and needle-resistant chil-
Athabasca University
Athabasca, Alberta
dren. While the main purpose of the study was to
explore nurses’ ideas about the experience of
immunizing children who are frightened of
BACKGROUND Despite increasing evidence that immuni- needles, a secondary purpose was to consider
zation procedures can be stressful for children, little is
approaches that are helpful in decreasing nurse
known about what the experience of immunizing
frightened and needle-resistant children can be like for stress. The research was guided by the question:
nurses. hat is it about immuni ing children ho strongly
resist needle injection that is a problem for public
health nurses
METHOD This article presents findings from a qualitative
research project designed to explore public health nurses’
feelings toward immunizing needle-resistant children. A
LITERATURE REVIEW
constructivist theoretical perspective and an action research
approach framed the study. Data sources included two A literature review revealed that a significant
survey questions and audio-recorded transcribed data from number of children and adults are frightened of
three focus groups. Participants included 35 public health
procedures involving needle injections. Consider-
nurses from five different health units in one Canadian
province. The data were analyzed for themes and were able research has been undertaken to investigate
confirmed with participants through ongoing member adult responses that are both non-helpful and
checking. helpful in easing children’s distress during these
procedures (Cohen, Manimala, Blount, 2000
Duff, 2003 Frank, Blount, Smith, Manimala,
RESULTS The following four overarching themes were
identified and are used to explain and describe significant Martin, 1995 French, Painter, Coury, 1994
features of the immunization experience that were stressful Manimala Blount, 2000 Milgrom, Coldwell,
and problematic for nurses: (a) nurses experience stress Getz, Weinstein, Ramsay, 1997 Schecter,
when immunizing children who fear and resist needle injec- Bernstein, Beck, Hart, Scherzer, 1991 Smalley,
tion; (b) the strength of child resistance and some adult 1999). Yet, few resources are available to help
behavior creates an ethical dilemma for nurses; (c) some
adult responses make immunizing difficult and unsafe; and
nurses understand their own responses or to cope
(d) resources to help nurses cope with these situations are with their feelings of stress. Ives (2007) empha-
inconsistent. sized how healthcare agencies can begin to

406
 CHAPTER 18 Critiquing Qualitative Research 407

address the problem by creating a culture of

empathy and respect and by outlining clear poli-
NON-HELPFUL AND HELPFUL
cies on the use of force during immunizations.
ADULT RESPONSES
There is a gap, however, in our understanding Parent and caregiver responses, particularly
of how nurses themselves perceive the experience anxiety-related behaviors, in uence how children
of immunizing frightened and resistive children. can reduce stress, gain control, and cope with
immunization. Some adult responses have been
found to be non-helpful. Parents or caregivers
FEAR OF NEEDLES who overly reassured, overly empathized, apolo-
Literature from the fields of psychology, nursing, gized, criticized, or gave children control of the
pharmacology, medicine, and dentistry reveals procedure at the beginning increased children’s
fear of needles as one of children’s greatest fears, stress (Cohen et al., 2000 Frank et al., 1995).
with claims that up to 93 of some groups Further, parents and caregivers who criticized or
of children experience serious immunization- asked the child to indicate readiness to receive the
associated stress (Bowen Dammeyer, 1999 needle also increased children’s stress (Devine
Gaskell, Binns, Heyhoe, Jackson, 2005 Jacob- et al., 2004). Children coped best when their
son et al., 2001 Martin, Ramsay, Whitney, Fiset, mothers were present but watched only and
Weinstein, 1994 Peretz Efrat, 2000 Polillio remained minimally involved. Most children
Kiley, 1997 Smalley, 1999 man, Chambers, found the presence of their parents during a
McGrath, Kisely, 2006). Research re ects that needle procedure to be helpful (Duff, 2003
as many as 10 of adults experience needle O’Laughlin Ridley-Johnson, 1995).
phobia (Bowen Dammeyer Hamilton, 1995 Distraction strategies were consistently identi-
Polillio Kiley Smalley). Clearly, nurses can fied as helpful for short-lived pain (Duff, 2003
expect to encounter both children who are fright- Gaskell et al., 2005 Lawton Rose, 2003
ened and resistant to needles as well as parents or Manimala Blount, 2000 Sparks, 2001). With
caregivers who are also fearful. infants, playing with an object, sucking, belly-to-
In his seminal work exploring needle phobia, belly contact, and nonprocedural talk were helpful
Hamilton (1995) hypothesized that needle phobia (Blount, Devine, Cheng, Simons, Hayutin,
is learned as well as inherited. He noted how 2008). Similarly, with infants, adult verbaliza-
negative experiences associated with immuniza- tions associated with better pain outcomes
tion, laboratory work, dental visits, and other reduced crying (Bustos, Jaaniste, Salmon,
medical procedures can condition children and Champion, 2008). With children ages 4–6 years,
even those who witness the events toward becom- watching cartoons and being coached to attend to
ing fearful of needles. Physical restraint and the movie helped (Cohen, Blount, Panopoulos,
verbal abuse by healthcare personnel during chil- 1997). With children ages 5–18 years, bubbles,
dren’s medical procedures can lead to life-long books, music, virtual reality glasses, or handheld
fears of situations associated with needles, such video games helped (Windich-Biermeier, Sjoberg,
as physicians, nurses, examination rooms, and Dale, Eshelman, Guzzetta, 2007). With most
even antiseptic smells (Hamilton). Later, Duff children, preparing ahead (Duff), offering limited
(2003) argued that needle fear centers on antici- choices (Ellis, Sharp, Newhook, Cohen, 2004),
patory and procedural stress and advocated for and giving permission to cry (Cohen et al., 2000)
the inclusion of psychological approaches to help reduced stress. Deferring the procedure or refer-
children actively gain a sense of control over their ral to an alternate source such as play therapy
reactions. helped to avoid con ict and coercion (Duff
408 PART FIVE Critiquing Research

French et al., 1994 Milgrom et al., 1997 McTaggart, 1988, 1990 Stringer Genat,
Smalley, 1999). Distinguishing among adult 2004). Action research is a re ective, spiral
responses that are helpful and those that are non- process where nurses use research techniques to
helpful offers important guidance to nurses when examine their own practice carefully, systemati-
they work with children who resist needles. cally, and with the intention of applying their
However, responses to nurse stress are not as findings directly to their own and other nurses’
clearly defined. everyday practice. Kemmis and McTaggart
offered the seminal explanation that action
NURSE STRESS research is a deliberate, solution-oriented investi-
Stress can be experienced when demands exceed gation that is group or personally owned and con-
the personal and social resources an individual is ducted. It is characterized by spiralling cycles of
able to mobilize (Lazarus Folkman, 1984). problem identification, systematic data collec-
While it is beyond the scope of this article to tion, re ection, analysis, data-driven action taken,
present a detailed literature review of nurse stress, and, finally, problem redefinition. The linking of
a snapshot of current work in the area reveals the terms action and research highlights the
limited attention to nurses immunizing frightened essential features of this method: trying out ideas
and resistant children. The apparent need to in practice as a means of increasing knowledge
force an immunization has been identified as an (Kemmis McTaggart, 1988). Kemmis and
ethical dilemma for nurses, even constituting a McTaggart (1990) also suggested that the partici-
human rights burden (Hodges, Svoboda, Van patory nature of action research, where research-
Howe, 2002, p. 12). Nurses remembered moral ers collaborate with participants in order to
dilemmas when they were left to wonder, Could understand and improve practice, can reduce the
I have done anything else even years later, con- distance between researchers and participants and
tinuing to justify and absolve themselves from the . . . problems they intend to solve, or the
blame (Gunther and Thomas, 2006). Nurses felt lived experience they intend to interpret (p. 28).
powerless, angry, exhausted, and even burned-out Data sources included two survey questions
following their participation in situations they and audio-recorded transcribed data from three
believed were ethical and moral dilemmas focus groups. The survey was distributed anony-
(Thomas, 2009). Coping with the emotional mously via employee e-mail to 58 nurses from
needs of patients and families has consistently five different health units in one Canadian
been highly stressful for nurses (McVicar, 2003 province. Survey question one: Does your prac-
Sherman, 2004). Avoiding coping rather than tice involve immunizing children Survey ques-
identifying that a problem exists and focusing on tion two: Sometimes children who present for
coping with it was found to be a significant pre- immunization strongly resist needle injection.
dictor of mood disturbance for nurses (Healy Based on your experience, what is it about this
McKay, 2000). Given our limited understanding situation that is a problem for you This survey
of links that may exist between negative immu- generated 35 (60 ) responses, all of whom con-
nization experiences and nurse stress, it is essen- firmed that their practice did include immunizing
tial to explore the problem. children.
The survey was followed by three audio-taped
THE RESEARCH APPROACH and transcribed focus groups. The focus groups
This project was framed from a constructivist were 2 weeks apart, each with five to six female,
theoretical perspective (Appleton King, 2002) English-speaking, Caucasian and Indo-Canadian
and a naturalistic action research design (Kemmis nurses, from five different health units in one
 CHAPTER 18 Critiquing Qualitative Research 409

Canadian province. The participants were those an ethical dilemma for nurses (c) some adult
who had responded to the survey, and their experi- responses make immunizing difficult and unsafe
ence ranged from novice (less than 1 year experi- (d) resources to help nurses cope with these situ-
ence) to expert (up to 25 years experience) in two ations are inconsistent.
groups. The third group had no novice participants.
Focus groups are exible and cost-efficient, gener-
ate rich data, and tend to have high face validity THEME ONE: NURSES EXPERIENCE STRESS
(Krueger Casey, 2009 Morrison-Beedy, Cote- WHEN IMMUNIZING CHILDREN WHO
Arsenault, Feinstein, 2001 Speziale Carpen- FEAR AND RESIST NEEDLE INJECTION
ter, 2003 Webb Kevern, 2000). The following Nurses used the word dread in all three focus
questions guided the discussion: groups to describe their apprehension about
1. When you hear the phrase, a child who is immunizing needle-resistant children, especially
strongly resistant to needle injection, what as a new practitioner. They described the situa-
comes to mind tions as awkward, difficult, and complex, with
2. What is it about these situations that is chal- too many pieces or variables. Nurses frequently
lenging for you recounted actual experiences to illustrate specific
3. What sorts of things have made it easier for points. Feeling ustered and fearful of making
you to immunize children who resist the needle a medication error or harming the child, as well
injection as fear for the nurse’s own safety, was reported
4. What sorts of things have made it harder in the survey and across all groups. Empathy for
5. Do you have any thoughts on how these situ- the child’s incredible panic and fear was articu-
ations can be improved lated, noting the child’s terror and screaming,
Content from these data sources was analyzed kicking, and biting behaviors as very disturbing.
for themes. The transcripts were thoroughly read I think of how hard it is to be scared. Like that’s
and reread, and a systematic process of content so much work on the child’s part. It takes so much
analysis was developed (Loiselle, Profetto- energy.
McGrath, Polit, Beck, 2007 Speziale Crying was not seen as particularly difficult,
Carpenter, 2003) to create the categorization and but acting out behaviors and struggling to get
coding scheme that led to the themes. Trustwor- away, to get out of the room were problems.
thiness was established through ongoing interac- The child’s terror, that’s what gets to me. I feel
tion and member checking with participants to really badly for the child because they’re embar-
confirm authenticity. Full ethical approval was rassed . . . and they’re kind of ashamed. The
granted by a university and a health authority. nurses felt torn about the process. They found
The following four themes emerged from ana- it very disturbing to witness the child’s distress
lyzing the survey and focus-group data collected and felt complicit in an assault. They described
from, and confirmed with, nurses who routinely feelings of helplessness and uncertainty, wonder-
immunized children. The themes represent nurses’ ing how it might have been done differently.
perceptions of what it was about immunizing One nurse wrote, I don’t know how to make
frightened and resistant children that was a these situations more comfortable. Nurses felt
problem for them. Verbatim comments are itali- . . . pressured to just finish the job, no matter how
cized. The themes are as follows: (a) nurses expe- much the child resists. Novice practitioners were
rience stress when immunizing children who fear more likely to feel pressured. Throughout my
and resist needle injection (b) the strength of orientation it was very heavily implied, it does
child resistance and some adult behavior creates not matter the situation, you always vaccinate
410 PART FIVE Critiquing Research

children for as many vaccines as they’re eligible . . . the lack of respect it demonstrates to a child. In
for. And I just feel a lot of pressure to do that deciding on their behalf what is best for them I don’t
during that clinic visit. Sometimes, the pressure understand what makes that okay and at what age
we give the child the control to make that decision. A
came from parents. I’ve had two, three different problem for me is the subjectivity of deciding what’s
scenarios where . . . the anger from the parents in the child’s best interest; subjectivity in assessing
like, Whaddya mean . . .’ And they’re going to potential harm to the child versus benefit of vaccine.
argue with you. I (parent) will hold them down
and you ill do it.’ Within each group, two or more nurses re-
The nurses reported feeling drained, emotion- counted stories of especially challenging situa-
ally exhausted, fatigued, and unsupported. A tions they thought had been handled poorly and
sense of failure, guilt, heavy heartedness, and felt regret about their involvement in the process.
frustration was expressed, as well as a scary There’s some where you’re going—oh that was
feeling of being out of control. One group awful That didn’t feel right. I don’t feel good
likened the situation to a circus, with moms about that. Children kindergarten age and older
chasing (children) around to try to get them in and were viewed as the most challenging, although
there is a waiting room full of people. Nurses some nurses also identified strong toddlers as
described feeling hurt and annoyed when parents difficult.
blamed and labeled them the mean nurse or Nurses wondered, How much restraint is too
the stabber. Nurses were troubled by the poten- much A survey responder stated:
tial for emotional scarring and serious erosion The problem becomes one of the child’s right to
of trust in the child’s relationship with health object and refuse . . . some parents like to talk their
professionals. They suspected that past experi- children into shots; this takes quite a bit of time.
ences strongly in uenced the present and believed Others are quite physical in their restraint methods
children deserve to be better prepared for and I don’t know exactly when to step in and
say—that’s enough!
immunization.
One nurse remarked:

THEME TWO: THE STRENGTH OF CHILD I don’t think the end always justifies the means.
Because I had a father who came in with a son and he
RESISTANCE AS WELL AS SOME ADULT was really quite brutal with him. And we were really
BEHAVIOR CREATES AN ETHICAL part of that because, you know, it was our end that
DILEMMA FOR NURSES we wanted to go to and that was the reason why.
And I thought, I’m never doing that again. I’m just
A major theme that emerged was the con ict
going to say, “I’m sorry, I can’t do this. This is beyond
around the child’s right to refuse versus the right what I can be part of.”
to be protected from preventable diseases. I
think as a nurse, the challenge is combining that Another recalled . . . a mother actually physi-
gentle persuasion but with letting them make cally sat on her child and restrained him and
their own decision. And we’re taught in our slapped his face and told him how much she
profession you know, do no harm. So you feel loved him and told him to just do it. Okay, and
like you’re doing harm when you encounter situ- that’s always going to come to my mind. It was
ations where there’s such strong resistance. A like an assault, us actually harassing him.
nurse wanted to find a balance between helping A colleague added:
the child find courage and protecting him from Right, and then being torn between, Do I follow
very dangerous diseases. Another stated the through, give it to him, get it over with for him? Will
problem as he have to go through this again? Or do I hold back
 CHAPTER 18 Critiquing Qualitative Research 411

and say, “Not under these circumstances.” . . . It was a

very awkward situation. “What do I do?” And I THEME THREE: SOME ADULT RESPONSES
thought, “Let’s get it over with for him. He’ll have to MAKE IMMUNIZING MORE DIFFICULT
go through that all over again or be bullied at home.” AND UNSAFE
But somehow we were then part of that.
It almost kind of reminds you, you know, of One Non-helpful responses by adults, such as parents,
Flew Over the Cuckoo’s Nest, where they have to bind school staff, or other caregivers, were defined
them down and they give them the electrical shock across all data sources as a burden to nurses. So
treatments and they don’t want it. often what I find makes it really difficult . . . I’m
not quite sure where to go with it when the par-
A survey responder commented: Immuniza- enting responses are so inappropriate. Most fre-
tion of children is recommended, not mandatory, quently cited non-helpful responses were either
therefore children may have the right to refuse. inadequate or overly forceful restraint by the
Another wrote: parent shaming, threatening, yelling, slapping,
lying or, alternately, pitying, placating, bribing,
The problem I have is with the three to five year olds wishy-washy, and helpless parental behaviors.
who clearly verbalize they don’t want the shot. We Nurses complained of getting kicked and hit by a
hold them down and do it anyways. From a young struggling child and expressed frustration with
age we teach children to use their words. We teach
parents who had not explained the purpose of the
them to say “no” to a stranger who offers candy, rides
etc. We teach them to kick, scream, and run when a visit to the child.
stranger touches them or they feel threatened by In school situations, nurses felt frustrated
them, yet I am a stranger to this child who is saying when well-meaning adults interfered with the
“no” to me and I proceed and hurt the child. What process by attempting to take control.
message are we sending these children?
It isn’t suddenly about being poked anymore. There’s
a bunch of family dynamics there as well and they get
Children with developmental delays were par- the power stuff going, and you put the child in the
ticularly challenging. A nurse recalled immuniz- school situations, sometimes it’s with the classroom
ing a grade six boy with developmental delays, teacher, you know, that’s involved as well, and you
think, “Oh boy, how many do we need involved in
It was really hard, because he wasn’t going to
this really?” We sort of bring in all the skills you have,
sit still on his own. So we had a lot of hard deci- not just the needle part, but the kind of group
sions to make with that and mom held him down. skills too.
It was awful.
In one group, a few of the more experienced It’s tough for the nurse because, ultimately
nurses initially seemed somewhat dismissive of . . . we are in charge. Nurses reported that adults
the issue as a sort of necessary evil yet, even sometimes tease students in a way that increases
these nurses acknowledged with some surprise anxiety and that students often rile up each
after the group how much there was to talk other.
about on the topic. Challenging variables Nurses disliked having competitive elements
included sheer number of vaccines, complexi- introduced into the situation. For example, parents
ties of vaccine administration, language barriers, may complain if the nurse chooses not to proceed
lack of privacy in mass immunization clinics, cir- with the immunization, with comments such as:
culating myths about needles getting stuck or She couldn’t do it so I need another nurse. One
breaking off in people’s arms, unpredictability of nurse described her dismay if a parent would tell
some resistance, noise levels, too many people her, my baby didn’t cry at all last time . . . with
involved, and lack of time. the other nurse she didn’t cry at all.’ I don’t know
412 PART FIVE Critiquing Research

why they say that to me because it hurts, it jinxes your best interest to have this. So let’s work together
me. And finally, nurses were frustrated with with parents and help them to do this. But as far as
parents who project their own fears onto the child the forcing, I will not be a party to this.
or communicate to the child expectations of resis- We sort of learn like where e draw the line
tant behavior, thus generating a self-fulfilling too, and that’s hard sometimes. A nurse with less
prophecy. than 2 years experience said: It’s different in
different places . . . like its sic okay for me here,
to say we don’t do that and I’m comfortable with
THEME FOUR: RESOURCES TO HELP that. But in another environment there might be
NURSES COPE EFFECTIVELY WITH more pressure I think, to get the thing done in a
THESE SITUATIONS ARE INCONSISTENT time frame.
AND INADEQUATE The nurses described being supported in
Nurses voiced how existing strategies and choosing to defer a vaccine as very important. A
resources to consistently support a positive immu- novice practitioner stated, I don’t think it’s made
nization outcome were inadequate, inconsistently clear to us that we can say no, that we don’t have
available, and poorly disseminated. Nurses to do it. One survey response stated:
described strategies they used to help in these
Trying to put the child at ease who has become very
situations with mixed results. Most of the strate- anxious. This can be very draining and it can be
gies were learned through trial and error or direct difficult to know when to call it off. If you call it off,
observation. A nurse with more than 10 years of then the parent (if a kindergarten immunization) is
immunization experience stated: In a school then quite often angry. Sometimes it seems like there
setting, I see it as a learning opportunity of just should be a policy or a sign that backs this up. The
sign or policy stating we will not use force to
sitting back and seeing how somebody else immunize.
handles it. I’m thinking, Thank God, I’m not the
one who has to deal with it. Collaborating with colleagues and being able
Nurses reported that crude forcible restraint is to debrief were highly valued. Occasionally,
no longer as common as it once was. I remember nurses recruited each other to assist with restraint,
a principal holding a kid against the wall actually, yet, as one nurse pointed out, It’s the same thing
believe that again, like if you’re getting another nurse. And
I think we’re better at saying we can’t do it than, let’s
then there are two of you holding the kid down.
say, fifteen years ago. I think we used to sit on kids Another agreed, Yeah, it makes it like a gang
more than we do now. I certainly, more now than I mentality. You know, we’re all ganging up
used to, just will say, “I can’t do this” . . . whereas on him.
before . . . we used to get a couple of us in there and The nurses discussed what sorts of things
really, with the parent’s permission of course, but
could make it easier for them to effectively
were more forceful.
manage situations with resistant children. They
Several nurses described how they learned, recommended combination vaccines labeled
sometimes through bitter experience, where to set trays to hold pre-filled syringes well-ventilated,
boundaries. soundproof clinic rooms separate waiting rooms
for before and after immunization and time to
And also, the holding down or the forcing, I think . . .
debrief after a difficult session. Strategies identi-
I do not have to give that, force that on that child. So I
think that’s something I’ve come to in my practice is fied as helpful included giving limited choices,
that the child does not have to have it. We will not using a calm voice, preparing parents for crying
force this child to have it . . . and so that, yes, it is in and giving children permission to cry, remaining
 CHAPTER 18 Critiquing Qualitative Research 413

firm but not threatening, using stickers to cele- fears onto the child. Referral to parent education
brate effort, and having distraction and calming sessions was a strategy employed where avail-
tools, such as puppets, bubbles, comfort dolls, able. One nurse identified the focus-group session
and cartoon videos, in waiting areas. Giving chil- itself as a useful opportunity to troubleshoot
dren time to express themselves but without and brainstorm ideas. Another talked about
engaging in endless negotiation is also important. building up your repertoire of tools and
Anesthesia was not discussed except in one explained how she benefited by learning strate-
survey response suggesting pre-procedural child gies from other nurses that would have never
sedation. occurred to me. The nurses expressed strong
Nurses desire skills to effectively manage interest in educational materials that could be
immunization procedures. I don’t have enough used by parents and children to better prepare for
skills to know what the best response or tech- an immunization appointment.
niques are to get the immunization done in a way
that is most positive for everyone involved.
DISCUSSION
I must admit, I’m better . . . more compassionate with
kids that I perceive as being truly afraid (than with) These four themes, developed from discussions
those that I think are . . . just being smart alecks. with nurses who routinely immunize children
Sometimes you get a child where you think, “Oh, who fear and resist needles, illustrate how this
you’re just trying to pull my chain and get things riled procedure is problematic and stressful for nurses.
up here.” Or you see a child that is truly just terrified
The intensity of nurse stress is re ected in the
and I’m better with the kids that are [truly terrified],
and maybe I might not even be reading it right. language participants used to describe their expe-
riences and in the vividness of their memories.
I . . . would like to learn about more tech- The words dread, awful, traumatizing,
niques for self-calming. Another wrote, Parents failure, assault, terror, fear, and shame
are often unaware of their child’s ability to learn appeared frequently in the data. Casting this
some of these skills and at how young an age it response against Lazarus and Folkman’s (1984)
can be taught. Nurses viewed the clinic visit as classic explanation that stress results when
an opportunity for children to acquire adaptive demands exceed the personal and social
coping skills and experience mastery in an honest, resources an individual is able to mobilize, study
respectful, supportive environment. Having findings lead us to question whether other nurses
enough time to prepare and also to debrief with are also feeling that the demands of immunizing
parent and child was seen as important. needle-resistant children exceed their ability
to cope.
There has been no time to prepare them in
anticipation of them being that way [so wound up, The comments re ect how the experience of
not being able to focus and calm down]. We have forcing compliance from children generates
nothing to offer these families. No opportunity to ethical and moral dilemmas for nurses. Bioethi-
teach the parents . . . we’re rushed and the parents cists Hodges et al. (2002) emphasized how
are in a hurry and there’s nothing else in place in heightened scrutiny is essential in situations
another time to prepare them. We wind up being a
part of it.
where children, who are unlikely to be able to
provide meaningful consent, are subjected to pro-
They identified a need to provide parents with phylactic interventions such as immunization.
clear direction about positioning, secure hold, and And yet, the issue may not be formally addressed
what not to say to their child, for example, limit with explicit policies and procedures in the prac-
bribes and threats, and avoid projecting parent tice arena. With the exception of the present
414 PART FIVE Critiquing Research

study, the literature has not yet begun to acknowl- nizing children and to more positive outcomes
edge that a problem exists. for all.
Nurses’ descriptions of their memories of
immunizing needle-resistant children were con- REFERENCES
sistent with the moral distress Gunther and Appleton, J. J. V., King, L. (2002). Journeying from the
Thomas (2006) described in their exploration of philosophical contemplation of constructivism to the
patient care events that were unforgettable to methodological pragmatics of health services research.
ournal of Advanced Nursing, (6), 641-648.
nurses. In both studies, nurses wondered whether
Blount, R., Devine, K., Cheng, P., Simons, L.,
they could or should have done things differently Hayutin, L. (2008). The impact of adult behaviors
even years later. Descriptions of their memories and vocalizations on infant distress during immuniza-
in the present study also re ected a sense of pow- tions. ournal of Pediatric Psychology, (10),
erlessness. Feelings of moral distress, powerless- 1163-1174.
ness, anxiety, and anger all contribute to the stress Bowen, A., Dammeyer, M. (1999). Reducing
children’s immunization distress in a primary care
and burnout Thomas (2009) identified as a per- setting. ournal of Pediatric Nursing, (5), 296-303.
sistent issue among nurses. However, nurses’ Bustos, T., Jaaniste, T., Salmon, K., Champion, D.
stress related to immunizing needle-resistant chil- (2008). Evaluation of a brief parent intervention
dren has not been previously included in discus- teaching coping-promoting behavior for the infant
sions of moral distress. immunization context. Behavior Modi cation, (4),
450-467.
CONCLUSION Cohen, L., Blount, R., Panopoulos, G. (1997). Nurse
coaching and cartoon distraction: An effective and
This article presented findings from a naturalistic practical intervention to reduce child, parent and
action research study that explored nurses’ per- nurse distress during immunizations. ournal of
ceptions of immunizing frightened and resistant Pediatric Psychology, (3), 355-370.
children. In contrast to other studies that focused Cohen, L., Manimala, R., Blount, R. (2000). Easier
said than done: What parents say they do and what
mainly on recipients of vaccines, this project they do during children’s immunizations. Children s
extends existing knowledge by describing nurses’ ealth Care, (2), 79-87.
re ections on their own experiences with immu- Devine, K., Benoit, M., Simons, L., Cheng, P., Seri, L.,
nizing by identifying four overarching themes. Blount, R. (2004). Psychological interventions for
This research found nurses experience stress acute pediatric pain. The Suffering Child, , 1-22.
Duff, A. (2003). Incorporating psychological approaches
when immunizing children who fear and resist
into routine paediatric venipuncture. Archives of
needle injection, the strength of child resistance Disease in Childhood, , 931-937.
and some adult behavior creates an ethical Ellis, J., Sharp, D., Newhook, K., Cohen, J. (2004).
dilemma for nurses, some adult responses make Selling comfort: A survey of interventions for needle
immunizing difficult and unsafe, and resources to procedures in a pediatric hospital. Pain Management
help nurses cope with these situations are incon- Nursing, (4), 144-152.
Frank, N., Blount, R., Smith, A., Manimala, R.,
sistent. This article calls for the creation of more Martin, J. (1995). Parent and staff behavior, previous
opportunities to explore whether or not immuniz- child medical experience, and maternal anxiety as
ing needle-resistant children is a problem for they relate to child procedural distress and coping.
other nurses. Articulating that a problem exists, ournal of Pediatric Psychology, (3), 277-289.
that needle procedures are often stressful, and that French, G., Painter, E., Coury, D. (1994). Blowing
away shot pain: A technique for pain management
the experience can leave nurses feeling morally
during immunization. Pediatrics, (3), 384-388.
and ethically con icted is an important first step. Gaskell, S., Binns, F., Heyhoe, M., Jackson, B. (2005).
Further study could lead to more consistent Taking the sting out of needles: Education for staff in
support for nurses who are responsible for immu- primary care. Paediatric Nursing, (4), 24-28.
 CHAPTER 18 Critiquing Qualitative Research 415

Gunther, M., Thomas, S. (2006). Nurses’ narratives injections. ournal of the American Dental Associa-
of unforgettable patient care events. ournal of tion, , 756-762.
Nursing Scholarship, (4), 370-376. Morrison-Beedy, D., Cote-Arsenault, D.,
Hamilton, J. (1995). Needle phobia: A neglected Feinstein, N. (2001). Maximizing results with focus
diagnosis. ournal of amily Practice, (2), 169-175. groups: Moderator and analysis issues. Applied
Healy, C., Mckay, M. (2000). Nursing stress: The Nursing Research, (1), 48-53.
effects of coping strategies and job satisfaction in a O’Laughlin, E., Ridley-Johnson, R. (1995). Maternal
sample of Australian nurses. ournal of Advanced presence during children’s routine immunizations:
Nursing, (3), 681-688. The effect of mother as observer in reducing child
Hodges, F. M., Svoboda, J. S., Van Howe, R. S. distress. Children s ealth Care, (3), 175-191.
(2002). Prophylactic interventions on children: Peretz, B., Efrat, J. (2000). Dental anxiety among
Balancing human rights with public health. ournal young adolescent patients in Israel. International
of Medical Ethics, , 10-16. ournal of Paediatric Dentistry, , 126-132.
Ives, M. (2007). Model empathy and respect when Polillio, A. M., Kiley, J. (1997). Does a needless injec-
immunizing children who fear needles. The Canadian tion system reduce anxiety in children receiving intra-
Nurse, (4), 6-7. muscular injections Pediatric Nursing, (1), 46-49.
Jacobson, R., Swan, A., Adegbenro, A., Ludington, S., Schecter, N. L., Bernstein, B. A., Beck, A., Hart, L.,
Wollan, P., Poland, G., et al. (2001). Making vaccines Scherzer, L. (1991). Individual differences in
more acceptable: Methods to prevent and minimize children’s response to pain: Role of temperament and
pain. Vaccine, , 2418-2427. parental characteristics. Pediatrics, (2), 171-177.
Kemmis, S., McTaggart, R. (1988). The action Sherman, D. W. (2004). Nurses’ stress and burnout. The
research planner (3rd ed.). Geelong, Victoria, American ournal of Nursing, (5), 48-56.
Australia: Deakin niversity Press. Smalley, A. (1999). Needle phobia. Paediatric Nursing,
Kemmis, S., McTaggart, R. (1990). The action (2), 17-20.
research reader (3rd ed.). Geelong, Victoria, Sparks, L. (2001). Taking the ouch out of injections
Australia: Deakin niversity Press. for children: sing distraction to decrease pain. Ameri-
Krueger, R. A., Casey, M. (2009). ocus groups: A can ournal of Maternal Child Nursing, (2), 72-78.
practical guide for applied research (4th ed.). Speziale, H., Carpenter, D. (2003). ualitative
Thousand Oaks, CA: Sage. research in nursing: Advancing the humanistic
Lawton, L., Rose, P. (2003). Changing practice in imperative (3rd ed.). Philadelphia: Lippincott,
invasive procedures: The experience of the Krishnan Williams Wilkins.
Chandran children’s centre. ournal of Child ealth Stringer, E., Genat, W. (2004). Action research in
Care, (4), 248-257. health. pper Saddle River, NJ: Pearson Prentice
Lazarus, R., Folkman, S. (1984). Stress, appraisal, Hall.
and coping. New York: Springer. Thomas, S. (2009). Transforming nurses stress and
Loiselle, C., Profetto-McGrath, J., Polit, D., Beck, C. anger: Steps to ard healing (3rd ed.). New York:
(2007). Canadian essentials of nursing research (2nd Springer.
ed.). Philadelphia: Lippincott, Williams Wilkins. man, L. S., Chambers, C. T., McGrath, P. J.,
Manimala, M., Blount, R. (2000). The effects of Kisely, S. (2006). Psychological interventions for
parental reassurance versus distraction on child needle-related procedural pain and distress in children
distress and coping during immunizations. Children s and adolescents. Cochrane Database of Systematic
ealth Care, (3), 161-177. Revie s, .
Martin, M., Ramsay, D., Whitney, C., Fiset, L., Webb, C., Kevern, J. (2000). Focus groups as a
Weinstein, P. (1994). Topical anesthesia: Differentiat- research method: A critique of some aspects of their
ing the pharmacological and psychological contribu- use in nursing research. ournal of Advanced
tions to efficacy. Anesthesia Progress: A ournal for Nursing, (6), 798-805.
Pain and An iety Control in Dentistry, (2), 40-47. Windich-Biermeier, A., Sjoberg, I., Dale, J. C.,
McVicar, A. (2003). Workplace stress in nursing: A Eshelman, D., Guzzetta, C. (2007). Effects of
literature review. ournal of Advanced Nursing, (6), distraction on pain, fear, and distress during venous
633-642. port access and venipuncture in children and
Milgrom, P., Coldwell, S., Getz, T., Weinstein, P., adolescents with cancer. ournal of Pediatric
Ramsay, D. (1997). Four dimensions of fear of dental ncology Nursing, (1), 8-19.
416 PART FIVE Critiquing Research

experiences, re ections, and feelings of public

INTRODUCTION TO CRITIQUE 1 health nurses who immunize fearful and needle-
The preceding article (Ives Melrose, 2010) is resistant children (p. 29). In addition to explora-
an example of an action research design. The tion of the nurses’ perceptions, the authors were
article is critically examined here for its rigour as interested in approaches that would decrease
an action research study, its contribution to nurses’ stress. The researchers clearly assumed
nursing, and its usefulness in practice. The crite- that nurses who immunize children, who in
ria listed in Box 18-1 are used to guide the turn fear and resist needles, find the experience
critique. stressful. This assumption was supported in the
literature review. In many qualitative studies,
Title researchers conduct a cursory literature review so
The title of the article captured the essence of the that the findings unfold naturally without a bias.
study however, it did not address the secondary In this case, the literature review probably helped
purpose of the article to consider approaches that frame the design of the study. The literature
would help reduce nurse stress during immuniza- review focused on the fear of needles, helpful and
tion. nonhelpful adult responses, and nurse stress.

Abstract Method
The abstract met the requirements of a good The method was described in a section called
abstract: It contained the background and method The research approach. The authors used a con-
(including the purpose and the sample size), and structivist theoretical perspective to explore the
it concluded with the results. experiences, re ections, and feelings of the par-
ticipants. They did not describe the philosophical
Statement of the Phenomenon underpinnings or approach of this method in any
of Interest detail but were quite explicit in describing the
Ives and Melrose (2010) clearly stated the phe- second research method of naturalistic action
nomenon of interest expressed in the research research design. It is important that the authors
question What is it about immunizing children identify the theoretical underpinnings of their
who strongly resist needle injection that is a chosen method and how the two approaches in
problem for public health nurses (p. 29). The one study support and inform the other. Appleton
authors acknowledged that needle phobia is a and King (2002), referenced in the study, stated
common problem and that extensive research has that constructivist and participative research
explored nonhelpful and helpful responses of (action research) share common methodological
adults to reduce children’s distress during a needle steps. It is difficult to assess whether Ives and
injection. Very little research, however, focuses Melrose’s (2010) study adhered to the processes
on the nurse’s response to immunizing fearful for constructivist and action research, inasmuch
children and how to cope with the subsequent as the authors did not clearly explain how they
stress. Because nurses’ responses are unknown in incorporated the methodologies. They did indi-
this field of study, the qualitative research method cate that action research has an action component
is appropriate. in which action is taken to improve practice.
Although Ives and Melrose did gather the data
Purpose from the participants on how the situation could
The purpose of the research project was twofold. be improved, no action was taken in this study to
First the authors wanted to investigate the try out new ideas.
 CHAPTER 18 Critiquing Qualitative Research 417

The questions were formulated to elicit the par-

Sampling ticipants’ descriptions of their experiences and
A survey was sent to a sample of nurses working understanding and, in addition, how practice can
in five different public health units. Of the 58 be improved.
nurses who received preliminary survey ques- Focus groups are commonly used in qualita-
tions, 35 responded that they immunized children tive research, and their use was appropriate in this
as part of their practice. Although it was not study, although researchers need to be cautious
stated explicitly, it appears that of the 35 respon- about the development of groupthink. in which
dents, 15 to 18 nurses participated in focus groups. one dominant participant can in uence other
In qualitative research, the most common type of group members (Streubert Carpenter, 2011).
sampling is purposive or purposeful sampling. Because the researchers did not identify how they
According to Streubert and Carpenter (2011), would manage this situation or gather data from
individuals are selected to participate in qualita- other participants through individual interview or
tive research based on their first-hand experience observation, this is a potential weakness of the
with a culture, social process, or phenomenon data-collection method.
of interest (p. 28). This ability to describe an Ethical approval was sought and received from
experience from one’s own perspective is what a university and a health authority who had juris-
makes selection purposive. In this study, all par- diction over the health units.
ticipants had experience with immunizations of
children thus the selection of participants was
appropriate. Data Analysis
The authors indicated that they analyzed the data
Data Collection for themes. Each of the transcripts from the focus
The primary source of data was a series of three groups was thoroughly read and reread, and a
audiorecorded and transcribed focus groups. The systemic process of content analysis was devel-
researchers described the focus groups as 2 weeks oped . . . to create the categorization and coding
apart, each with five to six female, English- scheme that led to the themes (p. 32). No further
speaking, Caucasian and Indo-Canadian nurses, detail was provided.
from five different health units in one Canadian In general, the measure of rigour in qualitative
province (p. 32). The focus groups were selected research is trustworthiness. Trustworthiness
an early survey that differentiated between nurses includes the concepts of credibility, auditability,
who immunized children and those who did not. and fittingness. It is difficult to determine whether
Five questions were used to guide the discussion: the researchers addressed the auditability (to
1. When you hear the phrase, a child who is what extent another researcher or a reader can
strongly resistant to needle injection, what follow method and conclusions drawn by the
comes to mind original researcher s ) and ttingness (how
2. What is it about these situations that is chal- applicable the study findings are to others in
lenging for you similar situations ) of the data (see Chapter 14 for
3. What sorts of things have made it easier for complete discussion). Although the researchers
you to immunize children who resist the did specify that they established trustworthiness
needle injection by ongoing interaction and member checking
4. What sorts of things have made it harder with participants to confirm authenticity (p. 32),
5. Do you have any thoughts on how these they did not clarify how they accomplished
situations can be improved this.
418 PART FIVE Critiquing Research

Findings Conclusions, Implications,

The authors reported their findings clearly and and Recommendations
comprehensively. To provide authenticity of the The authors resummarized the findings of the
findings, they provided quotations from the par- nurses’ perceptions of immunizing needle-
ticipants, allowing the voices of the nurses to be resistant children and called for the creation of
heard. Each of the four themes that emerged were more opportunities to explore whether this is a
confirmed with the nurses who participated. The problem for other nurses. Because one of the iden-
description of each theme was thorough and com- tified purposes of this study was to consider
pelling in the depth of description. The findings approaches that are helpful to decrease nurses’
seemed to accurately re ect the conceptualization stress, the lack of recommendations is a major gap
of the emergent themes. The findings were then in this study: Action research is specifically
placed in the context of what is known about designed as a research method whose outcome
nurse stress and immunizations with needle- is the implementation of an action or change
resistant children, particularly through the litera- (Streubert Carpenter, 2011, p. 308). Because it
ture on stress, ethical and moral dilemmas, and is clear from the study that administering injec-
moral distress. tions to resistant children is an ethical and moral
dilemma for nurses, causing stress and internal
con ict, the lack of recommendations is unfortu-
nate. Careful application of action research
methods in this study could have resulted in sig-
nificant changes in practice and contribution to
nursing knowledge.
 CRITIQUING A QUALITATIVE RESEARCH STUDY
The study “Women’s Decisions to Seek Treatment for the Symptoms of Potential Cardiac Illness”
by Sheila A. Turris (2009) is critiqued here. The article is presented in its entirety and is followed
CRITIQUE 2 by the critique on pp. 430–432. (From Journal of Nursing Scholarship, 41[1], 5-12. © 2009 Sigma
Theta Tau International.)

Women’s Decisions to Seek Treatment for the

Symptoms of Potential Cardiac Illness
Sheila A. Turris, RN, PhD estimated 310,000 people died before reaching an
Vancouver Coastal Health
Vancouver, British Columbia emergency department (ED) for treatment of a
cardiac event (American Heart Association,
Cardiac disease is the number one killer of Canadian 2008). Attention is being directed to a mandate of
women and therapies are often highly time dependent. treatment within 1 hour of symptom onset, given
Consequently, understanding the forms of knowledge that the number of deaths that occur in the first hour
women draw upon to understand and interpret their symp- (McKinley et al., 2004). Treatment-seeking delay
toms of a potential cardiac event may be important in (TDS), in the context of cardiac illness, is a per-
relation to efforts to reduce treatment-seeking delay. Data
for this grounded-theory study were drawn from in-depth
sistent clinical issue.
interviews with 16 women who went to one of two emer- When considering the literature with regard to
gency departments between June 2005 and June 2006 for TSD, gender and social support have also gar-
the treatment of symptoms indicating potential cardiac nered a great deal of attention. Specifically, there
illness. The basic social process of maintaining integrity is considerable debate about gender expectations
explained the women’s interpretations of and actions in
concerning cardiac symptoms. For example, a
relation to their symptoms. This paper indicates the findings
as they relate to specific forms of knowledge the women persistent belief is held by both men and women
used to make decisions about treatment-seeking, including that men are more prone than are women to acute
embodied, temporal, rational, and relational ways of myocardial infarction (Hammond, Salamonson,
knowing. Davidson, Everett, Andrew, 2007 Harralson,
2007 Le er Bondy, 2004). One perception is
CLINICAL RELEVANCE: The results of this study contribute that women handle stress more effectively and
to the body of knowledge with regard to women’s decisions are therefore perceived as being protected from
about seeking treatment for the symptoms of potential
cardiac illness.
myocardial infarctions (Richards, Reid, Watt,
2002). Some evidence also exists to show that
KEYWORDS: Cardiovascular, grounded theory, women’s
women are more likely to delay treatment than
health, qualitative methodology/qualitative research, are men: however, contradictory evidence is
research methods reported. Although some researchers have found
that being women is associated with greater delay
Heart disease is the leading cause of death (Bowker et al., 2000 Hannegan et al., 2000
in Canada (Canadian Institute of Health Informa- Goldberg, Gurwitz, Gore, 1999) others have
tion, 2006) and the number one killer of Canadian found that gender contributes no unique variance
women (Statistics Canada, 2004). If a cardiac to delay time (Moser, McKinley, Dracut,
arrest takes place at home, survival rates are Cluing, 2005).
dismal (Herlitz, Eek, Holmberg, Engdahl, Hol- Social support has also been studied with a
mberg, 2002). Last year in the nited States, an particular focus on advice seeking after the onset

419
420 PART FIVE Critiquing Research

of symptoms indicative of cardiac illness. Hospital B is a community hospital, located in a

Researchers have consistently found an associa- residential suburb. Hospital B is not designated
tion between seeking advice from a physician and as a trauma hospital, and staff did not have the
prolonged decision time (Bowker el al., 2000 infrastructure necessary to carry out rescue angio-
erwic, Ryan, DeVon, Drell, 2003). graphic procedures that might be required for the
Similarly, seeking advice from a relative treatment of acute myocardial infarction.
(Rosenkid, 2004) and living alone (Banks
Dracup, 2006) are typically associated with a Sampling and Procedures
delay in seeking treatment. In this grounded At the beginning of the study, the sample was a
theory study, a sample of women explained their purposive sample drawn from a group of women
decision-making processes in relation to visiting who sought ED care for the treatment of symp-
an ED for treatment of symptoms indicating toms of a potential cardiac illness at one of two
potential cardiac illness. The findings are reported hospital EDs. To participate in this study, indi-
as they relate to the specific forms of knowledge viduals had to be: women, visiting either of the
the women used to make decisions about treat- two EDs for treatment of the symptoms of a poten-
ment seeking. tial cardiac event, hemodynamically stable, 18
years of age or older, able to speak English, and
METHODS competent to provide informed consent. If a
Grounded theory (GT), the method used for this woman met the inclusion criteria, permission was
study, involves a systematic generation of theory requested for contact several days following dis-
from data. These data are constantly compared, charge to arrange a convenient time and location
at increasing levels of abstraction, searching for for a more formal, audiotaped interview, and she
evidence of a core category. One goal of GT is to was provided with an information letter about the
discover the main concern of participants and study. Each participant had more than 24 hours to
how they are trying to address that concern. In consider whether she wanted to participate. Fol-
GT, one does not aim for the truth but rather lowing discharge, the women were contacted by
the goal is to explain what is going on in relation telephone and a mutually convenient date, time,
to the phenomenon of concern. The research and location for an interview was arranged. Before
question was: In the context of having symp- being interviewed, written consent was obtained,
toms suggestive of cardiac illness, what process and a copy of the consent form was provided. The
guides women as they interpret their symptoms, women were also reminded that they could choose
make decisions about their treatment options, and not to respond to particular interview questions, or
subsequently make sense of their experiences of to request that the tape recorder be turned off at
ED care any time. In these ways, ongoing informed consent
was addressed. The audiotaped interviews took
Setting between 30 and 90 minutes to complete. The tran-
This project was undertaken with the approval of scripts were password protected and identifying
the university’s behavioural research ethics board, data were removed.
as well as the relevant hospitals’ ethics boards.
Data were collected from two EDs in a western Data Analysis
Canadian province, targeting one tertiary-care Data collection and analysis were done concur-
and one community hospital, both with more than rently. The accounts of 10 participants were
50,000 ED visits a year. Hospital A is a tertiary- obtained as a starting point for this project. Sam-
care referral centre for the province. In contrast, pling continued to ensure that categories were
 CHAPTER 18 Critiquing Qualitative Research 421

dense and well developed support for the rela- (Bonner Tolhurst, 2001 Glaser, 1992). In com-
tionships among the categories and the contribu- mencing the project, the initial research question
tions of categories to the core category were was clear and the design of the study was appropri-
explored. Theoretical sampling was used to collect ate for the research goals. Each successive inter-
further information. For example, when analysis view with a participant allowed confirmation (or
of the data suggested age might play a role in obtaining a contrasting account) and further devel-
women’s decisions about treatment seeking, the oping themes from earlier accounts. In these ways,
accounts of younger women were actively sought. the emerging theory was confirmed to fit the data.
pon the conclusion of the project, a total of 16 Fit is described as the extent to which a theory fits
women had been interviewed. or describes the situations in the social area under
Analysis of the data occurred over several study. Glaser and Strauss (1999) pointed out that a
months and involved open, selective, and theo- lack of fit results in the need to force the data to fit
retical coding. Each level of coding moved the the core category emerging. During biweekly
data analysis toward increasing levels of abstrac- meetings of the research team, discussing the early
tion. Open coding was the first step. Incidents findings allowed identification of gaps in the anal-
were compared to other incidents and similarities ysis and contributed to development of an inte-
and differences were sought, compared with other grated theory.
concepts, and subsequently grouped to form cat- Relevance, in the context of GT, is the idea that
egories. This process is termed constant compara- the theory produced should be relevant to action
tive analysis. Initially, labels for the categories in the area it purports to explain (Lomberg
were drawn from the interviews directly. As the Kirkevold, 2003). One might ask, to what extent
analysis proceeded, the labels became increas- does the theory generated explain actual prob-
ingly explanatory. Gradually, information was lems and basic social psychological processes in
grouped according to relationships and interac- the research setting The issue of relevance was
tions important concepts became apparent in addressed in the following specific ways. As data
successive accounts. were collected, emerging themes were checked
These concepts contributed to development of with the participants, as well as with a group of
increasingly complex categories to explain what colleagues, asking specifically about the explana-
was going on in the women’s accounts. Catego- tory power of the emerging theory.
ries had two essential features: they were both Finally, with regard to rigor, the issue of modi-
analytic and sensitizing. The former refers to the fiability was considered. Modifiability is a quality
quality of being sufficiently abstract and the latter of the theory produced (Glaser, 1992). A theory
refers to the ability of a code to generate a mean- should be able to evolve and encompass new
ingful picture (Glaser Strauss, 1999). Initially, information. For example, a rigorous GT explain-
the categories were general and broad they ing women’s decisions in relation to their symp-
became more specific and better developed as toms might be used as a starting point for
new data were constantly compared with that researching men’s decisions.
obtained in earlier interviews. Evidence of the In a GT approach, the goal is to formulate
core category was sought throughout the analytic hypotheses based on conceptual ideas and to dis-
process. cover participants’ main concern and how they
continually try to resolve it this is done through
Rigor the analysis of empirical data. The findings
The specific criteria for rigor in a GT study includes reported in this manuscript are from a larger study
fit, work, relevance, and modifiability focused on the investigation of women’s
422 PART FIVE Critiquing Research

decisions about the symptoms of potential cardiac by particular ways of knowing. Four ways of
events, and their subsequent experiences of care knowing were evident in the data including:
in an ED (Tunis Johnson, 2008). The following embodied, temporal, rational, and relational
discussion indicates the findings as they relate knowing. The women used these ways of knowing
specifically to the women’s decisions about to understand and interpret their symptoms within
seeking treatment. In the larger study, the basic a larger personal and social context. The under-
social process of maintaining integrity was identi- standings gleaned from particular ways of
fied encompassing three phases: resisting disrup- knowing provided information about specific
tion, suspending agency, and integrating new threats to the women’s personal, social, and phys-
knowledge and experiences. In the first phase, ical integrity and, most important, shaped deci-
resisting disruption, participants drew on particu- sions about treatment seeking.
lar forms of knowledge to make decisions about
their symptoms of potential cardiac events. The Embodied Knowing
findings indicated below further our understand- Embodied knowing indicated information about
ing about the decisions women make in relation the subjective, visceral experience of symptoms
to the symptoms of potential cardiac events and in relation to quality, intensity, and effects. The
therefore have potentially important implications visceral experience of the symptoms was promi-
for nursing practice. nent in each account of the participants in this
study. For example, one participant reported,

FINDINGS I didn’t feel any stress. I wasn’t tense at all . . . I

started feeling chest pain . . . chest tightness more
Ages of participants ranged from 29 to 89 years. than chest pain . . . I remember that I was a little bit
Slightly fewer than half were employed at the short of breath and my head was pounding and I
time of the interview. Roughly one fourth of the noticed that my hands were shaking.
participants were married and half had children. Important features of embodied knowing
The majority of women experienced the onset of included the number of symptoms experienced
symptoms while at home 7 participants reported the intensity of those symptoms the development
that family members were present when the of any new symptoms and evaluation of symp-
symptoms began. In relation to the amount of toms based on past experiences, as well as the
time that elapsed between onset of the symptoms, degree to which the symptoms interfered with
delay ranged from approximately 30 minutes to normal daily activities. With rare exceptions, the
more than 48 hours. Twenty percent of the par- understanding gained from embodied knowing
ticipants were transported to the ED by ambu- delayed treatment seeking. In fact, the minimiza-
lance 62 drove in a private vehicle. Roughly tion of symptoms was a recurring theme. The
one third of the participants received a cardiac- participants made particular assumptions about
related diagnosis. what constituted an emergency. The criteria for
According to accounts of the participants, an emergency included visibility. The illness or
decisions about treatment seeking were in u- injury had to be readily apparent to others and not
enced primarily by the context of the women’s simply an internal experience. For example, when
lives as mothers, daughters, and wives and only presented with a scenario in which a fall occurred
secondarily by the women’s interpretations of, and resulted in an ankle fracture, the women
and conclusions about, their symptoms. Resisting expressed no hesitation about visiting an ED. In
disruption of their personal, social, and physical contrast, the symptoms of a cardiac event are
integrity was a dynamic process in uenced internal, for the most part, and not always obvious
 CHAPTER 18 Critiquing Qualitative Research 423

to bystanders or family members. As one women (e.g., meal times) and were sometimes pivotal in
explained, the women’s decisions about treatment seeking.
If it’s not so significant that people are turning around Symptoms could be minimized during the night—
on the street looking as I go past, I could wait . . . that is, a woman might say to herself, I’ll go to
until I’ve got a day off, or I’ll find something is open the hospital in the morning if I still have the
on a Saturday, but it will wait . . . it will definitely pain. Similarly, symptoms that started before
wait, it’s not that bad. . . . If I had a compound breakfast were judged to be serious only alter
fracture I don’t think I’d hesitate to bother somebody,
but I think things like whatever was the matter with additional meals had passed. These time seg-
me that isn’t completely debilitating, you can still go ments were often seen as milestones. Most of the
to the store, you can still . . . look after your husband, participants in this study expressed a desire to
your children, or your dog, you know can certainly wait and see whether the symptoms resolved
wait until tomorrow. spontaneously.
Ideas about gender were used to downgrade Temporal knowing also drew upon the past
the level of threat attached to the experience of experiences of the participants. For example,
the symptoms. For example, participants fre- several of the women spoke about interpreting
quently mentioned the belief that women have a their recent event in light of a past cardiac event,
higher pain tolerance than men and accordingly in order to estimate the threat level. One partici-
the experience of pain was often dismissed or pant had had an acute myocardial infarction 6
rated as unimportant. One participant said, months before the interview. She spoke about her
I think if I had been given half a chance, by about
need to analyze every ache and pain in light of
9:30 or l0:00 that morning I might have come home this event.
and just lie down and decided I’ll just see if it doesn’t [My] assessment skills before were very on until this
go away on its own . . . I think women are more stoic whole issue came up and then all of a sudden this . . .
than men and I think it would . . . you know, when a it just means that all the ways I looked at discomfort
man is going [making sound of breathing hard] . . . I and all the ways I looked at pain and handled them in
think women just think, “Oh, I’ll take an aspirin.” no nonsense ways and everything and this just makes
you sort of second guess it, you know?
Although embodied knowledge primarily
delayed treatment seeking, two important excep- The use of a past event to evaluate the current
tions were noted. First, the development of threat was time limited: With the passage of time,
new symptoms usually prompted participants to the past event became less significant. If a previ-
reconsider their initial evaluation of the signifi- ous cardiac event had occurred relatively recently,
cance of their condition, and this reconsideration the participants were likely to associate the new
usually resulted in a prompt visit to the ED. event with that event and to promptly seek
Second, it the symptoms interfered with the medical attention. However, if the original event
women’s ability to work, drive, or take care of was in the distant past, participants were more
their families, this constituted appropriate grounds likely to delay, while considering many other
for seeking treatment. causes for their symptoms.

Temporal Knowing Rational Knowing

Temporal knowing indicated the contributions of Rational knowing indicated knowledge gleaned
time in relation to the women’s interpretation of from the media, reported statistics, and general
their symptoms. The women did not measure information about women and heart disease.
time in minutes and hours. Instead, time was Almost without exception, participants in this
divided into sections based on social convention study noted that women have different symptoms
424 PART FIVE Critiquing Research

of a heart attack than do men. An assessment of you’ve got to think about it and you don’t want to be
personal risk for heart disease, drawing on their . . . you don’t want to collapse at work and have
knowledge of their family history and personal somebody say, “You know, you shouldn’t have come
to work.” As it is at work, they’re always saying to me.
health history, was also prominent in their “I always worry about you driving yourself to the
accounts. Yet even when participants expressed hospital,” and I’m thinking, “Oh my God!” I said, “I
concern about their level of risk (i.e., strong keep to the outside lane.”
family history of premature heart disease), a
prompt visit to an ED did not always occur. The Second, the effect of treatment-seeking deci-
only time that rational knowledge was prominent sions on the social roles held by the women
in reducing treatment delay was when such within their families was often a critical in uence
knowledge was taken up by family members. on their ultimate decisions. In the following
Family members used this knowledge to force example, a participant explained her perspective
action on the part of the participants. Often, the about the social costs to her if she acted on her
primary reason participants came to the ED for experience of symptoms, admitting illness.
treatment was because a daughter or other family Well . . . my family demographics . . . the whole thing
member had insisted on that particular course of . . . the boys always rule [laughs]. That seems to be
action. In this situation, the women were not able sort of how things go . . . just family politics. . . . You
to resist disruption and visited an ED, however sort of feel a bit like the family black sheep and yet
reluctantly. I’m the one they all phone. But you just don’t want to
give them any more . . . anything else that makes you
Relational Knowing feel like a flake.

Relational knowing indicates the knowledge that Other women emphasized that responsibility
arises from being part of a social world composed at home, particularly in relation to young chil-
of relationships and interactions that shape behav- dren, in uenced decisions about seeking treat-
iour and meaning. Important features of relational ment. One woman explained that she had not told
knowing included perceptions about the social anyone at all about the chest pain, partly because
consequences of seeking or delaying medical she had children and no child-care. She also
treatment. Relational knowing was pivotal in refused to call a friend and take that person away
in uencing decisions. from her own children.
In terms of the social costs associated with Third, relational knowing was to a great extent
choosing to seek treatment, four issues of concern about understanding treatment-seeking options
were described. First, the participants discussed within a larger social framework. Specifically, the
the dangers associated with not seeking treatment women expressed concern that they would
and carrying on with the normal activities of their waste health-care resources. The women who
daily lives. Considerations about the dangers of delayed seeking treatment explained that under-
driving (e.g., colliding with another car in the lying the decision to wait and see was a desire
process of driving oneself to the ED) and the risk to balance on the one hand, the reality that some-
to others if one collapsed at work and accidentally thing might be wrong, and on the other hand, the
injured someone else in the process were exam- reality that if nothing was medically wrong, social
ples of the concerns expressed by participants. embarrassment and wasted health-care resources
For example, would result.
I wasn’t doing very much, so it wasn’t like it was a big I kept thinking, with me, that it . . . will go away on
stressful deal, but you know, when you’re running out its own and particularly having been into emergency
the door to go to work and everything like this and Monday morning, you just looked around and you
 CHAPTER 18 Critiquing Qualitative Research 425

thought, “These people have just so much to do, wives, and good employees. As discussed above,
there aren’t enough of them, they’re overwhelmed, I particular forms of knowledge were employed in
don’t want to bother these people with this little the decisions made by the women in relation to
whatever it is I’ve got because I’m not bleeding, I’m
not throwing [up].” And you look around and the
their treatment seeking. The findings of this study
place is just jumping. You do feel a bit like, “Oh, wow! indicate that attending to role responsibilities as
What am I doing here? Don’t bother them.” That wives and mothers was judged to be vital, whereas
probably says it quite well. “Wow! There’s sick people attending to symptoms that might indicate cardiac
here. What am I doing here?” disease was given a lower priority.
Finally, relational knowing prompted the
women to explore not only the cause of their
symptoms, but to interpret those symptoms within DISCUSSION
the context of their values and the quality of their More than 30 years have passed since Carper
lives. As one participant stated, (1978) published her paper, Ways of Knowing.
Although this paper does not extend Carper’s
I would prefer to go . . . if I’m going to pop off
work, our analysis of these data confirms Carper’s
someday, that I pop off with something that’s final:
that I don’t get rescued to be a “case,” you know. But theory specifically, there are many ways to know
I’m not ready to go yet: I’ve got things to do. and understand the world. Women making deci-
sions about treatment seeking draw upon particu-
In contrast, several of the participants in this lar forms of knowledge, here characterized as
study shared their experiences of multiple losses ways of knowing. In nursing and the social
such as children, partners, and close friends, as sciences, this terminology is not uncommon.
well as being concerned about their growing Belenky, Clinchy, Goldberger, and Tarule (1986)
dependence on their family members. They used the term to describe received, subjective,
reported that given their age, and these losses, procedural, and constructed forms of knowledge.
they were not unhappy at the thought of experi- se of the phrase is in line with Carper’s (1978)
encing a cardiac event that might result in death. ways of knowing in nursing practice (e.g., empiri-
Several women provided examples of perfect cal, ethical, esthetic, and personal) or Benner’s
days they had had, and in this context spoke (1984) case, patient, and person knowledge,
about death and expressed an acceptance that it which is used by expert nurses.
may be my time and I have lived a good life. In terms of the larger body of literature on
One participant stated, If it is God’s will, He will TSD, several findings of this study are notewor-
take you home. I have had a good life and I am thy. Although the results of this study were in
ready. For these women, the onset of symptoms concordance with the findings of other research-
was, if not welcome, then timely. Not surpris- ers who reported that rational knowledge—that
ingly, this group of women did not tend to seek is, knowledge about women’s risk of heart
treatment in a timely fashion, but rather delayed— disease, or the symptoms of a heart attack—plays
often for days—until a family member little role in reducing TSD (Dempsey, Dracup,
intervened. Moser, 1995), the extent to which such informa-
In summary, the women in this study analyzed tion was used by family members to urge action
their symptoms and managed the situation, was surprising. In fact, the results of this study
actively problem solving. Their goal was to resist are in con ict with reports that consultation with
any disruption to the rhythm of their daily lives family members increases delay (Hartford, Sjolin,
or to their images of themselves as healthy and Herlitz, 1993 Rosenfeld, 2004). For the par-
capable of continuing to be good mothers, good ticipants, consulting family members for advice
426 PART FIVE Critiquing Research

typically resulted in an immediate visit to an ED, treatment, they paused to re ect on their lives and
a visit that might not have taken place at expressed a sense of acceptance that the symp-
all without such a consultation. The results of toms they were experiencing might be timely
this study indicate that the relationship given their situations in relation to age, marital
between TSD and seeking advice should be re- status, and life satisfaction.
examined. Because women who do not visit an Con icting reports in healthcare literature,
ED are not included in studies, there might be regarding the association between past medical
misinterpretation regarding the effect of family history (e.g., previous acute myocardial infarc-
intervention. tion) and TSD have persisted for the past two
Although there are consistent reports of women decades (Ho, Eisenberg, Litwin, Schaeffer,
underestimating their risk of heart disease as Damon, 1989 Leitch, Birbara, Freedman,
compared with men (Oliver-McNeil Artinian, Wilcox, Harris, 1989: Wu, hang, Li, Hong,
2002), participants in this study were well aware Huang, 2004). The results of the present
that women were at equal risk and that women’s study show that this might in part be because of
symptoms might differ from those of men in a the length of time that has passed between the
similar situation, suggesting that the message is original event and the recurrence, indicating that
getting out to the public. However, the data from further exploration of this relationship might be
this study do not confirm that this knowledge useful. For the participants in this study, if a
affected the women’s decisions about treatment prior health event occurred within 18 months of
seeking. the ED visit, the women were likely to attribute
In addition, it was noteworthy that none of their symptoms to a cardiac cause, rather than a
the participants in this study had knowledge more benign gastrointestinal or musculoskeletal
about medications or medical procedures that cause. If subsequent studies show an association
might reduce damage to the heart muscle during between a reduction in delay and recent past
a myocardial infarction. Therefore, it might be medical history, this might be emphasized in
important to consider that particular types of interactions with primary healthcare profession-
information are more effective than others in als. For example, a physician or nurse practitio-
relation to reducing TSD. For example, several ner might say, As more time passes, you will
researchers report that although education be tempted to relax and become less vigilant
about the symptoms of a myocardial infarction about the symptoms of a heart attack. This
did not make a difference, knowledge about would put you at risk.
thrombolytic therapy and angioplasty did
reduce delay (Dracup, McKinley, Moser,
1997: Moser et al., 2006). LIMITATIONS
Advancing age is sometimes reported to be When interpreting the results of this study, several
associated with longer delays. The results of this limitations should be considered, particularly in
study indicate that further exploration of this rela- relation to the design of the study. First, the
tionship might be useful. A basic assumption of accounts and experiences of women who did not
this study was that reducing TSD is worthwhile come to an ED for treatment are missing. It is
because such a reduction addresses the clinical possible, and indeed probable, that decision
problem of premature death from an acute myo- making differs for this group in comparison to the
cardial infarction. However, the women in this sample for this study. Second, women who were
study who were more than 80 years of age critically ill were not recruited for participation
reported that when deciding whether to seek in this study because of hemodynamic instability.
 CHAPTER 18 Critiquing Qualitative Research 427

Third, one must bear in mind that this is a single underpinning women’s decisions about their
study. As such, the results add to our understand- symptoms allows an exploration of decision
ing about treatment-seeking delay and experi- making from women’s perspectives, rather than
ences of ED care but are an insufficient basis approaching the discussion exclusively from the
upon which to recommend change. Finally, quali- perspective of healthcare professionals.
tative findings are not open to generalization but
might further add to information about phenom-
ena of interest. CONCLUSIONS
This project arose from a strong interest in out-
comes for women who experience symptoms of
IMPLICATIONS FOR NURSING PRACTICE potential cardiac illness and seek treatment in an
The subset of findings reported in this manuscript ED. A unique contribution of this study is a broad
suggest that TSD is to a great extent a social focus on women’s decisions in relation to poten-
phenomenon, and not, as currently portrayed in tial symptoms of cardiac illness rather than an
the healthcare literature, primarily a product of exclusive focus on those who have actually been
individual factors. Participants in this study made diagnosed with an acute myocardial infarction. A
decisions about treatment seeking based on par- clear understanding with regard to the decisions
ticular ways of knowing. They used temporal, undertaken by women experiencing symptoms of
embodied, rational, and relational knowledge to potential cardiac illness can inform nursing
understand and interpret their symptoms of a interventions before a cardiac event, preventing
potential cardiac event. nderstanding the death and reducing morbidity. This work advances
forms of knowledge used by women in making theory development in relation to TSD, making a
decisions indicates the process by which deci- contribution to existing theory and research about
sions about treatment are made and allows us to this clinical issue.
have conversations with women about those deci-
sions before a crisis arises. For example, under-
standing that women tend to prioritize family ACKNOWLEDGEMENTS
needs above their own health needs might give us This research was supported by scholarships and
an opportunity to open a conversation about deci- awards from the Canadian Institutes of Health
sion making in the context of the symptoms of an Research (CIHR), the Michael Smith Foundation
acute cardiac event during nonurgent healthcare for Health Research, the Nexus Research nit at
encounters. the niversity of British Columbia, and the i
Theory about TSD is currently focused primar- Eta Chapter of Sigma Theta Tau International. I
ily on the proximate causes ignoring, for the most also acknowledge the helpful comments of Drs.
part, the larger social context of treatment seeking Joy Johnson, Pamela Ratner, and Bonita Long,
highlighted by the participants in this study. and thank the study participants.
nderstanding TSD as a social phenomenon may
widen the focus of research, and this is vital
because only 41 of the variance in decision CLINICAL RESOURCES
times is explained by a model that includes American Diabetes Association: http://www.
gender, age, past medical history, comorbidities, diabetes.org/diabetes-heart-disease-stroke.jsp
educational level, and the presence of family American Heart and Stroke Foundation: http://
members (Wu et al., 2004). Developing an under- www.amhrt.org/presenter.jhtml identifier
standing with regard to the forms of knowledge 1200000
428 PART FIVE Critiquing Research

British Heart Foundation: http://www.bhf.org. Glaser, B. G. (1992). Basics of grounded theory. Mill
uk/ Valley, CA: Sociology Press.
Glaser, B. G., Strauss, A. A. (1999). The discovery of
Canadian Heart and Stroke Foundation: http://
grounded theory: Strategies for ualitative research
ww2.heartandstroke.ca/splash/ (2nd ed.). New York: Aldine de Gruyter.
Framingham Heart Study: http://www. Goldberg, R. J., Gurwitz, J. H., Gore, J. M. (1999).
frarningham.com/heart/ Duration of and temporal trends (1994-1997) in
National Heart, Lung, and Blood Institute prehospital delay in patients with acute myocardial
(2009): http://www.nhlbi.nih.gov//educational/ infarction. Archives of internal Medicine, ,
2141-2147.
hearttruth/ Hammond, J., Salamonson. Y., Davidson. P., Everett, B.,
Andrew, S. (2007). Why do women underestimate
REFERENCES the risk of cardiac disease Australian Critical Care,
American Heart Association. (2008). eart disease and (2), 53-59.
stro e statistic, update (pp. 1-41). Dallas, T : Harralson, T. L. (2007). Factors in uencing delay in
Author. seeking treatment for acute ischemic symptoms
Banks, A. D., Dracup, K. (2006). Factors associated among lower income, urban women. eart and ung,
with prolonged prehospital delay of African Ameri- (2), 96-104.
cans with acute myocardial infarction. American Hartford, M., Sjolin, M., Herlitz, J. (1993). Symp-
ournal of Critical Care, , 149-157. toms, thoughts, and environmental factors in
Belenky, M. F., Clinchy. B., Goldberger, M., Tarule, suspected acute myocardial infarction. eart and
J. (1986). omen s ay of no ing: The development ung, (1), 64-70.
of self, voice, and mind. New York: Basic. Herlitz, J., Eek, M., Holmberg, M., Engdahl, J.,
Benner, P. (1984). rom novice to e pert. New York: Holmberg, S. (2002). Characteristics and outcome
Addison Wesley. among patients having out of hospital cardiac arrest
Bonner, A., Tolhurst, G. (2001). Insider-outsider at home compared with elsewhere. eart ournal, ,
perspectives of participant observation. Nurse 579-582.
Researcher, (4), 7-19. Ho, M. T., Eisenberg, M. S., Litwin, P. E., Schaeffer, S.
Bowker, T. J., Turner, R. M., Wood, T. L., Roberts, T. M., Damon, S. K. (1989). Delay between onset of
L., Curzen, M., Gandhi, M., et al. (2000). A national chest pain and seeking medical care: The effect of
survey of acute myocardial infarction and ischaemia public education. Annals of Emergency Medicine, ,
(SAMII) in the K: Characteristics, management and 727-711.
in-hospital outcome in women compared to men in Le er, L. L., Bondy, K. N. (2004). Women’s delay in
patients under 70 years. European eart ournal, , seeking treatment with myocardial infarction: A
1458-1463. meta-synthesis. ournal of Cardiovascular Nursing,
Canadian Institute of Health Information. (2006). ealth , 251-268.
care in Canada. Ottawa: Author. Leitch, J. W., Birbara, T., Freedman, B., Wilcox, I.,
Carper, B. A. (1978). Fundamental patterns of knowing Harris, P. (1989). Factors in uencing the time from
in nursing. Advances in Nursing Science, (I), 13-23. onset of chest pain to arrival at hospital. Medical
Dempsey, S. J., Dracup, K., Moser, D. K. (1995). ournal of Australia, , 6-8.
Women’s decision to seek care for symptoms of Lomberg, K., Kirkevold, M. (2003). Truth and
acute myocardial infarction. eart and ung, , validity in grounded theory—a reconsidered realist
444-456. interpretation of the criteria: Fit, work, relevance and
Dracup, K., McKinley, S. M., Moser, D. K. (1997). modifiability. Nursing Philosophy, , 189-200.
Australian patients’ delay in response to heart attack McKinley, S. M., Dracup, K., Moser, D. K., Ball, C.,
symptoms. Medical ournal of Australia, , Yamasaki, K., Kim, C.-J., et al. (2004). International
233-236. comparison of factors associated with delay in
Finnegan, J. R., Meischke, H., apka, J. G., Leviton, L., presentations for AMI treatment. European ournal of
Meshack, A., Benjamin-Garner, R., et al. (2000). Cardiovascular Nursing, , 225-230.
Patient delay in seeking care for heart attack Moser, D. K., Kimble, L. P., Alberts, M., Alonzo, A. A.,
symptoms: Findings from focus groups conducted in Croft, J. B., Dracup, K., et al. (2006). Reducing delay
five S regions. Preventive Medicine, , 205-213. in seeking treatment by patients with acute coronary
 CHAPTER 18 Critiquing Qualitative Research 429

syndrome and stroke: A scientific statement from the Decision trajectories and their predictors. Nursing
American Heart Association Council on Cardiovascu- Research, , 225-236.
lar Nursing and Stroke Council. Circulation, , Statistics Canada. (2004). Age-standardized mortality
168-182. rates (female). Retrieved October 26, 2004, from
Moser, D. K., McKinley, S. M., Dracup, K., Chung, lutp://www.statcan.gc.ca/daily-quotidien/040927/
M. L. (2005). Gender difference in reasons patients dq040927a-eng.htm
delay in seeking treatment for acute myocardial Turris, S. A., Johnson, J. L. (2008). Maintaining
infarction symptoms. Patient Education and Counsel- integrity: Women and treatment seeking for the
ling, , 45-54. symptoms of potential cardiac illness. ualitative
Oliver-McNeil, S., Artinian, N. T. (2002). Women’s ealth Research, (11), 1461-1476.
perceptions of personal cardiovascular risk and their Wu, Y., hang, Y., Li, Y., Hong, B., Huang, C.
risk-reducing behaviors. American ournal of Critical (2004). Factors associated with the extent of
Care, , 221-227. care-seeking delay for patient with acute myocardial
Richards, H. M., Reid, M. E., Watt, G. C. M. (2002). infarction in Beijing. Chinese Medical ournal, ,
Why do men and women respond differently to chest 1772-1777.
pain ournal of the American Medical omen s erwic, J. J., Ryan, C. J., DeVon, H. A., Drell, M. J.
Association, (2), 79-81. (2003). Treatment seeking for acute myocardial
Rosenfeld, A. G. (2004). Treatment-seeking delay infarction symptoms. Nursing Research, ,
among women with acute myocardial infarction: 159-167.
430 PART FIVE Critiquing Research

INTRODUCTION TO CRITIQUE 2 Significance

The preceding article by Turris (2009) is an Because of the effects of cardiac disease on the
example of a grounded theory study. The article population and the serious repercussion of delay
is critically examined as follows for its rigour as in seeking treatment, this study has the potential
a grounded theory study, its contribution to to in uence nursing practice and the lives of
nursing, and its usefulness in practice. The crite- many women. Although treatment-seeking delay
ria listed in Box 18-2 are used to guide the has been studied extensively, it continues to be a
critique. persistent clinical issue. Some of the explanatory
evidence in the literature is con icting.
Title
The title of the article concisely captures the Method
essence of the study. The method used in this study was clearly identi-
fied as grounded theory. Grounded theory is used
Abstract to explain basic social processes and to develop
The abstract met the requirements of a good theory (Streubert Carpenter, 2011). In this
abstract: It contained the background, the method study, the basic social process under study was
(including the purpose and the sample size), and the process of maintaining integrity in relation to
the findings. The abstract also had a statement visiting an emergency department for treatment
about the clinical relevance of the study. of symptoms indicating potential cardiac illness
(Turris, 2009, p. 6). The grounded theory method
Focus/Topic is appropriate for this study, and the method pro-
The phenomenon of interest was stated clearly in cesses were followed and yielded the expected
the introduction to the study. Turris (2009) was results.
interested in understanding how and why women
interpret their symptoms of a potentially deadly Sampling
cardiac event and in understanding their treat- Turris (2009) stated that purposive sampling
ment-seeking behaviour. Many women delay was used for data collection, which is appropriate
treatment, which is a critical mistake, as life- for a grounded theory study. Purposive or pur-
saving treatment should occur within 1 hour of poseful sampling provides investigators in quali-
the onset of symptoms. Although there is a per- tative research with the opportunity to recruit
sistent belief that women do not suffer from heart participants who are knowledgeable about the
disease, it is the number one killer of women. phenomenon under study because they can
provide rich descriptions of their personal experi-
Purpose ences. In this case, Turris recruited women who
Turris (2009) was interested in developing a sought emergency care for the treatment of symp-
theory to explain women’s treatment-seeking toms of a potential cardiac illness at one of two
delay behaviour in the presence of cardiac symp- hospitals. The inclusion criteria included treat-
toms. Her research question concerned what ment of symptoms that were of a potential cardiac
process guided women as they interpreted symp- event, hemodynamic stability, age of 18 years or
toms suggestive of cardiac illness, chose their older, English speaking, and ability and willing-
treatment options, and subsequently assessed ness to provide informed consent.
their experiences of care in the emergency Turris (2009) also stated that theoretical sam-
department. pling was used after the initial purposive
 CHAPTER 18 Critiquing Qualitative Research 431

sampling. In theoretical sa ling which is used may potentially distort the findings. Glaser (1998)
primarily in grounded theory, individuals are recommended that adjusted conversational
interviewed in an effort to further develop spe- interviewing (p. 173) replace formal questions,
cific aspects of the emerging theory. In this case, whereas Corbin and Strauss (2008) supported
Turris noted that the data suggested that age interview guides with open-ended questions and
might play a role in women’s decisions about opportunities for open dialogue. Because the
seeking treatment, and therefore the accounts of transcripts in Turris’s study were password pro-
younger women were actively sought (p. 6). The tected, readers can assume that a series of inter-
sample expanded from the original 10 partici- view questions were asked. It is not clear whether
pants to 16 in total. Although this was not explic- theoretical notes were used as recommended in
itly stated, it appears that data collection stopped grounded theory methodology.
when no further information was forthcoming or
the data saturation (repetition of information) Data Analysis
was reached. True to grounded theory methodology, Turris
Turris (2009) carefully described how she (2009) conducted the data generation and analy-
recruited the participants and obtained informed sis concurrently. Over a period of several months,
consent. She discussed the setting for recruit- the data were analyzed through open, selective,
ment, identified as one large urban hospital and and theoretical coding. Beginning with open
one community hospital. The smaller community coding, Turris moved to constant comparative
hospital was not able to offer angiographic pro- analysis. Important concepts emerged to explain
cedures often necessary for patients with acute patterns in the women’s accounts. She reported
myocardial infarction. Although Turris carefully that the evidence core category, which occurred
accounted for these differences between the two over and over in the data, was sought throughout
settings, she did not describe how the samples of the analytic process (p. 6).
participants differed from each other in each Turris (2009) dedicated an entire section of
setting. the report to the concept of rigour. She identified
how the study met each of the specific criteria of
Data Generation fittingness, work, relevance, and modifiability.
The primary source of data was a formal audio- For example, fittingness, whereby the theory
taped interview lasting 30 to 90 minutes. Tran- describes the phenomenon under study, was dis-
scripts from the interview were password cussed by the research team on a biweekly basis.
protected, and any identifying information was In addition, the criterion of relevance was met
removed. Turris (2009) was careful to describe inasmuch as each of the emerging themes was
the ethical approval and how informed consent checked with the participants and colleagues.
was obtained. She indicated that participants
were reminded that they could withdraw from Empirical Grounding of the
the study at any time, not respond to particular Study Findings
questions, or request to have the tape recorder The main finding in this study was that partici-
turned off. pants make decisions about seeking treatment on
Turris (2009) did not include the interview the basis of particular ways of knowing. This
questions in the study report. This is typical in pattern is part of a large social phenomenon rather
grounded theory research, in which the use of than a product of individual factors. Participants
interview guides are discouraged because they were reported to be in uenced primarily by the
may be based on preconceived notions that context of their lives as mothers, daughters, and
432 PART FIVE Critiquing Research

wives (Turris, 2009, p. 6) and secondarily by p. 11). The implications for nursing practice are
their assessments of their symptoms. The women important however, more specific practical rec-
used ways of knowing (embodied, temporal, ommendations would have been helpful for the
rational, and relational) to understand and inter- nurse clinician.
pret their symptoms within a larger personal and The research study by Turris (2009) contrib-
social context (p. 6). Turris then described the utes to the body of nursing knowledge and offers
findings in sections focusing on each of new direction in the area of study on delay in
the ways of knowing, which originated from seeking treatment. This area of study can yield
Carper’s seminal work on Fundamental Patterns information that can potentially reduce death and
of Knowing in Nursing published in 1978. morbidity from a largely treatable disease process.
Participants’ quotations supported each of the
identified ways of knowing and supported the Evidence-Informed Practice Tip
main finding that their responsibilities in the roles Qualitative research may generate basic knowl-
of wives and mothers were more important and edge, hypotheses, and theories to be used in the design
vital than attending to their own symptoms of of other types of qualitative or quantitative studies.
cardiac disease. This thorough discussion of the However, qualitative research is not necessarily a pre-
liminary step to another type of research. It is a com-
findings provides credibility that the theory is plete and valuable end in itself.
valid.
In the Discussion section, Turris (2009)
reported the findings in the literature concerning
delay in seeking treatment and claimed that many
CRITICAL THINKING CHALLENGES
are noteworthy. For example, Turris found, in ■ Discuss the similarities and differences between
contrast to previous studies, that rather than the stylistic considerations of reporting a
delaying treatment, consultation with a family qualitative study versus a quantitative study in a
member resulted in an immediate visit to the professional journal.
emergency department. This and other findings ■ Are critiques of qualitative studies by consumers
have potentially significant implications for prac- of research, in the role of either a student or a
tice. Limitations of Turris’s study were appropri- practising nurse, valid? Which type of qualitative
study is the most difficult for consumers of
ately detailed.
research to critique? Discuss what assumptions
Conclusions, Implications, led you to this determination.
Discuss how nurses would go about
and Recommendations ■

incorporating qualitative research in evidence-

In the Implications for Nursing Practice section, informed practice. Give an example.
Turris (2009) indicated that the findings on the
forms of knowledge that women use to make
decisions would allow nurses to have discussions REFERENCES
with women about their decisions before a crisis Appleton, J. V., King, L. (2002). Journeying from the
arises. In addition, understanding delay in seeking philosophical contemplation of constructivism to the
treatment as a social phenomenon might widen methodological pragmatics of health services
the focus of research. nderstanding why women research. ournal of Advanced Nursing, (6),
experiencing symptoms of potential cardiac 641-648.
Barbour, R. S., Barbour, M. (2003). Evaluating and
illness make their decisions can inform’ nursing synthesizing qualitative research: The need to develop
interventions before a cardiac event, preventing a distinctive approach. ournal of Evaluation in
death and reducing morbidity (Turris, 2009, Clinical Practice, , 179-186.
 CHAPTER 18 Critiquing Qualitative Research 433

Carper, B. A. (1978). Fundamental patterns of knowing Morse, J. M., Penrod, J., Hupcey, J. E. (2000).
in nursing. Advanced Nursing Science, , 13-23. ualitative outcome analysis: Evaluating nursing
Cesario, S., Morin, K., Santa-Donato, A. (2002). interventions for complex clinical phenomena.
Evaluating the level of evidence of qualitative ournal of Nursing Scholarship, , 125-130.
research. ournal of bstetric, ynecological and Schepner-Hughes, N. (1992). Death ithout eeping:
Neonatal Nursing, , 708-714. The violence of everyday life in Bra il. Berkeley:
Chiovitti, R., Prian, N. (2003). Rigour and grounded niversity of California Press.
theory research. ournal of Advanced Nursing Straus, S. E., Glasziou, P., Richardson, W. S.,
Practice, (4), 427-435. Haynes, R. B. (2011). Evidence-based medicine: o
Corbin, J., Strauss, A. (2008). Basics of ualitative to practice and teach it (4th ed.). Edinburgh:
research (3rd ed.). Los Angeles: Sage. Churchill Livingstone.
Creswell, J. W. (2009). Research design: ualitative, Strauss, A., Corbin, J. (1990). Basics of ualitative
uantitative, and mi ed methods approaches (3rd research: rounded theory procedures and tech-
ed.). Thousand Oaks, CA: Sage. ni ues. Newbury Park, CA: Sage.
Gibson, B. E., Martin, D. K. (2003). ualitative Streubert, H. J., Carpenter, D. R. (2011). ualitative
research and evidence-based physiotherapy practice. research in nursing: Advancing the humanistic
Physiotherapy, , 350-358. imperative (5th ed.). Philadelphia: Wolters Kluwer.
Glaser, B. G. (1998). Doing grounded theory: Issues Turris, S. (2009). Women’s decisions to seek treatment
and discussions (2nd ed.). Mill Valley, CA: Sociology for the symptoms of potential cardiac disease.
Press. ournal of Nursing Scholarship, (1), 5-12.
Glesne, C. (2011). Becoming ualitative researchers: An Wilkin, K., Slevin, E. (2004). The meaning of caring
introduction (4th ed.). Toronto: Allyn Bacon. to nurses: An investigation into the nature of caring
Horsburgh, D. (2003). Evaluation of qualitative work in an intensive care unit. ournal of Clinical
research. ournal of Clinical Nursing, , 307-312. Nursing, , 50-59.
Ives, M. Melrose, S. (2010). Immunizing children
who fear and resist needles: Is it a problem for
nurses Nursing orum, (1), 29-39.
FOR FURTHER STUDY
Jackson, R. L., Drummond, D. K., Camara, S. Go to Evolve at http://evolve.elsevier.com/
(2007). What is qualitative research ualitative Canada/LoBiondo/Research for Audio Glossary, how-to
Research Reports in Communication, (1), 21-28. instructions for Writing Proposals for Funding, and
Lincoln, Y. S., Guba, E. (1985). Naturalistic in uiry. additional research articles for practice in reviewing and
Beverly Hills, CA: Sage. critiquing.
C H A PTER 1 9

Critiquing Quantitative Research

Nancy E. Kline | Mina D. Singh

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Identify the purpose of the critiquing process for a quantitative research report.
• Describe the criteria of each step of the critiquing process for a quantitative research report.
• Evaluate the strengths and weaknesses of a quantitative research report.
• Discuss the implications of the findings of a quantitative research report for nursing practice.
• Construct a critique of a quantitative research report.

KEY TERM
scientific merit

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

434
 CHAPTER 19 Critiquing Quantitative Research 435

AS REINFORCED THROUGHOUT EACH CHAPTER of this merit of the report. Key to the critique is the
book, it is important not only to conduct and read strength of evidence that each study produces
research but also to use research for evidence- individually and collectively.
informed practice. As nurse researchers increase This chapter presents critiques of two studies
the depth (quality) and breadth (quantity) of in which research questions were tested with dif-
research methods from descriptive research ferent quantitative designs. The critiquing criteria
designs to randomized clinical trials, the data to designed to assist research consumers in judging
support clinical interventions and quality out- the relative value of a research report are found
comes are becoming more readily available. Each at the ends of previous chapters. These critiquing
published study, regardless of its design, re ects criteria have been summarized to create an abbre-
a level of evidence, but the critique of each study viated set of questions that will be used as a
covers much more than the level of evidence pro- framework for the two sample research critiques
duced by the design. When you critique a research (Box 19-1). These critiques exemplify the process
study, examine each component to determine the of evaluating reported research for potential

BOX 19-1
MAJOR CONTENT SECTIONS OF A RESEARCH REPORT AND RELATED CRITIQUING GUIDELINES
PROBLEM STATEMENT AND PURPOSE (SEE CHAPTER 4) HYPOTHESES OR RESEARCH QUESTIONS
1. What is the problem explored in, or the purpose of, (SEE CHAPTER 4)
the research study? 1. What hypotheses or research questions are stated in
2. Does the statement about the problem or purpose the study? Are they appropriately stated?
express a relationship between two or more variables 2. If research questions are stated, are they used in
(e.g., between an independent variable and a addition to hypotheses or to guide an exploratory
dependent variable)? If so, what is the relationship? Is study?
it testable? 3. What are the independent and dependent variables
3. Does the statement about the problem or purpose in the statement of each hypothesis or research
specify the nature of the population being studied? question?
What is it? 4. If hypotheses are stated, is the form of the statement
4. What significance of the problem—if any—has the statistical (null) or research?
investigator identified? 5. What is the direction of the relationship in each
hypothesis, if indicated?
REVIEW OF THE LITERATURE AND THEORETICAL 6. Are the hypotheses testable?
FRAMEWORK (SEE CHAPTERS 2 AND 5)
1. What concepts are included in the review? Of SAMPLE (SEE CHAPTER 12)
particular importance, note which concepts are the 1. How was the sample selected?
independent and dependent variables and how they 2. What type of sampling method is used in the study?
are conceptually defined. Is it appropriate for the design?
2. Does the literature review make the relationships 3. Does the sample reflect the population as identified
among the variables explicit or place the variables in the problem or purpose statement?
within a theoretical or conceptual framework? What 4. Is the sample size appropriate? How is it
are the relationships? substantiated?
3. What gaps or conflicts in knowledge of the problem 5. To what population may the findings be generalized?
are identified? How is this study intended to fill those What are the limitations in generalizability?
gaps or resolve those conflicts?
4. Are the references cited by the author mostly RESEARCH DESIGN (SEE CHAPTERS 10 TO 11)
primary or secondary sources? Give an example of 1. What type of design is used in the study?
each. 2. What is the rationale for the design classification?
5. What are the operational definitions of the 3. Does the choice of design seem logical for the
independent and dependent variables? Do they proposed research problem, theoretical framework,
reflect the conceptual definitions? literature review, and hypothesis?

Continued
BOX 19-1
MAJOR CONTENT SECTIONS OF A RESEARCH REPORT AND RELATED CRITIQUING GUIDELINES—cont’d
INTERNAL VALIDITY (SEE CHAPTER 9) RELIABILITY AND VALIDITY (SEE CHAPTER 14)
1. Discuss each threat to the internal validity of the 1. What type of reliability is reported for each
study. instrument?
2. Does the design have controls at an acceptable level 2. What level of reliability is reported? Is it acceptable?
for the threats to internal validity? 3. What type of validity is reported for each instrument?
4. Does the validity of each instrument seem adequate?
EXTERNAL VALIDITY (SEE CHAPTER 9) Why?
1. What are the limits to generalizability in terms of
external validity? ANALYSIS OF DATA (SEE CHAPTERS 15 AND 16)
1. What level of measurement is used to measure each
RESEARCH APPROACH (SEE CHAPTERS 7 AND 11) of the major variables?
1. Does the research approach fit with the purpose of 2. What descriptive or inferential statistics are reported?
the study? 3. Were these descriptive or inferential statistics
2. Is a mixed-methods approach, if used, appropriate for appropriate for the level of measurement for each
the study? variable?
4. Are the inferential statistics used appropriate for the
METHODS (SEE CHAPTER 13) intent of the hypotheses?
1. What data-collection methods are used in the study? 5. Does the author report the level of significance set for
2. Are the data-collection procedures similar for all the study? If so, what is it?
participants? 6. If tables or figures are used, do they meet the
following standards?
LEGAL/ETHICAL ISSUES (SEE CHAPTER 6) a. They supplement and economize the text.
1. Have the rights of participants been protected? How? b. They have precise titles and headings.
2. What indications are given that informed consent of c. They do not repeat the text.
the participants was ensured?
CONCLUSIONS, IMPLICATIONS, AND
INSTRUMENTS (SEE CHAPTER 13) RECOMMENDATIONS
1. Physiological measurement 1. If hypothesis testing was done, were the hypotheses
a. Is a rationale given for why a particular supported or not supported?
instrument or method was selected? If so, what 2. Are the results interpreted in the context of the
is it? problem or purpose, hypothesis (see this chapter),
b. What provision is made for maintaining the and theoretical framework or literature reviewed?
accuracy of the instrument and its use, if any? 3. What does the investigator identify as possible
2. Observational methods limitations or problems in the study in relation to the
a. Who did the observing? design, methods, and sample?
b. How were the observers trained to minimize 4. What relevance for nursing practice does the
bias? investigator identify, if any?
c. Did the observers have an observational guide? 5. What generalizations are made?
d. Were the observers required to make inferences 6. Are the generalizations within the scope of the
about what they saw? findings or beyond the scope of the findings?
e. Is there any reason to believe that the presence 7. What recommendations for future research are stated
of the observers affected the behaviour of the or implied?
participants?
3. Interviews APPLICATION AND UTILIZATION (SEE CHAPTER 20)
a. Who were the interviewers? How were they 1. Does the study appear to be valid? In other words,
trained to minimize bias? do its strengths for nursing practice outweigh its
b. Is there evidence of any interviewer bias? If so, weaknesses?
what is it? 2. Do other studies have similar findings?
4. Questionnaires 3. What risks or benefits are involved for patients if the
a. What is the type or format of the questionnaires research findings are used in practice?
(e.g., Likert-type, open-ended)? Are they 4. Is direct application of the research findings feasible in
consistent with the conceptual definitions? terms of time, effort, money, and legal/ethical risks?
5. Available data and records 5. How and under what circumstances are the findings
a. Are the records that were used appropriate to the applicable to nursing practice?
problem studied? 6. Should these results be applied to nursing practice?
b. Were the data used to describe the sample or for 7. Would it be possible to replicate this study in another
hypothesis testing? clinical practice setting?
 CHAPTER 19 Critiquing Quantitative Research 437

application to practice, thus extending the research judge which content is most important to com-
base for nursing. For clarification, refer to the municate to the profession. The decision is a
earlier chapters for detailed presentations of the function of the following:
critiquing criteria and explanations of the research The research design: experimental or non-
process. The criteria and examples in this chapter experimental
are applicable to quantitative studies in which The focus of the study: basic or clinical
researchers used experimental, quasiexperimen- The audience to whom the results will be
tal, and nonexperimental research designs that most appropriately communicated
provided levels II, III, and IV evidence. Each journal provides the guidelines for pre-
paring research manuscripts for publication, and
usually the following major headings are essen-
STYLISTIC CONSIDERATIONS tial sections of a research manuscript or research
As an evaluator, you should be aware of several report:
aspects of publishing before you begin to critique Introduction
research studies. First, different journals have dif- Method
ferent publication goals, and they target specific Results
professional nursing specialties. For example, the Discussion
Canadian ournal of Nursing Research publishes Depending on the stylistic considerations related
articles on the conduct or results of research to the author’s preferences and the journal’s
in nursing. The Canadian ncology Nursing requirements, the content included in the research
ournal also publishes research articles however, report is specific to each of the sections just
because its emphasis is broader, this journal mentioned.
also contains clinical and theoretical articles Stylistic variations (as factors in uencing the
relating to knowledge, experience, trends, and presentation of the research study) are very dis-
policies in oncological nursing. Consequently, the tinct features of a research report and can deter
style and content of a manuscript will vary from the focus of evaluating the reported research
according to the type of journal to which it is for scienti c erit that is, judging the overall
being submitted. quality or validity of a study. Constructive evalu-
Second, the author of a research article pre- ation is based on objective appraisal of the study’s
pares the manuscript by using both personal strengths and limitations. This step precedes con-
judgement and specific journal guidelines. Per- sideration of the relative worth of the findings for
sonal judgement refers to the researcher’s exper- clinical application to nursing practice. Judge-
tise that is developed in the course of designing, ments of the scientific merit of a research study
executing, and analyzing the study. As a result of are the hallmark of promoting a sound evidence
this expertise, the researcher is in a position to base for quality nursing practice.
CRITIQUING A QUANTITATIVE RESEARCH STUDY
The study “Older Adults’ Awareness of Community Health and Support Services for Dementia
Care” by Jenny Ploeg and colleagues (2009), is critiqued here. The article is presented in its
CRITIQUE 1 entirety and is followed by the critique on pp. 453–455. (From Canadian Journal of Aging,
28(4), 359–370. © 2009 by Cambridge University Press. Reprinted by permission.)

Older Adults’ Awareness of Community Health and Support

Services for Dementia Care

Jenny Ploeg, MScN, PhD Jennifer Millen Plenderleith

School of Nursing McMaster Centre for Gerontological Studies
Department of Health, Aging and Society McMaster University
McMaster University Hamilton, Ontario
Hamilton, Ontario
The article examines where older adults seek help in caring
for a parent with dementia and the factors associated
Margaret Denton, MA, PhD with their identification of community health and support
Department of Health, Aging and Society
services as sources of assistance. The authors conducted
Department of Sociology
McMaster Centre for Gerontological Studies
telephone interviews, using random digit dialing, of 1,152
McMaster University adults aged 50 and over in the city of Hamilton. Respon-
Hamilton, Ontario dents received a vignette that raised issues related to paren-
tal dementia. In identifying support sources, over 37 per
cent of respondents identified their physician, 33 per cent
Joseph Tindal, PhD identified informal support such as family and neighbors,
Department of Family Relations and Applied Nutrition
University of Guelph
and 31 per cent identified home health services. Only 18 per
Guelph, Ontario cent identified community support services. Female partici-
pants having higher levels of education were more likely to
identify their physician as a source of support. Knowing
Brian Hutchison, MD, MSc, CCFP, FCFP where to find information about community support
Department of Epidemiology and Biostatistics
services was associated with an increased likelihood of
Department of Family Medicine
Centre for Health Economics and Policy Analysis
mentioning physicians and home health services as sources
McMaster University of assistance.
Hamilton, Ontario
KEYWORDS community support services, awareness,
dementia, caregivers, vignette methodology
Kevin Brazil, MA, PhD
Department of Epidemiology and Biostatistics Dementa sic is an increasingly prevalent and
McMaster University
Hamilton, Ontario
important health concern in Canada (Canadian
Study of Health and Aging CSHA Working
Noori Akhtar-Danesh, MSc, PhD Group, 2000) and globally (Wimo, Winblad,
School of Nursing Aguero-Torres, von Strauss, 2003). Most older
McMaster University adults with dementia continue to live in their
Hamilton, Ontario
homes with the support of informal caregivers
(CSHA Working Group, 1994). Informal caregiv-
Jean Lillie, PhD ers of persons with dementia experience a signifi-
Department of Family Relations and Applied Nutrition
University of Guelph cant burden in their roles. Although many
Guelph, Ontario communities have an array of community health

438
 CHAPTER 19 Critiquing Quantitative Research 439

and support services to assist older adults with Dementia also places a significant burden on
dementia and their caregivers, the literature sug- informal caregivers. About half of all people with
gests that these services are underutilized and that dementia in Canada are living in the community,
one of the barriers to their use is the lack of and more than 98 per cent of them have a
awareness of such services (Strain Blandford, caregiver, usually an unpaid family member, rela-
2002). The research literature on service aware- tive, or friend (CSHA Working Group, 1994).
ness has important methodological limitations: in Caregivers of a family member with dementia
particular, acquiescence bias, whereby respon- are more likely to experience chronic health
dents over-report their awareness of services. The problems, depression, and social isolation, com-
purpose of this descriptive study is to describe pared to those caring for cognitively intact elderly
where older adults would turn to for help in (CSHA Working Group, 1994). Given the increas-
response to vignettes or short stories related to ing number of people with dementia, the impact
caring for a parent with dementia, and the socio- on informal caregivers of providing care, and
demographic factors associated with their choice the preference of older adults to age in place
of supports. We used vignette methodology to (Chappell, McDonald, Stones, 2008), increas-
avoid the acquiescence bias so common in service ing attention is being focused on strategies to
awareness research. support caregivers in their roles. One such
strategy involves the use of support provided by
community health and support services.
DEMENTIA
It has been estimated that the global prevalence Community Health and Support Services
of dementia in 2001 was more than 24 million Many communities have a broad array of com-
people aged 60 years or older, a prevalence rate munity health and support services available to
of approximately 3.9 per cent of that age group assist persons with dementia and their caregivers.
(Ferri, Prince, Brayne, Brodaty, Fratiglioni, Such services provide an alternative to institution-
Ganguli et al., 2005). This prevalence is projected alization. Service examples include home health
to double every 20 years to more than 81 million services—such as nursing and homemaking—as
people by 2040 (Ferri et al., 2005). In North well as community support services (CSSs). Com-
America, the prevalence of dementia among munity support services, as defined here, are deliv-
those aged 60 years and older in 2001 was 6.4 per ered in the home or community to assist people
cent (Ferri et al., 2005). The CSHA Working with health or social problems to maintain the
Group (2000) has estimated that there are 60,150 highest possible level of social functioning and
new cases of dementia per year in Canada. Of quality of life. Examples of CSSs are (a) adult day
those older Canadians with dementia, 64 per cent programs, (b) volunteer visiting, (c) caregiver
were diagnosed with Alzheimer’s disease (AD), support programs, (d) food services, (e) transpor-
19 per cent with vascular dementia, and 17 per tation services, and (f) organizations such as the
cent with other forms of dementia (Hill, Forbes, Alzheimer Society. Access to CSSs is particularly
Berthelot, Lindsay, McDowell, 1996). Demen- challenging because of the multiplicity of small
tia has been described as a major burden for agencies providing these services, the lack of a
health and social care systems (Wimo et al., central access point, and the lack of awareness of
2003). The net economic cost of dementia in such services. Further, the complexity of the health
Canada in 1991 was estimated to be at least CAN and social support system makes it challenging for
3.9 billion ( stbye Crosse, 1994), and it is older persons, their families, and health care pro-
likely to be much higher now, 17 years later. fessionals to navigate the system.
440 PART FIVE Critiquing Research

There is some evidence that use of community- and use of community-based services among 293
based services has positive benefits for frail older person-caregiver dyads (44 with cogni-
elders (with and without dementia) and their tive impairment). While they found few people
caregivers ( arit, Gaugler, Jarrott, 1999). In unaware of home-delivered meals (7.9 ),
their review of the literature, arit et al. (1999) in-home nursing (9.5 ), personal care (10.6 ),
found that such services resulted in lower and homemaking (11.2 ), a much larger propor-
levels of care-related stressors, perceived burden, tion was unaware of hospital respite (49.5 ),
depression, and anger. At the same time, patients nursing home respite (47.4 ), in-home respite
with dementia experienced improved life satis- (43.8 ), day hospitals (43.3 ), and day centres
faction and mood, engagement in activities, and (35.7 ). For these latter services (i.e., day centres,
fewer behavioural difficulties. In a qualitative day hospitals, in-home respite, nursing home
study of family caregivers of relatives with AD respite), the second most common reason given
or a related disorder, community services were for non-use was that caregivers were not aware
found to provide benefits including the experi- of the service. In that study, researchers provided
ence of community and support, a gain in knowl- participants with the categories of available com-
edge, receipt of personal renewal, and benefits to munity services.
the patient (Winslow, 2003). Studies of persons with dementia and their
caregivers have also demonstrated limited aware-
ness and utilization of community health and
LITERATURE REVIEW support services (Brodaty et al., 2005 Buono
nderutilization of community-based services— et al., 1999 Caserta, Lund, Wright, Redburn,
that is, the gap between expressed need and service 1987 Collins et al., 1991 Maslow, 1990 Vetter
use—has been recognized as a general problem in et al., 1998). For example, Collins et al. (1991)
the field of aging (Strain Blandford, 2002) and a found that among caregivers of persons with AD
particular problem in dementia care (Brodaty, not using specific services, the percentage who
Thomson, Thompson, Fine, 2005 Buono, were not aware of availability of services varied
Busato, Mazzetto, Paccagnella, Aleotti, anetti by service as follows: (a) support group (10 ),
et al., 1999 CSHA Working Group, 1994 Collins, (b) visiting nurse (18 ), (c) home-delivered
Stommel, Given, King, 1991 Forbes, Morgan, meals (20 ), (d) transportation service (30 ), (e)
Janzen, 2006 Vetter, Steiner, Kraus, Moises, counseling (36 ), (f ) day respite program (39 ),
Kropp, Moller et al., 1998). Research has sug- and (g) temporary overnight care (58 ). Another
gested that community health and support services study found that the main reason reported for
are underused, largely due to lack of awareness or non-use of services (e.g., counseling, support
knowledge of such services (Strain Blandford, groups, Meals on Wheels, and adult day care) was
2002 Vetter et al., 1998). Krout (1983) distin- that over 60 per cent of caregivers of persons with
guished between awareness and knowledge of AD (n = 36) were unaware of the availability of
services. Awareness is a general under-standing such services (Vetter et al., 1998). A third study
that a service exists. Knowledge involves knowl- found that 36 per cent of caregivers of dementia
edge of what the program is or does, where it is patients (n = 597) did not know whether com-
located, or how one gets involved with it (Krout, munity services were available or not (Caserta
1983, p. 155). Most of the following research et al., 1987).
addresses awareness, not knowledge, of services. Some research has been conducted on the
Strain and Blandford (2002), in the Manitoba factors associated with utilization of CSSs for
Study of Health and Aging, studied the awareness dementia care (Brodaty et al., 2005 Caserta
 CHAPTER 19 Critiquing Quantitative Research 441

et al., 1987 Strain Blandford, 2002 Vetter services and concerns such as inconsistency of
et al., 1998), but relatively little on the factors care provider, in exible care, and cost of
associated with awareness of such services. services.
Collins et al. (1991), in their study of family care- In addition to the limited research on factors
givers of patients with AD, found that older care- associated with awareness of community-based
givers were more likely to be uncertain about the services, there are important methodological lim-
availability of services, and that caregiver depres- itations of the service awareness literature, includ-
sion was associated with less knowledge of ing the literature related to dementia care services.
service availability. Our review of the factors In most studies, respondents have been provided
associated with, or assessed for their association with lists of service or agency names and asked
with, the use of CSSs for dementia care in the to indicate whether or not they were aware of or
research literature guided our selection of the using each one (Buono et al., 1999 Caserta et al.,
variables in this study. Specifically, researchers 1987 Collins et al., 1991 Strain Blandford,
have found that socio-economic variables (i.e., 2002 Vetter et al., 1998). This methodology
employment, income, education) (Collins et al., leads to acquiescence bias, the tendency of
1991 Cox, 1999 Ortiz Fitten, 2000), demo- respondents to answer the question positively
graphic variables (i.e., age, sex) (Collins et al., regardless of the content (Calsyn Winter,
1991 Robinson, Buckwalter, Reed, 2005), and 1999). In several studies, Calsyn, Roades, and
social variables (i.e., social support, social net- Calsyn (1992) provided older adults with a ficti-
works) (Caserta et al., 1987 Cotrell Engel, tious service or agency name and found that
1998) are associated with the use of dementia 20–30 per cent of respondents reported familiar-
care services. To date, little evidence exists of ity with that service. One approach to address
associations between the use of dementia care acquiescence bias in studies of service awareness
services and other variables such as language and is to use open-ended questions to solicit the name
disability (Brodaty et al., 2005 Ortiz Fitten, or types of agencies, but this approach has seldom
2000). been used in studies of service awareness.
There is limited Canadian information on The purpose of this study is to measure older
older persons’ awareness of community health persons’ awareness of community health and
and support services for situations of parental support services when presented with a scenario
dementia and the factors associated with such related to caring for a parent with dementia. We
awareness. Instead, the literature has focused on were also interested in the socio-demographic
use, availability, and acceptability of services, and other factors, including care-giving status,
and barriers to their use (Forbes, Markle-Reid, associated with older adults’ identification of
Hawranik, Peacock, Kingston, Morgan et al., community health and support services as sources
2008 Jansen, Forbes, Markle-Reid, Hawranik, of help in caring for a parent with dementia. We
Kingston, Peacock et al., 2009 Strain Bland- expected that caregivers might have greater
ford, 2002). For example, a qualitative study awareness of community health and support ser-
exploring the use and satisfaction with home and vices than non-caregivers, as research has found
community-based services for persons with that most caregivers have a viable informal
dementia from the perspective of family caregiv- network of secondary supports that may help
ers found that issues of availability and accept- address barriers to the use of formal services
ability captured caregivers’ experiences (Forbes (Cotrell Engel, 1998).
et al., 2008). Caregivers talked about the need for Our study addressed the following research
a continuum of home and community-based questions:
442 PART FIVE Critiquing Research

1. Where would older adults turn for assistance hypothetical situations that closely approximate
when faced with a scenario related to caring real-life decision-making situations. Respondents
for a parent with dementia are read the vignettes and asked to respond to the
2. What factors are associated with the identifica- hypothetical situations. Advantages of vignettes
tion of community health and support services are that they are interesting to the respondents,
as a source of assistance in caring for a parent they provide context, and they can be used to
with dementia address sensitive topics such as health care
3. Do caregivers have greater awareness of com- (Hughes Huby, 2002). In the case of research
munity health and support services than related to awareness of community health and
non-caregivers support services, the use of vignettes helps to
avoid acquiescence bias, in which lists of services
are presented to respondents, and investigator
METHOD bias, whereby the list of services is bounded
by the investigators’ awareness of available
The study design was a cross-sectional descrip- services.
tive survey, best used to obtain a description of a The vignettes used in this study were devel-
phenomenon (de Vaus, 2002). Although this oped by front-line service providers to represent
study design can establish association between realistic and familiar situations faced by older
variables, it cannot determine causation (Streiner adults for which community health and support
Norman, 1998). The study involved a tele- services would be appropriate thus, they have
phone survey of adults aged 50 years and older high face and content validity. The vignettes were
residing in the city of Hamilton, Ontario, Canada. pretested and some modifications were made.
Study methods have been described previously
(Denton, Ploeg, Tindale, Hutchison, Brazil, Akhtar- Data Collection
Danesh et al., 2008) and are summarized here. Awareness of community health and support ser-
vices was measured through a telephone survey
Study Setting of older adults. A survey firm was contracted to
Hamilton, Ontario, is Canada’s ninth largest city complete the interviews using a Computer
with a population of nearly 700,000 (Statistics Assisted Dialing Information (CADI) system.
Canada, 2006a). In 2006, Hamilton had a higher Interviews were completed within a six-week
percentage (15.1 ) of adults aged 65 years and period beginning mid-February 2006. English-
older compared to Canada (13.7 ) as a whole speaking residents aged 50 years of age and older
(Statistics Canada, 2006b). Hamilton has an array were invited to participate in the study. The
of community health and support service agen- sample was obtained by randomly selecting tele-
cies available for persons with dementia and their phone numbers from a list of telephone numbers
caregivers. for all residents of Hamilton.
Each participant was read four short vignettes.
Vignette Methodology Participants were asked, If you were in this situ-
We used a vignette methodology to address the ation, what would you do and prompted with
issue of acquiescence bias in the literature. The Anything else up to four times, in order to
use of vignettes or short stories is an established establish multiple sources of assistance. We then
research methodology (Hughes Huby, 2002 asked, Can you name an organization or program
Schoenberg Ravdal, 2000 Spalding Phillips, in our community that you would turn to in that
2007). Vignettes are short descriptions of situation and used up to four prompts until a
 CHAPTER 19 Critiquing Quantitative Research 443

CSS was named. We also collected demographic, The 12 vignettes were divided into three groups
economic, health, and social information about (panels) containing four vignettes a vignette
participants. Participants were asked if they had related to caring for a parent with dementia was
provided any unpaid care or assistance to one or included in each group of four vignettes. The total
more seniors in the past 12 months. If required, sample size was 3 × 384 = 1,152. Thus, 384 older
the following probes were provided as examples adults responded to each vignette, with 1,152 par-
of caregiving: visiting seniors helping them ticipants responding in total.
with shopping, banking, personal care (bathing,
assisting with dressing) and taking medications. Data Analysis
If participants indicated they had provided care, Interviewers entered participants’ responses to
they were considered as caregivers for the pur- the vignettes as verbatim responses. These
pose of this analysis. The study received ethics responses were coded and recoded by the survey
approval from McMaster niversity Research firm’s coders, working collaboratively with the
Ethics Board. research team, into 150 initial categories and then
into 20 meaningful categories for the purpose of
Sample Size analysis (see Table 19C1-2). We had several
Sample size calculation was described in an rounds of peer checking with community partners
earlier paper (Denton et al., 2008). Of the total 12 to ensure that the reduction from 150 to 20 cat-
vignettes used in the 2008 study, three vignettes egories was accurate and appropriate. The survey
addressed issues of caring for a parent with
dementia (see Table 19C1-1), and the responses
to these vignettes are reported in this paper. A TABLE 19C1-2
sample size of 384 was needed for each vignette. CATEGORIES OF PARTICIPANT RESPONSES
TO VIGNETTES
1. Community support services
TABLE 19C1-1 2. Spouse
3. Son/daughter
VIGNETTES
4. Friends and neighbours
VIGNETTE 5. Relatives
NUMBER PANEL VIGNETTE 6. Physician
1 A You are the main caregiver for your 7. Emergency
parent who has Alzheimer’s 8. Clinics/hospitals
disease. You have discovered that 9. Other health professionals
your mother has been taking 10. Non-health professionals
more pills than she should. 11. Pastor/clergy/faith community
2 B You are an only child of a parent 12. Social and recreation services
with Alzheimer’s disease. For 13. Nothing
years you have been bringing 14. Home health services
him meals, doing his laundry, 15. Long-term care/residential care
and paying his bills. Your spouse 16. Self-help/personal strategy
is sick, and now you have to 17. Government
help him/her, too. You are feeling 18. Information and referral sources
overwhelmed and frustrated. 19. Disease-specific agencies
3 C Your mother, who lives with you, is 20. Community Care Access Centre
very confused and can’t be left Note: For the purpose of this analysis: Informal supports included
alone. You want to keep her at spouse, son/daughter, friends, neighbours, relatives, and self-help
home, but you have to go to strategies. Home health services included home health services and
work. The rest of the family are Community Care Access Centre. Community support services included
working and cannot help. 37 agencies such as adult day programs, Alzheimer Society,
transportation ser-vices, and Meals on Wheels.
444 PART FIVE Critiquing Research

firm provided a Statistical Package for the Social Centre CCAC which is a one-stop access centre
Sciences (SPSS) file of the data to the for home health services covered under the
researchers. Ontario Health Insurance Plan) (c) CSSs (d)
Age was measured as a categorical variable long-term or institutional care and (e) their phy-
(age 50–60, age 61–70, and 71 and older, with sician. These responses were given by 81.3 per
age 50–60 as the reference category). Sex was cent of participants as their first response.
measured as 1 for females and 0 for males (refer- To answer the second question, What factors
ence category). Marital status was measured as a are associated with the identification of commu-
categorical variable but was recoded to 1 = nity health and support services as a source of
married and 0 = not married (reference category) assistance in caring for a parent with dementia ,
for the purpose of the regression analysis. Four we used logistic regression, as the identification
levels of education were measured: (a) less than was measured as a dichotomous variable. We
high school (reference category), (b) some high included the following variables in the regression
school or graduated from high school, (c) other analysis: age, sex, education, country of birth,
post-secondary education including community self-rated health, marital status, membership in
college and apprenticeship to the trades, and (d) clubs or organizations, and functional limitations
university or higher education. Income was mea- of the care recipient. The inclusion of caregiver
sured in four categories. Being foreign born was status as a variable in the regression analysis per-
measured as 1 = yes and 0 = born in Canada mitted us to address the third question: Do care-
(reference category). Self-rated health was mea- givers have greater awareness of community
sured as a categorical variable (excellent or very health and support services than non-caregivers
good as the reference category good, and fair or sing logistic regression, we regressed the iden-
poor). The respondent’s functional health was tification of services on the variables just
measured as having a limitation at home or identified.
outside the home (1 = yes, 0 = no as the reference Odds ratios are presented. An odds ratio greater
category). Membership in clubs or voluntary than 1.0 indicates an increased likelihood of iden-
organizations such as seniors centres, church, or tification of services a 95 per cent confidence
social groups was measured as 1 = member and interval of an odds ratio that does not include 1.0
0 = non-member (reference category). Caregiving indicates a statistically significant result.
status was measured as 1 = caregiver for a senior
in the past year and 0 = non-caregiver (reference
category). Participants were asked where they RESULTS
would find information about CSSs with three
opportunities to respond. The number of sources Participants
of information was summed and ranged from 0 A total of 22,072 different telephone numbers in
to 3. Hamilton were called, and 15,857 households
To answer the first question, Where would were contacted. We were unable to contact 6,215
older adults turn for assistance when faced with households because either the number was not in
a scenario related to caring for a parent with service, the line was busy, an answering machine
dementia , we focused on the five most frequent took the call, or there was no answer. Of the
responses given by respondents: (a) informal 15,857 households contacted, 10,373 had a resi-
sources including family, friends, and neighbours dent aged 50 years or older. Following exclusion
as well as self-help strategies (b) home health of 1,034 households due to language barriers and
services (including the Community Care Access illness, and 8,180 refusals, we conducted 1,159
 CHAPTER 19 Critiquing Quantitative Research 445

interviews seven of these were removed from the between caregivers and non-caregivers on the
database due to incomplete data, leaving 1,152 variables of sex, country of birth, language, and
usable interviews, for a response rate of 12.4 per marital status (see Table 19C1-3). However, care-
cent (1,159/9,339) of eligible households. givers were younger, had higher education and
Participants represented a wide cross-section income levels, better self-rated health, more
of older adults living in Hamilton. The demo- information sources, and were less likely to have
graphic profile of the sample is described in Table functional limitations than non-caregivers.
19C1-3. Over two thirds of participants were Most caregivers (54 ) provided care to one
female and most were married (63 ), over the person, while 46 per cent provided care to two or
age of 60 (57 ), and born in Canada (71 ). more persons. For the first care recipient men-
Almost half had high school education or less tioned, the care recipient was most likely to be a
(47 ), and 28 per cent had university education. parent (46 ) or friend (31 ). Most of the first
Household income varied with the most frequent care recipients (88 ) described by the caregivers
category being CAN 60,000 or more (39 ). A had a physical or mental condition or a health
high proportion reported excellent, good, or very problem that reduced the amount or kind of activ-
good self-rated health (54 ), and most reported ity they could perform. Most caregivers provided
no functional limitations (60 ). A high propor- care daily (26 ) or at least once per week (45 ).
tion (56 ) reported two or more information
sources. A comparison of our sample to Hamilton
data from the 2006 and 2001 Census of Popula- Services and Supports Identified
tion was presented earlier (Denton et al., 2008). by Participants
Our sample included a higher proportion of Across all three vignettes, the percentage of par-
females, people who were Canadian born, and ticipants who identified each type of community
people with incomes higher than the Hamilton health and support service as a source of assis-
population had (Denton et al., 2008). tance in a situation of parental dementia is illus-
Analysis was conducted on the full sample trated in Figure 19C1-1. The bottom part of each
(n = 1,152) instead of the three separate groups bar illustrates the percentage of respondents by
responding to each vignette. To ensure that the first response, and the top part of each bar illus-
three groups were not significantly different on trates the percentage of respondents who men-
demographic variables, a chi-square test was con- tioned a specific care source after prompting.
ducted to compare each demographic variable When faced with a situation of parental dementia,
between the three groups. No significant differ- the highest percentage of participants identified a
ences were found between groups on any of the physician and the physician’s office staff as
demographic variables (education, sex, marital sources of support. This source was named by 25
status, language, country of birth, health) except per cent of the respondents as their first choice,
age: χ2 (4, n = 1,152) = 11.2, p = 0.03. Based on and overall (i.e., multiple response) by 37 per
these findings, our groups were similar enough to cent of respondents. This response was closely
combine into one group for analysis. followed by those who mentioned informal
Of the total sample of 1,152 respondents, 474 sources (20 first response, 34 overall) and
(41 ) identified themselves as having provided home health services (19 first response, 31
care to a senior in the previous 12 months. A chi- overall). Only nine per cent of participants men-
square test was conducted to compare each vari- tioned a CSS as their first choice (18 overall),
able between caregiver and non-caregiver groups. and eight per cent mentioned long-term care as
There were no statistically significant differences their first choice (13 overall).
446 PART FIVE Critiquing Research

TABLE 19C1-3
DEMOGRAPHIC DESCRIPTION OF PARTICIPANTS
TOTAL SAMPLE CAREGIVERS NON-CAREGIVERS
DEMOGRAPHIC VARIABLE (n = 1,152) % (n = 474) % (n = 678) % CHI SQUARE

AGE*
50–60 42.6 49.2 38.1 χ2 (2, n = 1,152) = 23.1, p = 0.000
61–70 29.5 30.2 29.1
71+ 27.9 20.7 32.9
GENDER
Male 28.7 26.8 30.1 χ2 (1, n = 1,152) = 1.5, p = 0.224
Female 71.3 73.2 69.9
EDUCATION*
Less than high school 5.0 2.3 6.9 χ2 (3, n = 1,142) = 18.9, p = 0.000
Some or all of high school 41.9 39.1 44.0
Community college non-university 25.4 26.5 24.6
certificate, trade
University or higher 27.7 32.1 24.6
COUNTRY OF BIRTH
Born in Canada 71.4 72.4 70.6 χ2 (1, n = 1,142) = 0.4, p = 0.529
Foreign born 27.8 26.8 28.5
LANGUAGE
English 94.6 95.4 94.1 χ2 (2, n = 1,152) = 1.0, p = 0.593
French 0.6 0.6 0.6
Other 4.8 4.0 5.3
MARITAL STATUS
Married, common law 63.1 67.2 60.2 χ2 (3, n = 1,151) = 7.1, p = 0.068
Widowed 19.2 16.5 21.1
Divorced, separated 11.6 9.9 12.7
Single, never married 6.5 6.3 6.0
HOUSEHOLD INCOME ($)*†
$20, 000 or less 15.0 9.6 18.8 χ2 (4, n = 912) = 24.9, p = 0.000
$20, 001–$40, 000 27.6 25.5 29.1
$40, 001–$60, 000 18.6 18.4 18.8
$60, 001–$80, 000 16.7 21.3 13.4
$80,001+ 22.0 25.3 19.8
SELF-REPORTED HEALTH*
Excellent, very good 54.5 60.7 50.1 χ2 (2, n = 1,147) = 13.5, p = 0.001
Good 28.2 25.6 30.1
Fair, poor 17.3 13.7 19.7
FUNCTIONAL LIMITATIONS*
No 60.2 66.7 55.9 χ2 (1, n = 1,137) = 12.6, p = 0.000
Yes 39.8 33.6 44.1
NUMBER OF INFORMATION SOURCES*
0 13.8 9.5 16.8 χ2 (3, n = 1,152) = 31.6, p = 0.000
1 29.8 26.8 31.9
2 30.6 30.2 30.8
3 25.9 33.5 20.5
Number of study respondents varies across independent variables due to missing data.
*Significant difference found between caregiver and non-caregiver groups.
†
The number of study responses to the household income was 376 (caregiver), 536 (non-caregiver), 912 (both males and females) respectively.
 CHAPTER 19 Critiquing Quantitative Research 447

100 informal sources such as family, friends, and

Multiple Response
90 neighbours decreased significantly with age.
First Response
80
Looking at the effects of sex, the odds of identify-
ing their physicians as a source of support
70
increased by 40 per cent if caregivers were female
60 rather than male.
% 50 The odds of mentioning their physician as a
40 source of support increased by 250 per cent for
11.9 those participants with higher levels of education
30
13.4 11.8 (i.e., community college, trade school, or univer-
20
9 sity) compared to those with less than high school
10 20.1 25.3 5.2 18.8 education. Knowing where to obtain information
7.7 9.4
0 about CSSs increased the likelihood of mention-
Informal Physician LTC CSS HHS ing physicians and home health services as sup-
Sources
ports for situations of parental dementia. Married
FIGURE 19C1-1 Percentage of respondents stating where participants had lower odds of identifying
they would turn for help by first and multiple responses.
[CSS: community support services; HHS: home health informal sources of support for situations of
service; LTC: long-term care.] parental dementia, compared to those who were
unmarried.

Factors Associated with Identification of

Sources of Help DISCUSSION
Next, we turn to the socio-demographic and other Our study results provide insight into where older
factors that may be associated with the choice of Canadians would turn for help when faced with
support service. Here we focus on the overall a situation of parental dementia and the factors
response, that is, when participants mentioned a associated with their choice of supports. When
home health or CSS at any point in their answer, older adults were presented with vignettes that
either as their first response or in response to the described situations of parental dementia, they
prompt Anything else . Table 19C1-4 presents most frequently indicated they would turn to their
the odds ratios and 95 per cent confidence inter- physician, informal sources of help such as family
vals for each variable by type of community and friends, home health services, CSSs, and
health or support service: (a) informal sources, long-term care services. Many CSSs are specifi-
(b) physician, (c) long-term care, (d) CSSs, and cally targeted to assist older adults and their care-
(e) home health services. givers, yet only 18 per cent of middle-aged or
Findings from the logistic regressions showed older adults mentioned these services as sources
the following explanatory variables not to be sta- of help in situations of parental dementia.
tistically significantly associated with any type of The most frequently mentioned CSS was the
community health or support service: country of Alzheimer Society, mentioned by an average of
birth, self-rated health, membership in clubs or 18 per cent of respondents across the three
organizations, being a caregiver, and having any vignettes and the 10 opportunities to respond.
functional limitations in or outside the home. In These results are much lower than the 50 per cent
contrast, holding all other explanatory variables to 92 per cent of participants who were aware of
within the model constant when faced with a situ- selected community health and support services
ation of parental dementia, the odds of turning to in the study by Strain and Blandford (2002). It is
448 PART FIVE Critiquing Research

TABLE 19C1-4
LOGISTIC REGRESSION: COMMUNITY HEALTH AND SUPPORT SERVICES AND ASSOCIATIONS
WITH DETERMINANTS
COMMUNITY
INFORMAL LONG-TERM SUPPORT HOME HEALTH
SOURCES PHYSICIAN CARE SERVICES SERVICES
VARIABLES OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)

AGE (YEARS)
50–60 (ref) 1.0 1.0 1.0 1.0 1.0
61–70 0.7 (0.5–1.0) 0.8 (0.6–1.1) 1.3 (0.9–2.0) 1.0 (0.7–1.5) 1.1 (0.8–1.5)
71+ 0.6** (0.4–0.8) 0.8 (0.6–1.2) 1.2 (0.7–2.0) 0.6 (0.4–1.0) 1.2 (0.8–1.7)
GENDER
Male (ref) 1.0 1.0 1.0 1.0 1.0
Female 1.0 (0.7–1.3) 1.4* (1.1–1.9) 1.0 (0.7–1.4) 1.3 (0.9–1.9) 1.4 (1.0–1.8)
EDUCATION
Less than high school (ref) 1.0 1.0 1.0 1.0 1.0
Some or all high school 1.0 (0.5–1.9) 3.0** (1.4–6.5) 1.0 (0.4–2.6) 1.0 (0.4–2.2) 0.9 (0.5–1.8)
Community college; 0.9 (0.5–1.8) 3.6** (1.6–8.2) 1.5 (0.6–3.7) 1.1 (0.5–2.7) 1.1 (0.5–2.3)
non-university; trade
University or higher 0.9 (0.5–1.7) 3.5** (1.6–7.9) 1.2 (0.5–3.2) 1.6 (0.7–3.9) 1.4 (0.7–2.7)
COUNTRY OF BIRTH
Canadian (ref) 1.0 1.0 1.0 1.0 1.0
Foreign born 1.2 (0.9–1.6) 0.9 (0.7–1.2) 1.2 (0.8–1.8) 0.9 (0.7–1.3) 0.9 (0.6–1.2)
SELF-RATED HEALTH
Excellent, very good (ref) 1.0 1.0 1.0 1.0 1.0
Good 1.0 (0.7–1.4) 0.9 (0.7–1.2) 1.4 (0.9–2.1) 0.7 (0.5–1.0) 1.0 (0.7–1.4)
Fair, poor 1.0 (0.7–1.6) 0.9 (0.6–1.4) 1.0 (0.5–1.7) 0.7 (0.4–1.2) 0.7 (0.5–1.1)
MEMBERSHIP
No (ref) 1.0 1.0 1.0 1.0 1.0
Yes 0.9 (0.7–1.2) 0.9 (0.7–1.2) 0.9 (0.6–1.4) 1.3 (1.0–1.8) 1.3 (1.0–1.7)
CAREGIVER
No (ref) 1.0 1.0 1.0 1.0 1.0
Yes 0.9 (0.7–1.1) 1.1 (0.9–1.5) 1.3 (0.9–1.9) 1.1 (0.8–1.6) 1.3 (1.0–1.8)
†
SUM OF INFORMATION SOURCES
Continuous variable (0–3) 1.2 (1.0–1.3) 1.2** (1.1–1.4) 0.9 (0.8–1.1) 1.1 (1.0–1.3) 1.3*** (1.2–1.5)
MARRIED
No (ref) 1.0 1.0 1.0 1.0 1.0
Yes 0.7* (0.5–0.9) 1.0 (0.7–1.3) 1.5 (1.0–2.2) 1.0 (0.7–1.4) 1.0 (0.8–1.4)
FUNCTIONAL LIMITATION—IN OR OUTSIDE HOME
No (ref) 1.0 1.0 1.0 1.0 1.0
Yes 1.0 (0.8–1.4) 1.4 (1.0–1.9) 1.1 (0.7–1.6) 1.2 (0.8–1.7) 0.8 (0.6–1.1)
Chi-Square 22.664 44.000*** 13.239 38.083*** 61.465***
−2 Log Likelihood 1,399.157 1,430.218 844.508 1,037.466 1,315.740
Cox & Snell R Square 0.020 0.039 0.012 0.034 0.054
Nagelkerke R Square 0.028 0.053 0.022 0.054 0.076
Overall Percentage Correctly 66.3 62.5 87.1 81.2 69.1
Predicted
* p < 0.05, ** p < 0.01, *** p < 0.001
OR = odds ratio
† The OR represents each unit increase in the variable.
 CHAPTER 19 Critiquing Quantitative Research 449

possible that our vignette methodology, by avoid- (2008), in using data from this project for her PhD
ing acquiescence bias, provides a more accurate studies, found that people who provide care are
estimate of older adults’ awareness of CSSs. much more likely to have knowledge of CSSs
The highest percentage of participants (37 ) than those who do not provide care.
indicated they would turn to their physician for Several factors limit the generalizability of
help in situations where a parent has dementia. study findings. First, we had a low response rate
These results are similar to previous research to the telephone survey, consistent with other
findings that 40 per cent of adults aged 65 and studies of access to services (Calsyn Winter,
older spontaneously mentioned the facilitating 2000). The barrage of telemarketing and the use
role played by physicians in formal, community- of caller ID and telephone answering services
based or home health services (Schoenberg, make it difficult to achieve high response rates
Campbell, Johnson, 1999). However, other in telephone interviews. Our sample over-
researchers have found that physicians have represented older adults who were female, Cana-
insufficient information about available services dian born, and had high levels of education. As a
for dementia (Bruce Paterson, 2000 Fortinsky, result, the levels of awareness of community
1998) and that they do not refer patients to support health and support services found in this study are
services early enough, despite prolonged and likely in ated, since previous research has shown
often severe caregiver stress (Bruce Paterson, that women and those with higher education
2000 Bruce, Paley, nderwood, Roberts, levels have higher levels of service awareness
Steed, 2002). One of the issues is that CSSs for (Calsyn Roades, 1993 Calsyn, Roades,
dementia care are not currently integrated within Klinkenberg, 1998). Further, we have limited
a privately or publicly funded system. For information about the caregivers in this study. For
example, accessing home health services or a example, we do not know if they were caring for
family doctor does not necessarily provide a link a person with dementia. Although we studied
to CSSs for dementia care. Integrated access to older adults from one city only, limiting our
such services would likely benefit older adults ability to generalize our results, we believe that
and their caregivers. our findings have wider applicability.
The analysis revealed that few of the socio- The largest proportion of these older adults
demographic and other determinants we assessed indicated they would turn to their physicians and
were significantly associated with identification informal supports as sources of help in situations
of CSSs, home health services, long-term care, of parental dementia. Research has found that
physicians, and informal sources as sources of both professionals and informal supports play
support. The use of physicians as sources of help important mediating roles in linking dementia
for situations of parental dementia may be less caregivers to formal services (Cotrell Engel,
likely for some groups than others, including men 1998). Physicians take on a range of important
and those with less education. The use of infor- mediation roles such as ordering services, provid-
mal sources of help such as family and neigh- ing linkages or facilitating connections between
bours may decrease with age. Knowing where to services/agencies, advising or recommending ser-
look for information about CSSs was associated vices, supplying information on services, and
with increased likelihood of mentioning physi- providing reassurance about services (Schoen-
cians and home health services as potential berg et al., 1999). There is a need to develop and
sources of assistance. Further research is needed evaluate strategies to help physicians and other
on the factors associated with identification of health care providers to improve the links between
dementia care services. For example, Lillie older adults and their caregivers with appropriate
450 PART FIVE Critiquing Research

CSSs. Such strategies may include educational particular, is troubling, given that the very purpose
initiatives, promoting effective interprofessional of these services is to help people retain social
teamwork and collaboration, and the use of tech- functioning and quality of life in the community.
nology such as Internet and email to provide Action must be taken to improve the ability of
information specific to available community sup- physicians and other health care providers to help
ports for persons with dementia and their caregiv- make these linkages possible.
ers (Cantegreil-Kallen, Turbelin, Angel, Flahault,
Rigaud, 2006 Fortinsky, 2001).
Our study results suggest that efforts should ACKNOWLEDGEMENTS
also be made to increase the awareness of older We thank the following organizations and their
persons and their caregivers related to available employees for the valuable contribution that they
services for situations of parental dementia. have made in the development and analysis of
Harris, Bayer, and Tadd (2002) suggested that this research: Catholic Family Services of
older persons may differ in their preferred ways Hamilton—Linda Dayler Coalition of Commu-
of obtaining information and that multiple nity Health and Support Services—Lynne
approaches should be used including health pro- Edwards Community Information Hamilton—
fessionals, organizations providing information Lesley Russell Seniors Activation Maintenance
and advice, family and friends, lea ets or written Program—Lynne Edwards and Dave Banko
documents, television and radio or video, tele- Grocer-Ease—Bev Morgan Hamilton Commu-
phone, and the Internet. Some Ontario communi- nity Care Access Centre—Sherry Parsley, Tom
ties are introducing 211 as a telephone information Peirce, and Dianne Thompson Ontario Commu-
service, and this may be a promising approach to nity Support Association (OCSA)—Susan Thorn-
the lack of awareness (www.211canada.ca). Par- ing and Taru Virkamaki Regional Geriatric
ticular efforts should be made to address the Program (Central)—David Jewell Social Plan-
information needs of immigrant and culturally ning and Research Council of Hamilton (SPRC)—
diverse groups. Don Jaffray and nited Way of Burlington
In our evolving program of research in this and Greater Hamilton—Monica uinlan. This
area, we are conducting a study on how primary research was funded by Canadian Institutes of
care physicians and allied health care providers Health Research—Institute on Aging Ontario
working in their offices help to link older adults Ministry of Health and Long-term Care nited
to CSSs. We are also writing a paper that exam- Way of Burlington and Greater Hamilton and
ines how useful a social-determinants-of-health Social Sciences and Humanities Research Council
model is in predicting knowledge of community through funding of the MCRI (Social and Eco-
health services. nomic Dimensions of an Aging Population).

REFERENCES
CONCLUSION
Brodaty, H., Thomson, C., Thompson, C., et al. (2005).
Lack of awareness of where to turn for help when Why caregivers of people with dementia and memory
faced with situations of parental dementia is a loss don’t use services. International ournal of
serious problem that needs to be addressed so that eriatric Psychiatry, , 537-546.
Bruce, D. G., Paley, G. A., nderwood, P. J., Roberts,
older adults may continue to live in their own
D., Steed, D. (2002). Communication problems
homes using available supports, and so that care- between dementia carers and general practitioners:
givers avoid unnecessary burden. The lack of Effect on access to community support services.
awareness of community support services, in Medical ournal of Australia, , 186-188.
 CHAPTER 19 Critiquing Quantitative Research 451

Bruce, D. G., Paterson, A. (2000). Barriers to Cotrell, V., Engel, R. J. (1998). The role of secondary
community support for the dementia carer: A supports in mediating formal services to dementia
qualitative study. International ournal of eriatric caregivers. ournal of erontological Social or ,
Psychiatry, , 451-457. (3/4), 117-132.
Buono, M. D., Busato, R., Mazzetto, M., Paccagnella, Cox, C. (1999). Race and caregiving: Pattern of service
B., Aleotti, F., anetti, O., et al. (1999). Community use by African American and white caregivers of
care for patients with Alzheimer’s disease and persons with Alzheimer’s disease. ournal of
non-demented elderly people: se and satisfaction erontological Social or , (2), 5-19.
with services and unmet needs in family caregivers. Denton, M., Ploeg, J., Tindale, J., Hutchison, B., Brazil,
International ournal of eriatric Psychiatry, , K., Akhtar-Danesh, N., et al. (2008). Where would
915-924. you turn for help Older adults’ awareness of
Calsyn, R. J., Roades, L. A. (1993). Predicting community support services. Canadian ournal on
perceived service need, service awareness, and Aging, , 359-370.
service utilization. ournal of erontological Social de Vaus, D. (2002). Survey research. In T. Greenfield
or , (1), 59-76. (Ed.), Research methods for postgraduates (2nd ed.,
Calsyn, R. J., Roades, L. A., Calsyn, D. S. (1992). pp. 172-182). New York: Oxford.
Acquiescence in needs assessment studies of the Ferri, C. P., Prince, M., Brayne, C. H., Fratiglioni, L.,
elderly. The erontologist, , 246-252. Ganguli, M., et al. (2005). Global prevalence of
Calsyn, R., Roades, L. A., Klinkenberg, W. D. dementia: A Delphi consensus study. ancet, ,
(1998). sing theory to design needs assessment 2112- 2117.
studies of the elderly. Evaluation and Programming Forbes, D. A., Markle-Reid, M., Hawranik, P., Peacock,
Planning, , 277-286. S., Kingston, D., Morgan, D., et al. (2008). Availabil-
Calsyn, R., Winter, J. (1999). nderstanding and ity and acceptability of Canadian home and
control-ling response bias in needs assessment community-based services: Perspectives of family
studies. Evaluation Revie , , 399-417. caregivers of persons with dementia. ome ealth
Calsyn, R. J., Winter, J. P. (2000). Predicting Care Services uarterly, , 75-99.
different types of service use by the elderly: The Forbes, D. A., Morgan, D., Janzen, B. L. (2006).
strength of the behavioral model and the value of Rural and urban Canadians with dementia: se of
interaction terms. ournal of Applied erontology, health care services. Canadian ournal on Aging, ,
, 284-303. 321-330.
Canadian Study of Health and Aging Working Group. Fortinsky, R. H. (1998). How linked are physicians to
(1994). Patterns of caring for people with dementia community support services for their patients with
in Canada. Canadian ournal on Aging, , dementia ournal of Applied erontology, ,
470-487. 480-498.
Canadian Study of Health and Aging Working Group. Fortinsky, R. H. (2001). Health care triads and dementia
(2000). The incidence of dementia in Canada. care: Integrative framework and future directions.
Neurology, , 66-73. Aging Mental ealth, (Suppl. 1), S35-S48.
Cantegreil-Kallen, I., Turbelin, C., Angel, P., Flahault, Harris, M., Bayer, A., Tadd, W. (2002). Addressing
A., Rigaud, A. (2006). Dementia management in the information needs of older patients. Revie s in
France: Health care and support services in the Clinical erontology, , 5-11.
community. Dementia, , 317-326. Hill, G., Forbes, W., Berthelot, J. M., Lindsay, J.,
Caserta, M. S., Lund, D. A., Wright, S. D., Redburn, McDowell, I. (1996). Dementia among seniors.
D. E. (1987). Caregivers to dementia patients: The ealth Reports, (2), 7-10.
utilization of community services. The erontologist, Hughes, R., Huby, M. (2002). The application of
, 209-214. vignettes in social and nursing research. ournal of
Chappell, N., McDonald, L., Stones, M. (2008). Advanced Nursing, , 382-386.
Aging in contemporary Canada (2nd ed.). Toronto, Jansen, L., Forbes, D. A., Markle-Reid, M., Hawranik,
Ontario, Canada: Pearson-Prentice Hall. P., Kingston, D., Peacock, S., et al. (2009).
Collins, C., Stommel, M., Given, C. W., King, S. Formal care providers’ perceptions of home and
(1991). Knowledge and use of community services community-based services: Informing dementia care
among family caregivers of Alzheimer’s disease quality. ome ealth Care Services uarterly, ,
patients. Archives of Psychiatric Nursing, , 84-90. 1-23.
452 PART FIVE Critiquing Research

Krout, J. A. (1983). Knowledge and use of services by areas and census agglomerations—100 Data.
the elderly: A critical review of the literature. Retrieved September 30, 2008, from http://
International ournal of Aging and uman Develop- www12.statcan.ca/english/census06/data/highlights/
ment, , 153-167. agesex/pages/Page.cfm Lang E Geo CMA Code
Lillie, J. (2008). lder adults no ledge of community 01 Table 1 Data Count Sex 1 Sta rtRec 1
support services: Does social support ma e a Sort 7 Display Page CSDFilter 5000
difference npublished doctoral dissertation. Statistics Canada. (2006b). Population and dwelling
niversity of Guelph, Ontario, Canada. counts, for Canada and census subdivisions (munici-
Maslow, K. (1990). Linking persons with dementia to palities), 2006 and 2001 censuses. Retrieved
appropriate services: Summary of an OTA study. September 30, 2008, from http://www12.statcan.ca/
Pride Institute ournal of ong Term ome ealth english/census06/data/popdwell/Table.cfm T 201 S
Care, , 42-50. 3 O D RPP 150
Ortiz, F., Fitten, L. J. (2000). Barriers to healthcare Strain, L., Blandford, A. (2002). Community-based
access for cognitively impaired older Hispanics. services for the taking but few takers: Reasons for
Al heimer Disease and Associated Disorders, , nonuse. ournal of Applied erontology, , 220-235.
141-150. Streiner, D. L., Norman, G. R. (1998). PD
stbye, T., Crosse, E. (1994). Net economic costs of Epidemiology (2nd ed.). Hamilton, Ontario, Canada:
dementia in Canada. Canadian Medical Association B.C. Decker.
ournal, , 1457-1464. Vetter, P., Steiner, O., Kraus, S., Moises, H., Kropp, P.,
Robinson, K. M., Buckwalter, K. C., Reed, D. Moller, W. D., et al. (1998). Factors affecting the
(2005). Predictors of use of services among dementia utilization of home-care supports by caregiving
caregivers. estern ournal of Nursing Research, , relatives of Alzheimer patients. Dementia and
126-140. eriatric Cognitive Disorders, , 111-116.
Schoenberg, N. E., Campbell, K. A., Johnson, M. M. Wimo, A., Winblad, B., Aguero-Torres, H., von
(1999). Physicians and clergy as facilitators of formal Strauss, E. (2003). The magnitude of dementia
services for older adults. ournal of Aging Social occurrence in the world. Al heimer Disease
Policy, , 9-26. Associated Disorders, , 63-67.
Schoenberg, N. E., Ravdal, H. (2000). sing Winslow, B. W. (2003). Family caregivers’ experiences
vignettes in awareness and attitudinal research. with community services: A qualitative analysis.
International ournal Social Research Methodology, Public ealth Nursing, , 341-348.
, 63-74. arit, S. H., Gaugler, J. E., Jarrott, S. E. (1999).
Spalding, N. J., Phillips, T. (2007). Exploring the use seful services for families: Research findings and
of vignettes: From validity to trustworthiness. directions. International ournal of eriatric
ualitative ealth Research, , 954-962. Psychiatry, , 165-181.
Statistics Canada. (2006a). Age and sex, 2006 counts
for both sexes, for Canada and census metropolitan
 CHAPTER 19 Critiquing Quantitative Research 453

INTRODUCTION TO CRITIQUE 1 Review of Literature and Definitions

The article Older Adults’ Awareness of Com- Ploeg and colleagues’ (2009) article contained a
munity Health and Support Services for Demen- formal section titled Literature Review, even
tia Care (Ploeg et al., 2009) is examined here in though literature about dementia and community
terms of its quality and the potential usefulness health and support services is also cited later in
of the findings for application to nursing practice. the article. The formal section of the literature
The design of this study is level VI, inasmuch as review guides the reader through three main
it is a descriptive study. areas: (1) the concept of underutilization, high-
lighting the difference between awareness and
knowledge (2) factors associated with utiliza-
Title tion of CSSs and (3) methodological limitations
The title of the article captured the essence of the in previous studies.
study succinctly. Ploeg and colleagues (2009) noted the gaps
in each section. Regarding underutilization of
community-based services, they suggested that
Abstract most of the research addresses awareness, not
The abstract met the requirements of a good knowledge, of services (p. 361). nder factors
abstract it included the purpose, the sample size, associated with utilization, they further noted that
the method, the findings, and a short concluding very little research on the factors associated with
statement. awareness of CSSs for dementia care had been
conducted. In regard to methodological limita-
tions, they noted that acquiescence bias—in
Problem and Purpose which a respondent answers questions on a survey
Ploeg and colleagues (2009) stated that the overall in a positive manner without paying attention to
objective of the study was to measure older the content—was a problem. To support this
persons’ awareness of community health and assertion, they stated that in several studies,
support services when presented with a scenario Calsyn and colleagues (1992) had provided older
related to caring for a parent with dementia (p. adults with the name of a fictitious agency and
362) this objective was met. The significance of 20 to 30 of respondents reported familiarity
the problem was introduced clearly at the begin- with that service (p. 362).
ning of the article in the following statement: Several studies cited were more than 10 years
Although many communities have an array of older than the article itself, which may suggest
community health and support services . . . , the the paucity of research in this area. All of the
literature suggests . . . that one of the barriers to references were apparently primary sources. The
their use is the lack of awareness of such services literature review provided a logical argument for
(p. 360). This problem statement was very concise, Ploeg and colleagues’ (2009) study.
cited an appropriate reference, and presented a
persuasive argument. In addition, Ploeg and col- Development of a Conceptual
leagues specified the prevalence of Alzheimer’s Framework
disease and stated that this prevalence is pro- Ploeg and colleagues did not use a conceptual
jected to double every 20 years to more than 81 framework for the study. sing a conceptual
million people by 2040 (p. 360). This adds more framework could have provided a structure for
credence to the rationale for the research. the overarching concept of resource utilization.
454 PART FIVE Critiquing Research

dementia. Ploeg and colleagues focused

Hypotheses and Research Question on the responses to that vignette. The variables
Ploeg and colleagues (2009) explicitly stated the of this study were socioeconomic, demographic,
research questions as follows (p. 362): and social.
1. Where would older adults turn for assis-
tance when faced with a scenario related to Internal Validity
caring for a parent with dementia Although threats to internal validity are germane
2. What factors are associated with the identi- to experimental research, researchers need to pay
fication of community health and support attention to factors in a nonexperimental design
services as a source of assistance in caring that may potentially compromise a study. Possible
for a parent with dementia threats to internal validity include history which is
3. Do caregivers have greater awareness of important to acknowledge in this age of wide-
community health and support services than spread media campaigns around resources directed
non-caregivers (p. 362). at Alzheimer’s disease. Acquiescence bias was
They offered no hypotheses, but an example eliminated with the vignette methodology.
could have been that greater awareness of support
services would lead to increased utilization. External Validity
Generalizability is limited to the sample because
Sample the individuals were from a specific geographical
Ploeg and colleagues (2009) described the random area in Canada thus, generalizing to other popula-
sample, which consisted of 22,072 different tele- tions would be questionable. Ploeg and colleagues
phone numbers, and the final sample included (2009) also stated the sample over-represented
1,152 individuals (a response rate of 12.4 of eli- older adults who were female, Canadian born, and
gible particpants). The inclusion criteria were a had high levels of education (p. 368). To assess
minimum age of 50 years and ability to speak selection effects, Ploeg and colleagues compared
English. There were no exclusion criteria. Ploeg the demographics of the sample with the Hamil-
and colleagues stated that a power analysis was ton census data. The sample included higher pro-
performed and described in another paper. Restat- portions of women, people who were Canadian
ing the analysis in this article would have been born, and people with incomes higher than the the
helpful for readers to assess whether the sample average of the Hamilton population thus, they did
size was adequate and to determine the type of not represent the population.
statistical test for which the power analysis was
performed. Note that each type of statistical test— Legal/Ethical Issues
such as a t test, an analysis of variance (ANOVA), Ethical approval was received from an ethics
or regression—would have a different sample size. review board at the university this was necessary
before the study could proceed.
Research Design
Ploeg and colleagues (2009) stated that their Instruments
design was a cross-sectional descriptive survey, From the literature review, you would anticipate
which was appropriate for this study. The authors that Ploeg and colleagues would collect informa-
used vignettes to address the methodology issue tion on socioeconomics, demographics, and social
of acquiescence bias. The sample was divided into supports, which they did. They provided a detailed
three groups, and each group received vignettes, explanation of the vignette methodology with
one of which was about caring for a parent with samples of the vignettes and questions posed.
 CHAPTER 19 Critiquing Quantitative Research 455

Examples of results for the second question

Reliability and Validity showed that the older the caregivers were, the less
For the the narrative vignettes, reliability was not likely they were to turn to informal sources such
necessary. The authors stated that face and content as family, friends, and neighbours this change
validity of the vignettes was high inasmuch as the decreased significantly with age (odds ratio =
vignettes . . . were developed by front-line 0.6 Ploeg et al., 2009, p. 366). Furthermore,
service providers to represent realistic and famil- identifying their physicians as a source of
iar situations faced by older adults for which support was 40 more likely if caregivers were
community health and support services would be female (odds ratio = 1.4) and 250 more likely
appropriate (p. 362). (odds ratio = 3.5) if participants had higher levels
of education (Ploeg et al., 2009, p. 366).
The answer to the third question— Do caregiv-
Analysis and Findings of the Data ers have a greater awareness of community health
The variables mentioned in the article are on and support services than non-caregivers —is
either a nominal or an ordinal scale of measure- unclear, inasmuch as it is not mentioned.
ment. Gender, country of birth, marital status,
language, and functional limitations are measured Discussion
on a nominal scale, whereas age, education, The findings were explained within the context of
household income, self-reported health, and previous research and the CSSs with the comment
number of information sources constitute ordinal There is a need to develop and evaluate strate-
levels of measurement. gies to help physicians and other heath care pro-
Descriptive statistics were reported as means viders to improve the links between older adults
and standard deviations in both Table 19C1-1 (see and their caregivers with appropriate CSSs
p. 443) and the text a small amount of repetition (Ploeg et al., 2009, p. 368).
of the results highlighted the findings. Chi-square Also in this section, Ploeg and colleagues
analyses at the .05 level of significance were noted that they were conducting another study of
appropriately conducted for obtaining informa- how primary care physicians and allied health
tion on the association between the variables and care providers working in their offices help older
the total sample, caregivers, and noncaregivers adults access CSSs. In addition, they were writing
before logistic regression analyses were per- another paper in which they used an approach of
formed. The responses to the first question were social determinants of health.
outlined clearly in the listing of the top five places
from which older adults would seek assistance. Conclusions, Implications,
Figure 19C1-1 showed that the number one and Recommendations
resource for assistance was the physician. This Ploeg and colleagues (2009) commented brie y
finding was further noted in the text: When faced on participants’ lack of awareness of CSSs and
with a situation of parental dementia, the highest stated that action must be taken to assist physi-
percentage of participants identified a physician cians and other allied health care providers in
and the physician’s office staff as sources of linking individuals to CSSs that are focused on
support (Ploeg et al., 2009, p. 366). persons with Alzheimer’s disease.
CRITIQUING A QUANTITATIVE RESEARCH STUDY
The study “Self-Harm Intentions: Can They Be Distinguished Based Upon a History of Childhood
Physical and Sexual Abuse?” by Elaine E. Santa Mina (2010) is critiqued here. The article is
CRITIQUE 2 presented in its entirety and is followed by the critique on pp. 470–472. (From Canadian
Journal of Nursing Research, 42[4], 122–143.)

Self-Harm Intentions: Can They Be Distinguished Based Upon a

History of Childhood Physical and Sexual Abuse?

Elaine E. Santa Mina, RN, PhD might account for any differences. Conceptual-
Associate Director, Post-Diploma Degree Program izations of self-harm/suicidal intentions vary
Daphne Cockwell School of Nursing, Ryerson University
Toronto, Ontario from the wish-to-relieve disturbing thoughts and
feelings to the wish-to-die (Gratz, 2003 Leen-
A non-experimental, comparative design is used to mea-
aars, 1988). In addition, there is evidence that a
sures self-harm intention in clients with and without a
history of childhood physical and sexual abuse (CP/SA) history of childhood physical/sexual abuse (CP/
presenting to an emergency department with an episode SA) may play a key role in self-harm intentions
of self-harm behaviour. The traditional suicide literature (Pagura, Cox, Sareen, Enns, 2008). However,
identifies the key intention concepts of wish-to-die, lethality, it has not been definitively determined that a
hopelessness, and depression. However, the trauma litera- history of CP/SA distinguishes between self-
ture understands self-harm behaviour to be an adaptive
response to CP/SA and as such possibly helpful for manag-
harm intentions and suicidal behaviours.
ing intense affect and dissociation. The findings of this study This study investigates the in uence of CP/SA
demonstrate that a CP/SA history is not a distinguishing in differentiating between self-harm intentions
factor in self-harm intention. Almost all participants, regard- and suicidal behaviour in a clinical sample of
less of abuse history, gave multiple reasons for their self- adults admitted to an inner-city teaching hospital
harm behaviour, in addition to or other than the wish-to-die.
with self-harm/suicidal behaviour.
The striking similarity between the non-abused and abused
groups with regard to self-harm intention challenges clini-
cians to assess for the full range of intentions of people who
engage in self-harm and suicidal behaviour.
BACKGROUND
There are disparities in the reported rates of self-
KEYWORDS: child abuse, mental health/pyschosocial,
psychiatric nursing, stress and coping, theory harm/suicidal behaviours, suicides, and CP/SA.
These disparities result from differences in con-
Persistently high rates of self-harm/suicidal ceptual and operational definitions and corre-
behaviours, and their potential outcome of death sponding measures to report each phenomenon
by suicide, challenge clinicians to develop assess- (Santa Mina Gallop, 1998). Despite reported
ments that direct efficacious treatments and inconsistencies, the incidences remain high and
thereby reduce the incidence. Thorough clinical point to a continuing global health problem. In
assessments of self-harm/suicidal behaviours Western countries, adolescent self-harm rates
include an understanding of the motivations or range from 5 to 9 (Skegg, 2005) and in
intentions that drive them. It remains clinically college students the prevalence may be as high as
and empirically uncertain whether self-harm and 14 (Gratz, Dukes Conrad, Roemer, 2002). Of
suicidal intentions are distinguishable (Fliege, further concern is evidence that people who
Lee, Grimm, Klapp, 2009) and what factors engage in self-harm and suicidal behaviours have

456
 CHAPTER 19 Critiquing Quantitative Research 457

an increased risk of death by suicide (De Munck, intentions may serve to guide assessments of and
Portzky, Van Heeringen, 2009). Canadian and interventions for either similar or disparate clini-
American rates of death by suicide are compa- cal populations.
rable, at approximately 20 males and 5 females
per 100,000 population (Langois Morrison, LITERATURE REVIEW
2002 .S. National Center for Health Statistics,
2008). Also, CP/SA is known to be a risk factor Association Between CP/SA and Self-
for both self-harm and suicide in adults (Pagura Harm/Suicidal Behaviour
et al., 2008 Santa Mina Gallop, 1998). Inves- Three major reviews report the impact of CP/SA
tigations to better understand the possible distin- on adult suicide and self-harm (Beitchman et al.,
guishing role played by CP/SA in self-harm/ 1992 Browne Finkelhor, 1986 Santa Mina
suicidal intentions could lead to strategies for Gallop, 1998). These reviews report that adults
reducing the incidence of suicide through the use with a history of CP/SA are more likely to engage
of more precise assessments and interventions. in suicide and self-harm. Despite methodological
The theoretical and research literatures problems related to definitions, small sample
describe different suicide and self-harm etiolo- sizes, and the lack of control and comparison
gies and intentions (Walsh Rosen, 1988). The groups, there is evidence of a link between CP/
earliest empirical work describes suicide as an act SA and adult self-harm, suicidal ideation, suicidal
of self-annihilation (Freud, 1917) and suicidal behaviours, and suicide. Research also supports
intention as a phenomenon that uctuates along an association between self-harm and suicide.
a life–death continuum. The research literature Many people who engage in self-harm behaviour
supports associations between the wish-to-die, eventually attempt suicide or die by suicide
multiple etiologic risk factors, and the severity (Mann Currier, 2007).
of depression and hopelessness (Fuse, 1997).
However, contemporary trauma theorists suggest Childhood Physical and Sexual Abuse
that self-harm is a phenomenon separate from but Conceptual and operational definitions of CP/SA
related to suicide and that self-harm intentions differ (Santa Mina Gallop, 1998). The most
may counter the death wish and be a method for common definitions are incorporated into this
coping (Alexander, 1999 Arnold, 1995). Theo- study and are supported by the literature. Physical
rists suggest that many adults with a history of abuse is deliberate striking, hitting, punching, or
childhood CP/SA cope with the overwhelming burning a child less than 18 resulting in physical
feelings (affect dysregulation) and disturbances injury such as: bruises or fractures requiring
in memories and thoughts (dissociation) by medical intervention, distinguished from single
engaging in self-harm. The literature does support slaps, or interfamilial fights (Windle, Windle,
an association between CP/SA, the experience of Scheidt, Miller, 1995, p. 1323). Sexual abuse
intense emotional and cognitive sequelae, and is sexual contact, ranging from fondling to inter-
adult self-harm (Browne Finkelhor, 1986 course, between a child in mid-adolescence or
Santa Mina Gallop, 1998). Although an asso- younger and a person at least five years older
ciation between CP/SA and adult psychiatric dis- (Briere, 1992, p. 4).
orders has been widely reported (Brown
Anderson, 1991 Welch, Patterson, Shaw, Conceptualization of Suicide and
Stewart-Brown, 2009), its in uence on self-harm Self-Harm Intention
and suicidal intentions is not clear. Knowledge The suicide and self-harm literature and concep-
about the in uence of CP/SA in self-harm tual reviews do not reach consensus on the extent,
458 PART FIVE Critiquing Research

if any, of overlap of intentions between the two traditional suicide theories and empirical studies
phenomena, or whether they are in fact the same as well as recent trauma theories and investiga-
phenomenon (Fliege et al., 2009 Walsh Rosen, tions direct this study’s conceptualization of the
1988 Whitlock Knox, 2007). As this study phenomenon and the choice of instruments to
proposes that the intentions may differ based measure the variables. The conceptualization for
upon a history of CP/SA, the breadth of the study is that the sociological factor of gender
intentions—as identified across the literature—is and the cognitive and affective states of wish-to-
included in the study’s conceptualization. For this die, hopelessness, depression, and affective regu-
study, self-harm intention is defined, with the key lation dissociation, as described below, in uence
constructs from the self-harm and suicide litera- self-harm/suicidal intentions in the presence of a
ture and adapted from Connors’ (1996) work, as history of CP/SA.
the purpose(s) or meaning(s) of the self-harm
behaviour, the regulation of affect, the regulation Sociological Risk Factors
of dissociation, and/or the termination of life Sociological suicide theories articulate broad
associated with direct self-actions that lie outside social factors such as geography, culture, socio-
the realm of social acceptability and that hurt or economic status, marital status, age-related
harm the body (Santa Mina, 2005, p. 67). Self- factors, and gender that support a macro level of
harm behaviour is the class of actions, outside analysis to explain suicide (Fuse, 1997). For
the realm of social responsibility, that hurt or example, suicide attempts are more frequent in
harm the body including cutting, burning, slap- females (De Munck et al., 2009) yet males are at
ping, punching, scratching, gouging, harmful least three times more likely than females to die
enemas and douches, interfering with the healing by suicide (Statistics Canada, 2010). Also,
of wounds, inserting dangerous objects into females are reported to engage in more self-
the vagina or rectum, head-banging, . . . choking, cutting behaviours than males and are motivated
hitting oneself with objects, ingesting sharp by recurring issues of previous trauma (Arnold,
objects, and biting (Connors, 1996, p. 199). 1995). Therefore, this study investigates CP/SA
The conceptualization of self-harm/suicidal and gender as possible factors in uencing self-
intention and its measurement are informed by harm/suicidal intention.
the constructs that are known to be risk factors
for the behaviours/sociological risk factors the Suicide Intention
key suicide-intention constructs the wish-to-die The wish-to-die, which is a quintessential suicide
lethality, depression, and hopelessness (Leenaars, concept, emanates from traditional psychological
1988 Lester, 1991) and the management of emo- suicide theories (Freud, 1917 Menninger, 1935)
tions (affect) and cognitive disturbance (dissocia- and continues to be fundamental today in the
tion) (Gratz, 2003). Previous investigations have assessment of suicide risk (Registered Nurses
minimally integrated these constructs from soci- Association of Ontario, 2007). Suicide intention
ology and psychology to re ect the complexity of uctuates along a wish-to-live/wish-to-die con-
the phenomenon. However, the prevalence of tinuum the extent to which someone wants to end
trauma history in people who attempt or die by his or her life is an indication of intention to die
suicide and people who self-harm suggests that (suicide intention). The wish-to-die lays the
intentions may be broader than the intent-to-die groundwork for investigation of the lethality of
that is articulated in the classic suicide literature the suicidal behaviour. Highly lethal suicidal
(Fliege et al., 2006). Therefore, concepts from behaviour is associated with an increased
 CHAPTER 19 Critiquing Quantitative Research 459

likelihood of dying by suicide and is interpreted

as proximal to the wish-to-die (Linehan, Comtois, Affect Regulation
Brown, Heard, Wagner, 2006). The conceptualization of self-harm without the
intent-to-die has evolved within the trauma litera-
Depression ture. Prior to the 1980s, few authors (Jung, 1974
Depression is a medical diagnosis that includes Rosenthal, Rinzler, Walsh, Klausner, 1972)
the presence, in a 2-week period, of one of two alluded to trauma as a precursor of self-harm.
symptoms—depressed mood and decreased More recently, authors (Klonsky Moyer, 2008
interest in daily activities almost all day, most van der Kolk et al., 1996) have described an asso-
days—and four or more of the following symp- ciation between the constellation of emotional,
toms: an increase or decrease in weight in the behavioural, and cognitive disturbances and a
absence of dieting a decrease in appetite an history of childhood trauma in people who engage
increase or decrease in sleeping restlessness or a in self-harm behaviour. The emotions of rage,
decrease in physical activity fatigue or a decrease frustration, guilt, and shame are common in
in energy feelings of worthlessness decreased trauma survivors and are experienced intensely
ability to concentrate or make decisions and and rapidly (Herman, 1992). Emotional instabil-
recurring thoughts of death or suicide (American ity, or sudden uctuations in overwhelming feel-
Psychiatric Association, 1994). Depression is ings, is known as affective lability. It is theorized
highly associated with suicidal intention and sui- that self-harm is a response to these intense nega-
cidal behaviour (World Health Organization, tive feelings as it expresses them and thereby
2010) and childhood trauma, such as CP/SA restores emotional stability (Alexander, 1999
(Herman, 1992). Arnold, 1995).

Hopelessness Dissociation
Hopelessness is characterized by negative, con- A cognitive disturbance known as dissociation is
stricted thought content and patterns regarding experienced frequently by survivors of CP/SA.
the past, present, and future such that pessimism Dissociation is a temporary state of cognitive dis-
and meaninglessness permeate the experience of integration such that a person’s consciousness,
living (Beck, Weissman, Lester, Trexler, 1974 memory, identity and perception of environment
Mitchell, Garand, Dean, Panzak, Taylor, 2005). are momentarily disrupted along a continuum of
People who feel hopeless rigidly believe that their severity, from inattention to inability to integrate
perceptions are accurate and exclude other expla- affect, behaviour, and cognition (Mulder, Beau-
nations that lead to realistic interpretations trais, Joyce, Fergusson, 1998, p. 806). Separa-
(Lester, 1994). Suicide becomes the only way to tion of conscious awareness from traumatic
escape emotional suffering. Beck, Morris, and events enables a person to withdraw from the
Beck (1974) demonstrate the ability of the cogni- psychological, emotional, and cognitive pain. In
tive state of hopelessness to in uence the nature the dissociated state, the child observes him/
of emotions and the subsequent behaviour. In herself as though the abuse were happening to
their seminal work, Beck, Schuyler, and Herman someone else. Later in life, the person may
(1974) report that depression combined with re-experience dissociative states in the presence
hopelessness is a key factor in suicide intention of emotional distress and have temporary percep-
and may result in highly lethal suicidal behaviour tual experiences that are out of touch with the real
and death by suicide. world. The behavioural response to a state of
460 PART FIVE Critiquing Research

dissociation may be self-harm, which can help the expressed interest, then the research assistant,
person to regain a sense of what is real (Mangall who was not associated with either unit, met with
Yurkovich, 2008 van der Kolk et al., 1996). the client at his or her convenience in a private
office on the unit. Study details were described in
depth, with opportunities for the participant to ask
RESEARCH HYPOTHESES questions and withdraw at any time. Written
It was hypothesized that, compared to those informed consent was obtained. The research
without CP/SA, clients who engaged in a recent assistant was a graduate student in psychology
episode of self-harm behaviour and reported a with experience counselling people with psychi-
CP/SA history would have (1) less suicide inten- atric disorders and a history of trauma. Clinical
tion, (2) more reasons related to affect regulation staff were available should the participant become
and dissociation, (3) greater dissociation, (4) distressed. A contact number was provided to par-
greater affective lability, (5) less hopelessness, ticipants for crisis support should they become
and (6) no difference in depression. distressed at any time after completion of the
study.
The participants completed all self-report
METHOD instruments in hospital within 3 days of the
A non-experimental, comparative design was episode of self-harm behaviour. They received
used. A power analysis for the study was con- 10 plus public transit or parking expenses, if
ducted in advance of data collection to determine incurred. No participant reported distress requir-
sample-size adequacy. It was decided that a ing clinical or crisis support.
sample size of 64 per non-abused group and The instruments for measuring CP/SA were
abused group was required in order to detect a the Childhood Physical Abuse Scale (Briere,
moderate effect. Ethics approval was obtained 1992) and Russell’s Sexual Abuse Scale (Russell,
from the research ethics boards of the university 1999). The instruments for each hypothesis were
and the hospital. (1) the Beck Suicide Intent Scale (SIS) (Beck,
A convenience sample of clients who engaged Morris, et al., 1974) and the Self-In icted Injury
in self-harm behaviour was recruited from the Severity Form (SIISF) (Potter et al., 1998) (2)
inpatient and emergency units of the mental the Reasons for Self-Injury Inventory (SI )
health service in an inner-city teaching hospital (Alexander, 1999) (3) the Dissociative Experi-
the clients were recruited while still in hospital. ences Scale (DES) (Bernstein Carlson Putnam,
Recruitment took place weekdays only, due to 1993) and the Structured Interview for Disorders
constraints in providing around-the-clock research of Extreme Stress (SIDES) dissociative subscale
assistant support. The principal investigator (Pelcovitz et al., 1997) (4) the Trauma Symptom
informed all clinical staff about the study and the Checklist-40 (Briere, 1996) and the SIDES affec-
clinical staff identified clients who met the inclu- tive subscale (Pelcovitz et al.) (5) the Beck
sion criteria: 18 years of age or older English- Hopelessness Scale (BHS) (Beck, Weissman,
speaking presenting to the inpatient or emergency et al., 1974) and (6) the Beck Depression Inven-
unit with an episode of self-harm and medically tory II (BDI II) (Beck, Steer, Garbin, 1988).
deemed as a voluntary admission to the hospital,
competent to give consent under the Mental
ealth Act of ntario, Canada, Potential DATA ANALYSIS
participants were asked if they were interested in Descriptive statistics were calculated to charac-
learning more about the study. If the client terize the sample on the variables of interest.
 CHAPTER 19 Critiquing Quantitative Research 461

Inferential analyses were used to address the most frequent type (χ2 = 11.122 df = 5 p =
study purpose. Two-tailed t tests and chi-square 0.049). Self-harm types were evenly distributed
tests for independence (5 level of significance) between genders. The predominance of overdose
compared the non-abused and abused groups on as a self-harm method is consistent with method
suicide intention, lethality, reasons for self-harm, distributions found in other clinical studies (Haw,
dissociation, affect regulation, hopelessness, and Houston, Townsend, Hawton, 2002 Lester
depression. Of the 64 participants with abuse, 31 Beck, 1980 Schnyder Valach, 1997). The
had CP/SA, 17 had physical abuse only, and 16 majority of the abused participants in the sample
had sexual abuse only. Therefore, an ANOVA for reported isolated incidents of CP/SA. They rarely
four groups—non-abused, physically abused only, reported incidents of abuse at a very young age
sexually abused only, and CP/SA—constituted (pre-pubescence) or of a prolonged or severe
the secondary analysis. nature (over several years with force and penetra-
Three analyses of the SI items were con- tion). All participants were moderately suicidal
ducted. First, the total number of reasons for self- yet re ected an overall low level of lethality
harm per participant was calculated from the sum (SIISF 48/77 had no actual injury). The partici-
of the items selected. Next, the reason items pants had received immediate intervention after
were categorized into two subscales based on the self-harm episode, which may have served to
trauma theory, in order to test the hypotheses. lower the level of lethality.
Finally, a principal component analysis was con- The sample was highly dissociative, with high
ducted to extract five SI reason factors concep- affective lability. All participants reported a life-
tually grounded in trauma theory. The findings time presence of affective instability and 70
from this work are reported elsewhere (Santa reported a lifetime presence of dissociation,
Mina et al., 2006). with no significant between-group differences.
On the SIDES current presence items, 40
(n = 80) reported current presence of affective
RESULTS instability and 34 reported current presence
The research assistant approached 113 clients for of dissociation, also with no between-group dif-
possible participation. Of these, 83 agreed to take ferences. Overall, the sample reported moderate
part, for a response rate of 73 (non-abused, n = hopelessness and severe depression.
19 abused, n = 64). The reason given for refusal Self-harm/suicidal intentions did not differ for
was lack of interest and/or time. the non-abused and abused groups (Table 19C2-1),
Descriptive statistics indicated a demographi- nor did the level of lethality of the behaviour.
cally homogeneous sample. Although gender was However, the sample overwhelmingly (98 )
distributed evenly, females were more likely than reported multiple reasons for the self-harm
males to be abused rather than non-abused episode. Only 2 of the sample (n = 2) reported
(females, 7 : 1 males, 2 : 1 χ2 = 4.723 df = 1 suicide as the sole reason for the behaviour, while
p = 0.03). The participants ranged in age from 18 60 (n = 50) reported suicide plus other reasons,
to 69 years, with the majority (82 ) being such as to achieve a feeling of peace or to
between 21 and 50 years (x = 37.28 SD = 11.05). regain a sense of reality, and 38 (n = 31)
Due to the small sample size per group, analysis reported multiple reasons that did not include
of additional sociological factors did not produce suicide. The abused group revealed greater dis-
meaningful results. Types of self-harm behaviour sociation than the non-abused group, as well as
varied somewhat (overdose, n = 42 cutting, n = greater affective lability. The secondary ANOVA
20 other, n = 21), although overdose was the found no significant difference across the four
 462

TABLE 19C2-1
INSTRUMENTS, SCORING METHODS, AND RELIABILITY AND VALIDITY
PREVIOUSLY
REPORTED
INTERPRETATION OF RELIABILITY AND THIS STUDY’S RELIABILITY
CONCEPT INSTRUMENT NUMBER OF ITEMS SCORING METHOD SCORES VALIDITY AND VALIDITY

Childhood Childhood 3 items Yes/no and frequency Yes to any of the No reports Not tested
Physical Abuse Physical Abuse if yes items = history
Scalea of CPA
Childhood Sexual Russell’s Sexual 8 items Yes/no and frequency Yes to any of the Test-retest Not tested
Abuse Abuse Scaleb of events by items = history reliability
PART FIVE Critiquing Research

perpetrator of CSA r = .91–.99

Suicide Intent Beck’s Suicide 10 items Each item scored as Higher score Interrater α = .76; 95% confidence
Intent Scale 0, 1, or 2; possible indicative of reliability interval; α = .67–.83
(SIS)c range of scores greater intent; SIS r = .82–.95d;
0–30 scores in suicidal α = .85e
clients x =
7.68–10.2
Lethality Self-Inflicted Injury 4 categories of Based on chart 1 (no injury) to 4 High internal Interrater reliability on
Severity Form lethality severity review, self-harm (highly lethal consistency; method of self-harm r =
(SIISF)f for each of behaviour is trauma high interrater .94; interrater reliability
7 self-harm categorized by penetrating body reliability; high SIISF r = .77
behaviour self-harm method cavity) discriminate
methods and severity of and construct
lethality validityf
Reasons for Self-Injury 31 items total: Yes/no per item; total Yes = reason given Good face Strong α = 0.83; 95%
Self-Harm Questionnaire 2 subscales: sum of number of by participant for validityg confidence interval;
(SIQ)g (i) reasons due to reasons; sum of self-harm event; 0.78–0.86; affective
affect, 7 items; reasons for no = not a subscale: α = 0.72; 95%
(ii) reasons due to subscales; factor reason for confidence interval,
dissociation, analysis for reason self-harm given 0.62–0.80; dissociative
4 items themesh; range of by participant for subscale: α = 0.77; 95%
scores: total scale self-harm event confidence interval,
total number of 0.68–0.84; factor
reasons 0–31; analysis: factors I–V:
subscale reasons weak significant
due to affect 0–7; correlations with SIS,
subscale reasons BDI II, BHS, SIISF, TSC-40,
due to dissociation DESh
0–4
 PREVIOUSLY
REPORTED
INTERPRETATION OF RELIABILITY AND THIS STUDY’S RELIABILITY
CONCEPT INSTRUMENT NUMBER OF ITEMS SCORING METHOD SCORES VALIDITY AND VALIDITY
Affect Regulation Structured (i) symptom current (i) yes/no; Yes = present; no = α = .90i α = .83; 95% confidence
Interview for presence; (ii) yes/no; not present; interval = .76–.87
Disorders of (ii) symptom (iii) 0–3 per item 0 (none) to 3
Extreme Stress: lifetime presence; mean score on (frequently)
affect (ii) symptom subscale score: ≥ 2 is
dysregulation severity 19 items clinically
subscale, significant
19 itemsi
Affect Regulation Trauma Symptom 40 items 1 (never) to 4 (often) Higher score α = .89–.91j α = .90; 95% confidence
Checklist-40 Range of scores indicates greater interval = .86–.93
(TSC-40) 0–120 affective
dysregulation
Dissociation Structured (i) symptom current (i) yes/no; (i) and (ii) yes = Alteration α = .62; 95% confidence
Interview for presence; (ii) yes/no; present, no = not inattention = interval = .47–.73
Disorders of (ii) symptom (iii) 1–3 per item present; .76i
Extreme Stress lifetime presence; mean score on (iii) 0 (none) to 3
(SIDES): (iii) symptom subscale (frequently);
alteration in severity 5 items range of scores
attention 0–3; score ≥ 2 is
subscale clinically
5 itemsi significant
Dissociation Dissociative 28 items 0–100; Mean scores range Cut-off scores < 10 Test-retest α = .93; 95% confidence
Experiences 0 (never) to 100 from 0 to 100 = low reliability r = interval = .88–.95
Scale (DES)k (always); mean dissociation; .79–.96 at
scores 10–29.9 = intervals of 4
dissociation to 8 weeks;
common with internal
psychiatric reliability α =
disorders, not .83–.93 (p <
trauma; >30 = .00)k
dissociative
disordersk
Continued
CHAPTER 19 Critiquing Quantitative Research
463
 TABLE 19C2-1
464

INSTRUMENTS, SCORING METHODS, AND RELIABILITY AND VALIDITY—cont’d

PREVIOUSLY
REPORTED
INTERPRETATION OF RELIABILITY AND THIS STUDY’S RELIABILITY
CONCEPT INSTRUMENT NUMBER OF ITEMS SCORING METHOD SCORES VALIDITY AND VALIDITY
Hopelessness Beck Hopelessness 20-item true/false Sum total score 0–3 = no α = .92m α = .93; 95% confidence
Scale (BHS)l hopelessness; interval = .82–.95
4–8 = mild;
9–14 =
moderate; 15–20
= severe; 9 =
criteria for
PART FIVE Critiquing Research

sensitivity
(attempters
correctly
identified as
eventually dying
by suicide) and
specificity
(attempters
identified and
did not die by
suicide)l
Depression Beck Depression 21 items Scores range from 0 0–9 = no to α = .81 –86n α = .87; 95% confidence
Scale (BDI II)n to 63 minimal interval = .82–.91
depression;
10–18 = mild to
moderate; 19–29
= moderate to
severe; 30–63 =
severen
Note: α = coefficient alpha.
a
Briere (1992).
b
Russell (1999).
c
Beck, Morris, & Beck (1974).
d
Beck, Morris, et al. (1974).
e
Beck & Steer (1989).
f
Potter et al. (1998).
g
Alexander (1999).
h
Santa Mina et al. (2006).
i
Pelcovitz et al. (1997).
j
Elliot & Briere (1992).
k
Bernstein Carlson & Putnam (1993).
l
Beck, Weissman, Lester, & Trexler (1974).
m
Beck & Steer (1989).
n
Beck, Steer, & Garbin (1988).
 CHAPTER 19 Critiquing Quantitative Research 465

groups for any of the hypotheses, with one excep- between life and death (Shneidman, 1985). Yet
tion: On the SI affect sub-scale, the physical that conjecture misses the breadth of other power-
and sexual abuse group reported the largest mean ful intentions by narrowly focusing on the wish-
number of reasons, as compared to the other three to-live/wish-to-die continuum that subsequently
groups ( = 2.88, 3, 79 p = .04). On the factor may shape assessments and interventions.
analysis, the SI mean scores for factors I through Recruitment of an adequate sample size for a
V did not differ significantly across the groups. non-abused group proved to be very challenging
The ANOVA, in the secondary analyses, tested in this inner-city population. When adequacy of
for a difference in dissociation (DES) and depres- sample size for the abused group was achieved
sion (BDI II) across the four groups. Although (n = 64), the non-abused sample remained at 19.
there was no statistically significant difference in As data analysis demonstrated highly non-
dissociation or depression across the four groups, significant findings and risk of a type II error was
a trend of higher scores in the group with both low, the decision was made to stop data collec-
physical and sexual abuse was noted and exam- tion. However, this limitation is important in
ined. A linear trend was calculated on the DES placing the findings in context. Participants with
and BDI II and demonstrated a statistically sig- a broad history of childhood trauma, inclusive of
nificant increase in dissociation ( = 5.211 df = emotional and psychological abuse and physical
1 p = 0.025) and depression ( = 7.796 df = 1 neglect, may endorse responses to suicide inten-
p = 0.01), with an increase in abuse severity. tion, reasons for self-harm, affect regulation, dis-
sociation, depression, and hopelessness that are
similar to responses of those with a history of CP/
DISCUSSION SA. It is possible that some of the participants in
The results of this study refute the hypothesis that the non-abused group had experienced these
self-harm intentions can be distinguished based other types of childhood abuse, but it was beyond
on a history of CP/SA. Although the findings are the scope of this study to measure them. This
not statistically significant, they are clinically sig- could account for the absence of differences
nificant. The majority of participants, regardless between the groups, as the non-abused sample
of abuse history, reported multiple intentions, may have had other forms of childhood trauma
re ective of the key concepts in both suicide and and may have been as responsive in intentionality
self-harm intention. This finding provides direc- as those with CP/SA. Challenges in finding ade-
tion for practice and research. The participants quate sample sizes for the non-abused clinical
endorsed numerous, seemingly contradictory, group remain, as abuse prevalence is high in
intentions for self-destructive behaviours: to psychiatric populations. It is possible that these
bring myself back to reality, to achieve a feeling findings and a propensity towards an abused
of peace, to distract from feelings or thoughts population are representative of an inner-city
of suicide and a suicide attempt. Participants population. Comparisons between inner-city/
also reported moderate levels of suicidality (SIS) urban and rural populations, as well as between
and high levels of hopelessness (BHS) and emergency and community populations, may also
depression (BDI II), even if they did not report be helpful for obtaining comparator groups.
suicide attempt as a reason on the SI . This Future studies should aim for large sample sizes,
mixed clinical picture is consistent with the find- to accommodate the breadth of abuse types as
ings of other studies (Brown, Comtois, Linehan, well as the myriad sociological factors that can
2002 Holden McLeod, 2000). One might affect intentions. Alternative sampling strategies,
argue that it supports the notion of ambivalence such as highly resourced consecutive sampling,
466 PART FIVE Critiquing Research

would serve to reduce sampling bias and ensure may not be gender-based or associated with spe-
the inclusion of all patients with self-harm cific methods or diagnoses such as BPD.
who are treated and released from emergency The associations amongst the spectrum of
departments during evening/night shifts and childhood maltreatment and adult self-harm and
weekends. suicide are multifaceted and complex. The natures
Evidence also points to a complex relationship of both childhood maltreatment and self-harm/
between numerous types of childhood maltreat- suicide are highly sensitive and are fraught with
ment and adult psychopathology that are not issues of stigma and intense emotion. Although
limited to overt self-harm and suicidal behaviours quantitative, psychometrically sound instruments
(Welch et al., 2009). The relationship between for measuring the full range of types of childhood
childhood maltreatment and adult addictions, as abuse and types of self-harm and intentions may
one specific type of indirect self-harm behaviour, be useful, the nature of the phenomena may be
is an example. Indirect types of self-harm, such such that numeric measures do not fully capture
as risky behaviour inclusive of substance misuse the essence of the experience. ualitative methods
as discussed by Connors (1996), should be added may serve to enrich our knowledge of these mul-
as factors in our understanding of the complexity tifaceted problems. Both positivist and naturalist
and breadth of childhood maltreatment and adult paradigms need to be incorporated into future
self-harm and suicide. research, to fully inform us about how gender and
In the present study, the prevalence of CP/SA the spectrum of abuse types, methods, lethality,
in both men and women is noteworthy. The find- and intentions contribute to self-harm intention.
ings point to a similarity in self-harm intentions It may be that a triangulation of methods will lead
and methods regardless of gender, a notable soci- to a better understanding of this intricate issue.
ological risk factor for self-harm and suicide. The findings from this study indicate that
Much of the self-harm literature addresses females patients who are thinking about or engaging in
who cut themselves in order to cope with over- self-harm/suicidal behaviours need to be assessed
whelming thoughts and feelings in response to for the full spectrum of intentions, regardless of
their trauma histories. The reports are often gender, self-harm method, or known abuse
focused on female patients who cut themselves history. Research is needed to direct the develop-
and are diagnosed with borderline personality ment and utilization of clinical self-harm/suicide
disorder (BPD) (Andover, Pepper, Ryabchenko, assessments that are more inclusive of the breadth
Orrico, Gibb, 2005). In the present study, inter- of self-harm intentions. Fully informed assess-
estingly, men as well as women engaged in ments may guide clinicians to develop interven-
cutting to relieve intense affect and manage dis- tions targeted to patient needs. Efficacious,
sociation, and men and women overdosed to intention-based interventions have the potential
achieve the same purposes, regardless of abuse to diminish the tragic impact of self-destructive
histories. This suggests that males as well as behaviours on individuals, their families, and
females engage in various types of self-harm societies by reducing self-harm and suicidal
behaviours for reasons related to affect and dis- behaviours.
sociation. As gender did not present as a socio-
logical factor in uencing either the type of CONCLUSION
self-harm behaviour or behavioural intentions, This is the first published clinical study to inves-
clinicians and researchers may need to rethink tigate the in uence of CP/SA on self-harm/suicide
intentions in self-harm behaviours, as these may intentions among men and women admitted to an
not be exclusively the sequelae of CP/SA and acute-care unit for these behaviours. Although the
 CHAPTER 19 Critiquing Quantitative Research 467

literature points to CP/SA as a possible distin- Beck, A. T., Schuyler, D., Herman I. (1974).
guishing factor for intentions, the present findings Development of suicidal intent scales. In A. T. Beck,
H. L. P. Resnick, D. J. Lettieri (Eds.), The
demonstrate that self-harm and suicidal inten-
prediction of suicide (pp. 45-58). Bowdie, MD:
tions may be remarkably similar, regardless of a Charles Press.
history of CP/SA. In the majority of individuals Beck, A. T., Steer, R. A. (1989). Clinical predictors
across both abused and non-abused groups, of eventual suicide: A 5-10 year prospective study of
regardless of gender or self-harm type and self- suicide attempters. ournal of Affective Disorders, ,
harm/suicidal intentions, intentions were over- 203-209.
Beck, A. T., Steer, R. A., Garbin, M. G. (1988).
whelmingly varied, were seemingly contradictory, Psychometric properties of the Beck Depression
and re ected a need to manage distressing feel- Inventory: Twenty-five years of evaluation. Clinical
ings and thoughts within the context of the wish- Psychology Revie , , 77-100.
to-die. These findings challenge clinicians and Beck, A. T., Weissman, A., Lester, D., Trexler, L.
researchers to re-evaluate the breadth of their (1974). The measurement of pessimism: The
Hopelessness Scale. ournal of Consulting and
assessments and interventions and to incorporate
Clinical Psychology, (6), 861-865.
knowledge from both suicide and trauma theories Beck, R. W., Morris, J. B., Beck A. T. (1974).
in order to best care for people with this multidi- Cross-validation of the Suicidal Intent Scale.
mensional problem. Psychological Reports, , 445-446.
Beitchman, J. H., ucker, K. J., Hood, J. E., Da Costa,
G. A., Akman, D., Cassavia, E. (1992). A review
ACKNOWLEDGEMENTS of the long-term effects of child sexual abuse. Child
Abuse and Neglect, ,
This study was funded by a St. Michael’s Hospi- 101-118.
tal Community Mental Health Fellowship, Bernstein Carlson, B. E., Putnam, F. W. (1993). An
St. Michael’s Mental Health Service, in Conjunc- update on the Dissociative Experiences Scale.
tion with the Mental Health Systems Research Dissociation, (1), 16-29.
Briere, J. (1992). Child abuse trauma: Theory and
and Development Program, Department of Psy-
treatment of the lasting effects. Newbury Park, CA:
chiatry, niversity of Toronto. Sage.
I gratefully acknowledge the contributions of Briere, J. (1996). Psychometric review of the Trauma
my committee members for this study: Dr. Ruth Symptom Checklist-40. In B. H. Stamm (Ed.),
Gallop, Dr. Paul Links, Dr. Dorothy Pringle, and Measurement of stress, trauma, and adaptation
Dr. Ron Heslegrave. (pp. 373-376). Lutherville, MD: Sidran Press.
Retrieved December 1, 2001, from
www.johnbriere.com/tsc.htm.
REFERENCES Brown, G. R., Anderson, B. (1991). Psychiatric
Alexander, L. A. (1999). The functions of self-injury morbidity in adult inpatients with childhood histories
and its lin to traumatic events in college students of sexual and physical abuse. American ournal of
npublished doctoral dissertation, niversity of Psychiatry, (1), 55-61.
Massachusetts, Amherst. Brown, M. ., Comtois, K. A., Linehan, M. M.
American Psychiatric Association. (1994). Diagnostic (2002). Reasons for suicide attempts and non-suicidal
and statistical manual of mental disorders (4th ed.). self injury in women with borderline personality
Washington: Author. disorder. ournal of Abnormal Psychology, (1),
Andover, M., Pepper, C., Ryabchenko, K., Orrico, E., 198-202.
Gibb, B. (2005). Self-mutilation and symptoms of Browne, A., Finkelhor, D. (1986). Impact of child
depression, anxiety, and borderline personality. Suicide sexual abuse: A review of the research. Psychological
and ife Threatening Behavior, (5), 581-591. Bulletin, (1), 66-77.
Arnold, L. (1995). omen and self injury: A survey of Connors, R. (1996). Self-injury in trauma survivors: 1.
omen. Bristol: Mental Health Foundation, Bristol Functions and meanings. American rtho-psychiatric
Crisis Service for Women. Association, (2), 197-206.
468 PART FIVE Critiquing Research

De Munck, S., Portzky, G., Van Heeringen, K. Lester, D. (1991). The study of suicidal lives. Suicide
(2009). Epidemiological trends in attempted suicide and ife Threatening Behavior, (2), 164-173.
in adolescents and young adults between 1996 and Lester, D. (1994). Research note: A comparison of 15
2004. Crisis, (3), 115-119. theories of suicide. Suicide and ife Threatening
Elliot, D. M., Briere, J. (1992). Sexual abuse trauma Behavior, (1), 80-88.
among professional women: Validating the Trauma Lester, D., Beck, A. T. (1980). What the suicide’s
Symptom Checklist-40 (TSC-40). Child Abuse and choice of method signifies. ME A, (3), 271-277.
Neglect, , 391-398. Linehan, M. M., Comtois, K. A., Brown, M. ., Heard,
Fliege, H., Kocalevent, R., Walter, M., Beck, S., Gratz, H. L., Wagner, A. (2006). Suicide Attempt
K., Gutierrez, P. M., et al. (2006). Three assessment Self-Injury Interview (SASII): Development,
tools for deliberate self-harm and suicide behavior: reliability, and validity of a scale to assess suicide
Evaluation and psychopathological correlates. ournal attempts and intentional self-injury. Psychological
of Psychosomatic Research, , 113-121. Assessment, (3), 303-312.
Fliege, H., Lee, J.-R., Grimm, A., Klapp, B. (2009). Mangall, J., Yurkovich, E. (2008). A literature review
Risk factors and correlates of deliberate self-harm of deliberate self-harm. Perspectives in Psychiatric
behaviour: A systematic review. ournal of Psychoso- Care, (3), 175-184.
matic Research, , 477-493. Mann, J. J., Currier, D. (2007). Prevention of suicide.
Freud, S. (1917). Mourning and melancholia. In Psychiatric Annals, (5), 331-339.
J. Strachey (Ed.), Complete psychological works, Menninger, K. A. (1935). A psychoanalytic study of the
Vol . London: Hogarth. significance of self-mutilations. Psychoanalytic
Fuse, T. (1997). Theories of suicide. In T. Fuse (Ed.), uarterly, , 408-466.
Suicide: Individual and society (pp. 75-163). Toronto: Mitchell, A., Garand, L., Dean, D., Panzak, G.,
Canadian Scholars’ Press. Taylor, M. (2005). Suicide assessments in hospital
Gratz, K. L. (2003). Risk factors for and functions of emergency departments: Implications for patient
deliberate self-harm: An empirical and conceptual satisfaction and compliance. Topics in Emergency
review. Clinical Psychology, , 192-205. Medicine, (4), 302-312.
Gratz, K. L., Dukes Conrad, S., Roemer, L. (2002). Mulder, R. T., Beautrais, A. L., Joyce, P. R.,
Risk factors for deliberate self-harm among college Fergusson, D. M. (1998). Relationship between
students. American ournal of rthopsychiatry, (1), dissociation, childhood sexual abuse, childhood
128-140. physical abuse, and mental illness in a general
Haw, C., Houston, K., Townsend, E., Hawton, K. population. American ournal of Psychiatry, (6),
(2002). Deliberate self-harm patients with depressive 806-811.
disorders: Treatment and outcomes. ournal of Pagura, J., Cox, B., Sareen, J., Enns, M. (2008).
Affective Disorders, (1), 57-65. Factors associated with multiple versus single episode
Herman, J. L. (1992). Trauma and recovery. New York: suicide attempts in 1990–1992 and 2001–2003 nited
Basic Books. States National Comorbidity Surveys. ournal of
Holden, R., McLeod, L. (2000). The structure of the Nervous and Mental Disease, (11), 806-813.
Reasons for Attempting Suicide uestionnaire Pelcovitz, D., van der Kolk, B., Roth, S., Mandel, F.,
(RAS ) in a non-clinical adult population. Personal- Kaplan, S., Resick, P. (1997). Development of a
ity and Individual Differences, , 621-628. criteria set and a Structured Interview for Disorders
Jung, C. G. (1974). The relations between the ego and of Extreme Stress (SIDES). ournal of Traumatic
the unconscious (R. Hull, Trans.). In H. Read, M. Stress, (1), 3-12.
Fordam, G. Adler (Eds.), The collected works of Potter, L. B., Kresnow, M., Powell, K. E., O’Carroll,
C. G. Jung, Vol . London: Routledge Kegan Paul. P. W., Lee, R. K., Frankowski, R., et al. (1998).
(Original published in 1928.) Identification of nearly fatal suicide attempts:
Klonsky, E. D., Moyer, A. (2008). Childhood sexual Self-In icted Injury Severity Form. Suicide and ife
abuse and non-suicidal self-injury: Meta-analysis. Threatening Behavior, (2), 174-186.
British ournal of Psychiatry, , 166-170. Registered Nurses Association of Ontario. (2007). The
Langois, S., Morrison, P. (2002). Suicide deaths and assessment and care of adults at ris for suicidal
suicide attempts. ealth Records, (2), 9-22. ideation and behaviour: Best practice guideline.
Leenaars, A. A. (1988). Suicide notes: Predictive clues Toronto: Author. Retrieved January 8, 2010, from
and patterns. New York: Human Sciences Press. http://www.rnao.org/.
 CHAPTER 19 Critiquing Quantitative Research 469

Rosenthal, R. J., Rinzler, C., Walsh, R., Klausner, E. .S. National Center for Health Statistics. (2008).
(1972). Wrist-cutting syndrome: The meaning of a Health, nited States. Retrieved August 31, 2010,
gesture. American ournal of Psychiatry, (11), from http://www.cdc.gov/nchs/hus.htm.
1363-1363. van der Kolk, B. A., Pelcovitz, D., Roth, S., Mandel,
Russell, D. E. H. (1999). The secret trauma: Incest in F. S., McFarlane, A., Herman, J. L. (1996).
the lives of girls and omen. New York: Basic Dissociation, somatization, and affect dysregulation:
Books. The complexity of adaptation to trauma. American
Santa Mina, E. E. (2005). Intentions in self-harm ournal of Psychiatry, (7), 83-93.
behaviour in an emergency population: Can they be Walsh, B. W., Rosen, P. M (1988). Distinguishing
distinguished based upon a history of childhood self-mutilation from suicide: A review and commen-
physical and se ual abuse npublished doctoral tary. In B. W. Walsh P. M. Rosen (Eds.), Self-
dissertation, niversity of Toronto. mutilation theory, research, and treatment (pp.
Santa Mina, E. E., Gallop, R. (1998). Childhood 39-53). New York: Guilford.
sexual and physical abuse and adult self-harm and Welch, S., Patterson, J., Shaw, R., Stewart-Brown, S.
suicidal behavior: A literature review. Canadian (2009). Family relationships in childhood and
ournal of Psychiatry, , 793-800. common psychiatric disorders in later life: Systematic
Santa Mina, E. E., Gallop, R., Links, P., Heslegrave, R., review of prospective studies. British ournal of
Pringle, D., Wekerle, C., et al. (2006). The Self-Injury Psychiatry, , 392-398.
uestionnaire: Evaluation of the psychometric Whitlock, J., Knox, K. L. (2007). The relationship
properties in a clinical population. ournal of between self-injurious behaviour and suicide in a
Psychiatric and Mental ealth Nursing, , 221-227. young adult population. Archives of Pediatric and
Schnyder, ., Valach, L. (1997). Suicide attempters Adolescent Medicine, (7), 634-640.
in a psychiatric emergency room population. eneral Windle, M., Windle, R. C., Scheidt, D. M., Miller,
ospital Psychiatry, (2), 119-129. G. B. (1995). Physical and sexual abuse and
Shneidman, E. (1985). Classifications and approaches. associated mental disorders among alcoholic
In E. Shneidman (Ed.), De nition of suicide inpatients. American ournal of Psychiatry, (9),
(pp. 23-40). New York: John Wiley. 1322-1328.
Skegg, K. (2005). Self-harm. ancet, , 1471-1483. World Health Organization. (2010). Suicide prevention.
Statistics Canada. (2010). Summary tables: Suicides and Geneva: Author. Retrieved January 8, 2010, from
suicide rate, by sex and by age group. Retrieved http://www.who.int/mental health/prevention/en/
August 31, 2010, from http://www40.statcan.ca/101/ 2010.
cst01/hlth66a-eng.htm.
470 PART FIVE Critiquing Research

INTRODUCTION TO CRITIQUE 2 Review of Literature and Definitions

The article Self-Harm Intentions: Can They Be The literature review was well organized. In a
Distinguished Based pon a History of Child- systematic manner, the relationships among the
hood Physical and Sexual Abuse (Santa Mina, core elements of CP/SA and adult-harm, suicidal
2010) is examined in terms of its quality and the ideation, suicidal behaviours, and suicide were
potential usefulness of the findings for applica- clarified. The use of subheadings increased the
tion to nursing practice. The design of this was readability of the literature review. The references
level IV, inasmuch as it was a nonexperimental were mostly primary sources (e.g., Connors,
study. 1996 Windle, Windle, Scheidt, Miller, 1995),
with the addition of a few secondary sources
Title (e.g., Santa Mina Gallop, 1998). Most sources
The title re ects the major purpose of this study. had been published within 10 years of the article’s
publication, and those more than 10 years older
Abstract are landmark studies. Although the research on
The abstract meets most of the requirements of the study’s topic is considerable, as articulated in
thoroughness, except for the sample size. It con- the literature review, Santa Mina (2010) identified
tains the purpose, the research design, the find- a gap in the literature by commenting that The
ings, and the results. suicide and self-harm literature and conceptual
reviews do not reach consensus on the extent, if
Problem and Purpose any, of overlap of intentions between the two
Santa Mina (2010) stated that the purpose of the phenomena, or whether they are in fact the same
study was to investigate the in uence of CP/SA phenomenon (p. 125).
in differentiating between self-harm intentions
and suicidal behaviour in a clinical sample of Theoretical Framework
adults admitted to an inner-city teaching hospital The theoretical framework helped structure
with self-harm/suicidal behaviour (p. 123). the study, inasmuch as it consisted of conceptu-
The identification of the research problem is alizations from sociological suicide theories, with
very well organized, with an easy ow of the the following main factors: suicide intention,
linkages. Santa Mina (2010) outlined the signifi- depression, hopelessness, affect regulation, and
cance of the study by stating the incidence of dissociation. The linkages of these factors to
self-harm behaviour and suicidal behaviours. support the study are clearly articulated. In addi-
Next, the author offered information on the causes tion, conceptual definitions serve to enhance the
of and intentions involved in self-harm. The con- clarity.
cepts of affect dysregulation and dissociation
were introduced. The rationale for the study was Hypotheses and Research Question
explicitly stated as follows: Santa Mina (2010) explicitly stated the research
hypotheses as follows: that patients who engaged
Although an association between CP/SA and adult in a recent episode of self-harm behaviour and
psychiatric disorders has been widely reported . . ., its reported a history of CP/SA, in comparison with
influence on self-harm and suicidal intentions is not
those with no such history, would have (1) less
clear. Knowledge about the influence of CP/SA in
self-harm intentions may serve to guide assessments suicide intention, (2) more reasons related to
of and interventions for either similar or disparate affect regulation and dissociation, (3) greater dis-
clinical populations. (p. 124) sociation, (4) greater affective lability, (5) less
 CHAPTER 19 Critiquing Quantitative Research 471

hopelessness, and (6) no difference in depression the abused group was further divided: partici-
(p. 128). pants who had been physically abused only, those
No research question is stated, but such a state- who had been sexually abused only, and those
ment could be articulated from the hypotheses as with a history of CP/SA thus, the study, with the
follows: What is the difference in suicide inten- addition of the nonabused group, had a total of
tion, affect regulation and dissociation, affect four groups. The major dependent variables for
liability, level of hopelessness, and level of this study were suicide intention, reasons for self-
depression in individuals who reported a CP/SA harm, affect regulation and dissociation, hope-
history in comparison with those who did not lessness, and depression. Demographic variables,
have a CP/SA history such as age and gender, were used to determine
homogeneity in the sample.
Sample
A convenience sample was used this is appropri- Internal Validity
ate as it is a nonexperimental study. A probability Santa Mina (2010) made no mention of possible
method such as randomization (assignment to a threats to internal validity that would inordinately
CP/SA group or to a non-CP/SA group) would decrease confidence in the results. This study was
have been unethical. The inclusion criteria were based on self-report measures, which could have
age of 18 years or older ability to speak English led to social desirability bias or acquiescence
presenting to a hospital after an episode of self- bias, which in turn could have affected the results.
harm and being admitted voluntarily to the hos-
pital, with competence to give consent under the External Validity
Mental ealth Act of ntario, Canada, Generalizability is limited to the sample because
(Santa Mina, 2010, p. 129). The study had no of the effect of nonprobability convenience sam-
exclusion criteria. pling. Santa Mina (2010) decided to discontinue
Recruitment took place weekdays only, data collection because the sample size for the
because of constraints in support, and Santa Mina nonabused group was very small: data analysis
informed all clinical staff about the study. These yielded nonsignificant findings, and the risk of a
staff members identified patients who met the type II error was low (p. 137), and so this study
inclusion criteria. A power analysis indicated that exhibited selection effects.
64 patients were needed in each group (the abused
group and the nonabused group). However, the Legal/Ethical Issues
nonabused group contained only 19 participants, The university’s ethics review board gave
and so a final power analysis should have been approval for the study, and the relevant permis-
performed. sions from hospital leaders were obtained. The
informed consent form was obtained, and clinical
Research Design staff members were available to participants if
Santa Mina (2010) explicitly stated that she used they became upset. A contact phone number was
a nonexperimental, comparative design. There given to participants for crisis support should
were two groups: one of abused patients and one they become upset at any time after completion
of nonabused patients. These were naturally of the study.
occurring groups. This design is consistent with
the purpose of the research. Instruments
The independent variable was type of group Santa Mina (2010) used several instruments and
(abused, nonabused). Also, to conduct the ANOVA, provided a very detailed table outlining the
472 PART FIVE Critiquing Research

concept: number of items, scoring method, and group with both physical and sexual abuse was
interpretation of scores. She collected demo- noted.
graphic information about age and gender. Overall, the findings related to each hypothesis
are clearly stated.
Reliability and Validity
A very thorough description of the reliability and Discussion
validity of each instrument was provided in Table Santa Mina (2010) stated that although the results
19C2-1 for previous studies and Santa Mina’s were not statistically significant, they had clinical
study. significance inasmuch as participants, regardless
of abuse history, reported multiple intentions.
Results Participants also reported moderate levels of sui-
Santa Mina (2010) used descriptive statistics to cidality and high levels of hopelessness and
determine homogeneity between the abused and depression. The findings of this study were appro-
nonabused groups. Although gender was distrib- priately related to results of previous studies on
uted evenly, women were more likely to have self-harm and suicidal behaviours and the differ-
been abused. Overdose was found to be the most ences between the genders.
frequent type of self-harm behaviour, and its Several considerations for future research
prevalence was distributed evenly between the were mentioned: for example, Future studies
genders. should aim for large sample sizes, to accommo-
A description of the participants’ results is date the breadth of abuse types as well as the
detailed in the text (Santa Mina, 2010, p. 130). myriad sociological factors that can affect inten-
Evidence to support the hypotheses was detailed tions (Santa Mina, 2010, p. 137).
in the text as follows: Self-harm/suicidal inten- Clinical implications from the results are
tions did not differ for the non-abused and abused made: patients who are thinking about or engag-
groups (p. 130), and the sample overwhelm- ing in self-harm/suicidal behaviours need to be
ingly (98 ) reported multiple reasons for the assessed for the full spectrum of intentions,
self-harm episode. . . . The abused group revealed regardless of gender, self-harm method, or known
greater dissociation than the non-abused group, abuse history (Santa Mina, 2010, p. 138).
as well as greater affective liability (p. 136).
These results were, however, not statistically Limitations and Conclusion
significant. Santa Mina (2010) discussed the limitation of the
Santa Mina (2010) also examined the SI sample size and its effect on type II error and
affect subscale and found that participants placed it within the context of suicide intention,
who had been both physically and sexually reasons for self-harm, affect regulation, dissocia-
abused reported the largest mean number of tion, depression, and hopelessness (p. 137). She
reasons (p. 136) these results were statistically described the significance of her study: This is
significant. Santa Mina also tested for a differ- the first published clinical study to investigate the
ence in dissociation and depression across the in uence of CP/SA on self-harm/suicide inten-
four groups. Although the results were not statis- tions among men and women admitted to an acute-
tically significant, a trend of higher scores in the care unit for these behaviours (p. 139).
 CHAPTER 19 Critiquing Quantitative Research 473

Ploeg, J., Denton, M., Tindale, J., Hutchison, B., Brazil,

CRITICAL THINKING CHALLENGES K., Akhtar-Danesh, N., . . . Millen Plenderleith, J.
(2009). Older adults’ awareness of community health
■ Discuss the ways in which the stylistic and support services for dementia care. Canadian
considerations of a journal affect the researcher’s ournal of Aging, (4), 359-370.
ability to present the research findings of a Santa Mina, E. (2010). Self-harm intentions: Can they
quantitative study. be distinguished based upon a history of childhood
■ Are critiques of quantitative studies valid when a physical and sexual abuse Canadian ournal of
student or a practising nurse writes them? What Nursing Research, (4), 122-143.
level of quantitative study is best for you as a Santa Mina, E. E., Gallop, R. (1998). Childhood
consumer of research to critique? What sexual and physical abuse and adult self-harm and
assumptions did you use to make this suicidal behavior: A literature review. Canadian
ournal of Psychiatry, , 793-800.
determination?
Windle, M., Windle, R. C., Scheidt, D. M., Miller,
■ What is essential for you as a consumer of
G. B. (1995). Physical and sexual abuse and
research to use when you critique a quantitative associated mental disorders among alcoholic
research study? Discuss the ways you might use inpatients. American ournal of Psychiatry, (9),
Internet resources now or in the future when 1322-1328.
you critique studies.

REFERENCES
Calsyn, R. J., Roades, L. A., Calsyn, D. S. (1992).
FOR FURTHER STUDY
Acquiescence in needs assessment studies of the Go to Evolve at http://evolve.elsevier.com/
elderly. The erontologist, , 246-252. Canada/LoBiondo/Research for Audio Glossary, how-to
Connors, R. (1996). Self-injury in trauma survivors: instructions for Writing Proposals for Funding, and
1. Functions and meanings. American ournal of additional research articles for practice in reviewing and
rthopsychiatry, (2), 197-206. critiquing.
This page intentionally left blank
475

PA RT SI X
PART ONE Part Title

Application of Research:
Evidence-Informed Practice
20 Developing an Evidence-
Informed Practice
RESEARCH VIGNETTE thesia, reduced number of births
by caesarean section, and fewer
Nursing Best Practices reports by women of negative
labour experiences. In addition,
my PhD research also built upon
Barbara Davies, RN, PhD It is fascinating to contemplate another Cochrane systematic
Professor the forces that in uenced my review in which intermittent fetal
School of Nursing
career development and, in par- auscultation was found to reduce
Faculty of Health Sciences
University of Ottawa; ticular, the development of my rates of Caesarean section for
Co-Director, Nursing Best Practice Research research program on evidence- women at low risk for complica-
Unit informed practice. One of the tions and thus concluded that aus-
University of Ottawa and RNAO most in uential people in my cultation is preferable to electronic
Ottawa, Ontario
early career was Donna Diers, the fetal monitoring. Although the
Give to the world the best you have Dean of Nursing at Yale niver- provision of labour support and
and the best will come back to you. sity. I was employed as a clinical the use of intermittent fetal aus-
—Author unknown instructor at Yale and recall cultation were recommended by
As a graduating student in nursing Donna’s passionately pleading for the Society of Obstetricians and
at the niversity of Toronto in nursing research to provide the Gynaecologists of Canada and
1974, I had to supply a quotation data to demonstrate the power by the World Health Organization,
about my philosophy of life to of nursing care to improve health. several observational studies
accompany my yearbook photo- She strongly encouraged teaching revealed that the majority of
graph. I still recall being perplexed staff at Yale to conduct nursing women at low risk who gave birth
about what to state. Eventually, I research that would make a dif- in Canada were not receiving
selected the above quotation about ference in practice. either of these beneficial interven-
doing your best. Little did I know My PhD supervisor, Ellen tions. This lack of knowledge
that more than 30 years later, I Hodnett, who holds the Heather translation became the challenge
would be the codirector of the M. Reisman Chair in Perinatal of my doctoral studies. In other
Nursing Best Practice Research Nursing Research at the niver- words, the knowledge from the
nit at the niversity of Ottawa, sity of Toronto, also substantially Cochrane reviews was needed so
a dynamic collaboration with in uenced my knowledge and that women at childbirth would
the politically active Registered skills related to best practices. receive care that was consistent
Nurses’ Association of Ontario. Ellen is one of the early leaders of with clinical practice guidelines.
Our research unit won the Sigma the Cochrane Collaboration, an What could I do to help healthy
Theta Tau international practice- international research initiative to women at low risk receive sup-
academe innovative collaboration answer clinical questions by con- port at childbirth without the
award in 2009. In 2011, I received ducting systematic reviews. In encumbrance of continuously
the distinguished alumnus award Cochrane Reviews, authors strive being hooked up to monitoring
from my alma mater, now named to determine the most effective machines In collaboration with
the Lawrence S. Bloomberg Faculty health care interventions. Ellen my PhD thesis committee—which
of Nursing. These awards are two was the lead author of a review consisted of an obstetrician,
examples of the best coming about the provision of labour an epidemiologist, and nursing
back to me in recognition of many support to women at childbirth. research experts—I designed a
hours of hard work. The best also My PhD research built upon multifaceted intervention that
includes teaching undergraduate Dr. Hodnett’s systematic review included an educational work-
and graduate students to use findings that continuous labour shop for nurses with tools for
research to make a difference in the support provides the benefits of applying research results to prac-
health of Canadians. reduced use of analgesia or anaes- tice. Assessment forms, protocols

476
based on the research results, Fortunately, a very special op- translation research in maternal
and case study exercises were portunity arose at this time to infant health about public health
included. co-lead, with Dr. Nancy Edwards, networks to support smoking ces-
The intervention received posi- the evaluation of the emerging sation and breast-feeding counsel-
tive evaluations from the partici- Nursing Best Practice Guidelines ling by family practice nurses
pating nurses, and the majority of Program of the Registered Nurses’ for pregnant women. What topics
staff nurses (>80 ) attended the Association of Ontario. The guide- would you research as a graduate
workshop. I am pleased to report lines include a synthesis of the student
that this labour support workshop available research evidence devel- Some academicians criticize
and a fetal health surveillance oped by expert panels on priority the notion of best practices, com-
workshop, both of which were topics, such as fall prevention, menting on the harm done by a
first offered in 1995 as part of pressure ulcers, asthma control, hierarchy of evidence that unduly
my PhD research, are still being breast-feeding, and support for glorifies the randomized con-
offered regularly in Ottawa (http:// families. trolled trial as the gold standard
www.cmnrp.ca/en/pppeso/Home Over the next 11 years, Nancy of research methodology. My
p2974.html). and I—along with many others— view is that a randomized con-
A note of caution for other pro- conducted studies of implemen- trolled trial is the preferred
fessionals who are considering the tation interventions, surveys of research design to determine
development of similar know- nurses’ attitudes, qualitative inter- whether one treatment is more
ledge translation interventions to views of barriers and facilitators, effective than another treatment.
improve practice: Many nurses observations of teaching tools, However, health has multiple
and physicians have extensive and chart audits of nursing prac- determinants, and health care is
experience, and it is important to tice and patient outcomes. The provided in a complex environ-
recognize their knowledge and mandate of the Nursing Best Prac- ment by many types of health care
expertise. sing a respectful, col- tice Research nit (http://www. professionals. Thus, it is equally
laborative approach works well. nbpru.ca/) is to incorporate the important to conduct qualitative
Ask health care professionals best knowledge into nursing and research studies to better under-
what they think is needed to health care, enhance practice, and stand the factors in uencing
improve practice to be consistent improve health and system out- the promotion of health for all
with research results, and you will comes. The goal is to promote best people.
probably hear some excellent nursing knowledge. The research Finally, I offer a few words
ideas. unit has 52 individual members about values, preferences, and
Shortly after completing my and 22 organizational members. best practices. At the core of best
PhD in 2000, I received a Career Currently, my research on best practices is the notion that deci-
Scientist award from the Ministry practices addresses sustainability sions and interventions need to be
of Health and Long-Term Care factors, interprofessional collabo- patient centred. Assessing pre-
(Ontario) for a 5-year program ration, and system changes. I am ferences, clarifying values, and
of research on maternal-infant supervising or a thesis committee informing patients and their fami-
knowledge translation. Although I member for 12 graduate students lies about the risks and benefits of
was excited to be selected as one (MScN, PhD, postdoctoral) work- options are essential. Early in my
of six awardees (the only nurse) in ing on such important topics as research career, while doing my
a tough provincial competition, audit and feedback of patient out- MScN thesis on factors in uenc-
I was also apprehensive about comes, implementation of guide- ing the decisions of women of
whether I could be successful in lines in academic programs, and advanced maternal age to have
major national research competi- community environmental health. genetic amniocentesis, I discov-
tions, a requirement to continue to Some of these graduate stu- ered that different women per-
receive the ongoing funding. dents are continuing knowledge ceived the risk of 1 per 100

477
very differently. Each woman, her what constitutes best practices. less, the quest to achieve excel-
spouse, and the genetic counsellor Best practices require the integra- lence in nursing practice and
had their own and sometimes con- tion of research knowledge with health services still drives my
icting perspectives. Conducting patient preferences and clinical research journey. What drives
this genetic research heightened judgement. your work in practice, education,
my awareness of differences in In retrospect, although best or research What type of evi-
patient and provider values and practice is an admirable goal, I dence counts for you What does
the need for thoughtful re ection now question whether such a state best practices mean to you ■
about my own values regarding can ever be achieved. Neverthe-

478
C H A PTER 20

Developing an Evidence-
Informed Practice
Marita Titler | Susan Adams | Cherylyn Cameron

LEARNING OUTCOMES
After reading this chapter, you will be able to do the following:
• Differentiate among conduct of nursing research, research utilization, and evidence-informed
practice.
• Describe the steps of evidence-informed practice.
• Identify three barriers to evidence-informed practice and strategies to address each.
• List three sources for finding evidence.
• Describe strategies for implementing evidence-informed practice changes.
• Identify steps for evaluating an evidence-informed change in practice.
• Use research findings and other forms of evidence to improve the quality of care.

KEY TERMS
conduct of research evidence-informed practice problem-focused triggers
dissemination guidelines research utilization
evaluation knowledge-focused triggers translation science
evidence-informed practice opinion leaders

STUDY RESOURCES
Go to Evolve at http://evolve.elsevier.com/Canada/LoBiondo/Research for Audio
Glossary, how-to instructions for Writing Proposals for Funding, and additional
research articles for practice in reviewing and critiquing.

479
480 PART SIX Application of Research: Evidence-Informed Practice

EVIDENCE-INFORMED HEALTH CARE PRACTICES ARE organizational culture that values and supports
available for a number of conditions, such as use of evidence, and by localization of the evi-
asthma, smoking cessation, heart failure, and dence for use in a specific health care setting
management of diabetes. However, these prac- (Greenhalgh, Robert, MacFarlane, Bate,
tices are not always implemented in care delivery, Kyriakdiou, 2004 Rogers, 2003).
and variation in practices abound (Ward, Evans, Translation of research into practice is a mul-
Spies, Roberts, Wakefield, 2006). Availability tifaceted, systemic process of promoting adoption
of high-quality research does not ensure that the of evidence-informed practices in delivery of
findings will be used to affect patient outcomes. health care services that goes beyond dissemina-
Research findings in the nited States and the tion of evidence-informed guidelines (Rogers,
Netherlands suggest that 30 to 40 of patients 2003). isse ination is the communication of
are not receiving evidence-informed care, and research findings dissemination activities take
20 to 25 of patients are receiving unneeded many forms, including publications, conferences,
or potentially harmful care (Graham, Logan, consultations, and training programs (Adams
Harrison, Straus, Tetroe, Caswell, Robinson, Titler, 2010), but promoting knowledge uptake
2006). The use of evidence-informed practices is and changing practitioner behaviour requires
now an expected standard in many institutions to active interchange with those in direct care (Scott,
prevent nosocomial events such as injury from Plotnikoff, Karunamuni, Bize, Rodgers, 2008
falls, Foley catheter–associated urinary tract Titler, Herr, Brooks, ian-Jin, Ardery, Schilling,
infections, and stages 3 and 4 pressure ulcers. Clarke, 2009).
However, implementing such evidence-informed Although the science of translation is young,
safety practices is a challenge and requires use of the effectiveness of interventions for promoting
strategies that address the complexity and systems adoption of evidence-informed practices is being
of care, individual practitioners, senior leader- studied, and funding is supporting research in this
ship, and ultimately changing health care cultures area (Bootsmiller et al., 2004 Smith, Williams,
to be evidence-informed practice environments Owen, Rubenstein, Chaney, 2008 Stetler,
(Leape, 2005). Mc uenn, Demkis, Mittman, 2008).
Conduct of research is only the first step in In addition, more evidence is now available to
improving practice through the use of research. guide selection of strategies for translating
Because of the gap between discovery and use of research into practice (Brooks, Titler, Ardery,
knowledge in practice (Bootsmiller, Yankey, Herr, 2009 Gravel, L gar , Graham, 2006
Finch, Ward, Vaughn, Welke, Doebbeling, Titler, 2008). This chapter presents an overview
2004 Davey, Brown, Fenelon, Finch, Gould, of evidence-informed practice, the process of
Hartman, Wiffen, 2005 Titler, 2008), efforts implementing evidence in practice to improve
must be concentrated on developing methods to patient outcomes, and a description of translation
speed translation of research findings into prac- science.
tice. Development and dissemination of evidence-
informed practice guidelines are essential steps, OVERVIEW OF EVIDENCE-INFORMED
but each alone does little to promote knowledge PRACTICE
uptake by direct care providers (Clancy, Slutsky, The relationships among conduct, dissemination,
Patton, 2004). and use of research are illustrated in Figure 20-1.
Promoting use of evidence in practice is an Conduct o research is the analysis of data col-
active process that is facilitated partly by model- lected from a homogeneous group of participants
ling and imitation of other professionals who who meet study inclusion and exclusion criteria
have successfully adopted the innovation, by an for the purpose of answering specific research
 CHAPTER 20 Developing an Evidence-Informed Practice 481

Improve
quality of
care

Utilize
Identify
findings
questions
in practice

Disseminate Conduct
knowledge research

Generate
new
knowledge
FIGURE 20-1 Model of the relationship among conduct, dissemination, and use of research.
Redrawn from Weiler, K., Buckwalter, K., & Titler, M. (1994). Debate: Is nursing research used
in practice? In J. McCloskey & H. Grace (Eds.), Current issues in nursing (4th ed.). St. Louis:
Mosby.

questions or testing specified hypotheses. Re- to differentiate among the terms evidence-
search design, methods, and statistical analyses based practice and evidence-informed practice
are guided by the state of the science in the area Evidence-informed practice extends beyond the
of investigation. Traditionally, conduct of re- early definitions of evidence-based practice.
search has included dissemination of findings Many of the models explored in this chapter refer
through research reports in journals and at scien- to evidence-based practice, inasmuch as they
tific conferences. In comparison, research utili were developed before the use of the term
ation is the process of using research findings to evidence-informed practice The models are still
improve patient care this process involves imple- valid and help guide professionals toward actual-
menting sound research-based innovations in izing research into best practices. Best evidence
clinical practice dissemination of scientific includes empirical evidence from systematic
knowledge critique of studies synthesis of re- reviews, from randomized controlled trials, and
search findings determining applicability of from other scientific methods such as descriptive
findings for practice developing an evidence- and qualitative research, as well as information
informed standard or guideline implementing from case reports, scientific principles, and expert
the standard and evaluating the practice change opinion. When enough research evidence is avail-
with respect to staff, patients, and cost/resource able, practice should be guided by this evidence,
utilization (Titler, Kleiber, Steelman, Rakel, in conjunction with clinical expertise and patients’
Budreau, Everett, Goode, 2001). values. In some cases, however, a sufficient
Evidence in or ed ractice is the conscien- research base may not be available, and health
tious and judicious use of current best evidence care decision making is derived principally from
in conjunction with clinical expertise and patient nonresearch evidence sources such as expert
values to guide health care decisions (Titler, opinion and scientific principles (Titler et al.,
2006). As noted in Chapter 1, it is important 2001). When more research is completed in a
482 PART SIX Application of Research: Evidence-Informed Practice

specific area, the research evidence must be Cronenwett (1995) and others described two
incorporated into evidence-informed practice forms of using research evidence in practice: con-
(Titler, 2006). As illustrated in the knowledge ceptual and decision driven (Estabrooks, 2004).
generation and use cycle (see Figure 20-1), appli- Conceptual forms in uence the thinking of the
cation of research findings in practice may not health care professional, but not necessarily the
only improve quality care but also create new and action. Exposure to new scientific knowledge
exciting questions to be addressed through the occurs, but the new knowledge may not be used to
conduct of research. change or guide practice. An integrative review of
The terms research utili ation and evidence- the literature, formulation of a new theory, or gen-
informed practice are sometimes used inter- erating new hypotheses may be the result. se of
changeably. However, although these two terms knowledge in this way is referred to as no ledge
are related, they are not one and the same. creep or cognitive application It is often used by
Evidence-informed is a broader term that encom- individuals who read and incorporate research into
passes not only research utilization but also the their critical thinking (Weiss, 1980). Decision-
use of case reports and expert opinion in deciding driven forms of using evidence in practice encom-
the practices to be used in health care. If evidence- pass application of scientific knowledge as part of
informed practice is defined as the conscious and a new practice, policy, procedure, or intervention.
judicious use of the current best evidence in the In this type of application of research findings, a
care of patients and delivery of health care ser- critical decision is reached to endorse current prac-
vices, then research utilization is a subset of tice or to change it on the basis of review and
evidence-informed practice that focuses on the critique of studies applicable to that practice.
application of research findings. Examples of decision-driven models of using
research in practice are the Iowa Model of
Use of Evidence in Practice Evidence-Based Practice to Promote uality Care
Nursing has a rich history of using research in (Titler et al., 2001), the Ottawa Model of Research
practice, pioneered by Florence Nightingale, who se (OMR Logan Graham, 1998), the Pro-
used data to change practices that contributed to moting Action on Research Implementation in
high mortality rates in hospitals and communities Health Services (PARIHS) model (Rycroft-
(Nightingale, 1858, 1859, 1863a, 1863b). Malone, Kitson, Harvey, McCormack, Seers,
Although during the early and mid-1900s, few Titchen, Eastabrooks, 2002), and the Conduct
nurses built on the solid foundation of research and tilization of Research in Nursing (C RN)
utilization exemplified by Nightingale (Titler, model (Haller, Reynolds, Horseley, 1979
1993), the nursing profession has provided major Horsley, Crane, Crabtree, Wood, 1983).
leadership for improving care through applica- Multifaceted active dissemination strategies
tion of research findings in practice (Kirchhoff, are needed to promote use of research evidence
2004). Today nurses are being prepared as scien- in clinical and administrative health care decision
tists in nursing, leading the way in translation making, and they must address both the individ-
science, and, as a result, the scientific body of ual practitioner’s and the organization’s per-
nursing knowledge is growing (Estabrooks, spectives (Titler, 2008). When nurses decide
Derksen, Winther, Lavis, Scott, Wallin, individually what evidence to use in practice,
Profetto-McGrath, 2008 Titler, 2008 Titler considerable variability in practice patterns
et al., 2009). It is now every nurse’s responsibil- results, which can potentially lead to adverse
ity to facilitate the use of nursing knowledge patient outcomes. For example, a solely indi-
in practice. vidual perspective of evidence-informed
 CHAPTER 20 Developing an Evidence-Informed Practice 483

practice would leave the decision about use of extraneous variables (e.g., behaviours and facili-
pressure ulcer prevention practices to each nurse. tators) that may in uence adoption of evidence-
Some nurses may be familiar with the research informed practices (e.g., organizational size,
findings for pressure ulcer prevention, whereas characteristics of users ICEBeRG, 2006). Two
others may not be. As a result, different nurses models are explored in this chapter: the Iowa
may use con icting practices, especially inas- Model of Evidence-Based Practice to Promote
much as shifts change every 8 to 12 hours. From uality Care and the Ottawa Model of Research
an organizational perspective, policies and proce- se.
dures are based on research, and then adoption of
these practices by nurses is systematically pro- The Iowa Model of Evidence-Based Practice to
moted in the organization (Squires, Moralejo, Promote Quality Care
LeFort, 2007). An overview of the Iowa Model of Evidence-
Based Practice to Promote uality Care, as an
example of a practice model, is illustrated in
Models of Evidence-Informed Practice Figure 20-2. This model has been widely dis-
Multiple models of evidence-informed practice seminated and adopted in academic and clinical
and translation science are available. Common settings. Since the original publication of this
elements of these models are syntheses of evi- model in 1994 (Titler, Kleiber, Steelman, Goode,
dence, implementation, evaluation of the effect Rakel, Barry-Walker, Buckwalter, 1994),
on patient care, and consideration of the context/ Titler and colleagues have received more than
setting in which the evidence is implemented. 300 written requests to use the model for publica-
Grol and associates (2007) provided a summary tions, presentations, graduate and undergraduate
of models. Included in their summary relevant research courses, and clinical research programs.
to quality improvement and implementation of It is an organizational, collaborative model that
change in health care were cognitive, educational, incorporates conduct of research, use of research
motivational, social interactive, social learning, evidence, and other types of evidence (Titler
social network, and social in uence theories, et al., 2001). Titler and colleagues adopted the
as well as models related to team effective- definition of evidence-based practice as the con-
ness, professional development, and leadership. scientious and judicious use of current best evi-
Additional work by the Improved Clinical dence to guide health care decisions. Levels of
Effectiveness through Behavioural Research evidence range from randomized controlled trials
Group (ICEBeRG) resulted in the development to case reports and expert opinion.
of a database consisting of planned action models, In this model, knowledge- and problem-
frameworks, and theories that explicitly describe focused triggers lead staff members to question
both the concepts and action steps to be consid- current nursing practice and whether patient care
ered or taken. This database was developed can be improved through the use of research find-
from a search of social science, education, and ings. If, through the process of literature review
health literature that focused on practitioner and critique of studies, staff members find that the
or organizational change (http://www.iceberg- number of scientifically sound studies is not suf-
grebeci.ohri.ca/research/kt theories db.html). ficient for use as a base for practice, they consider
Although review of these models is beyond the conducting a study. Nurses in practice collaborate
scope of this chapter, implementing evidence in with scientists in nursing and other disciplines to
practice must be guided by a conceptual model to conduct clinical research that addresses practice
organize the strategies being used and to clarify problems encountered in the care of patients.
484 PART SIX Application of Research: Evidence-Informed Practice

Problem-Focused Triggers Knowledge-Focused Triggers

1. Risk management data 1. New research or other literature
2. Process improvement data 2. National agencies or organizational
3. Internal/external benchmarking data standards and guidelines
4. Financial data 3. Philosophies of care
5. Identification of clinical problem 4. Questions from institutional standards committee

Consider Is this topic

No
other a priority for the
triggers organization?

Yes
Form a team

Assemble relevant research and related literature

Critique and synthesize research for use in practice

Is there
Yes a sufficient No
research
base?
Pilot the Change in Practice
1. Select outcomes to be achieved Base Practice on Other Conduct
2. Collect baseline data Types of Evidence research
3. Design evidence-based practice (EBP) guideline(s) 1. Case reports
4. Implement EBP on pilot units 2. Expert opinion
5. Evaluate process and outcomes 3. Scientific principles
6. Modify the practice guideline 4. Theory

Is change
Continue to evaluate No appropriate for Yes
quality of care and Institute the change in practice
adoption in
new knowledge practice?
Monitor and Analyze Structure,
Disseminate results Process, and Outcome Data
• Environment
• Staff
! a decision point • Cost
• Patient and family
FIGURE 20-2 The Iowa Model of Evidence-Based Practice to Promote Quality Care.
Redrawn from Titler, M. G., Kleiber, C., Steelman, V. J., Rakel, B. A., et al. (2001). The Iowa
Model of Evidence-Based Practice to Promote Quality Care. Critical Care Nursing Clinics of North
America, 13(4), 497–509. Copyright Elsevier 2001.
 CHAPTER 20 Developing an Evidence-Informed Practice 485

Findings from such studies are then combined basis of evaluation data, and the change is imple-
with findings from existing scientific knowledge mented with additional patient populations for
to develop and implement these practices. If which it is appropriate. Patient/family, staff, and
research is insufficient for guiding practice, and fiscal outcomes are monitored. Organizational
if conducting a study is not feasible, other types support and administrative support are important
of evidence (e.g., case reports, expert opinion, factors for success in the use of evidence in care
scientific principles, theory) are used or combined delivery.
with available research evidence to guide prac-
tice. Priority is given to projects in which a high The Ottawa Model of Research Use
proportion of practice is guided by research evi- Logan and Graham (1998) developed the OMR ,
dence. Practice guidelines usually re ect research a model for interdisciplinary health care research
and nonresearch evidence and therefore are called use. The framework was created to be used
evidence-informed practice guidelines by policymakers seeking to increase the use of
An evidence-informed practice guideline is health research by practitioners, as well as by
developed from the available evidence. The rec- researchers interested in studying the process by
ommended practices, based on the relevant evi- which research becomes integrated into practice
dence, are compared with current practice, and a (p. 228). They identified the following six com-
decision is made about the necessity for a practice ponents of research utilization: (1) the practice
change. If a practice change is warranted, changes environment, (2) potential adopters, (3) the
are implemented through a process of planned evidence-informed innovation, (4) transfer strate-
change. The practice is first implemented with a gies, (5) adoption, and (6) health-related and
small group of patients, and it is evaluated. The other outcomes (Figure 20-3). Constant assess-
evidence-informed practice is then refined on the ment, monitoring, and evaluation parallel the

Practice Environment
• Structural
• Social
• Patients
• Other
Transfer Strategies Outcomes
Potential Adopters Adoption
• Diffusion • Patient
• Knowledge • Decision
• Dissemination • Practitioner
• Attitudes • Use
• Implementation • Economic
• Skill

Evidence-Informed Innovation
• Translation process
• Innovation

Assess + Monitor + Evaluate

FIGURE 20-3 The Ottawa Model of Health Care Research Use.
Adapted from Logan, J., & Graham, I. (1998). Toward a comprehensive interdisciplinary model
of health care research use. Science Communication, 20(2), 229. Reprinted by Permission of
SAGE Publications.
486 PART SIX Application of Research: Evidence-Informed Practice

progression through the components. As barriers the necessary resources to successfully complete
are identified, strategies are developed to sur- the project.
mount them and to enhance supports. Individuals should work collectively to achieve
consensus in topic selection. Working in groups to
STEPS OF EVIDENCE-INFORMED PRACTICE review performance improvement data, brainstorm
The Iowa Model of Evidence-Based Practice to about ideas, and achieve consensus about the final
Promote uality Care (Titler et al., 2001 see selection is helpful. For example, a unit staff
Figure 20-2), in conjunction with Rogers’ (1995, meeting may be used to discuss ideas for evidence-
2003) diffusion of innovations model, provides informed practice quality improvement commit-
guiding steps in actualizing evidence-informed tees may identify several practice areas in need of
practice. A team approach is most helpful in fos- attention (e.g., urinary tract infections in older
tering a specific evidence-informed practice, with patients, reducing the incidence of pressure ulcers)
one person in the group providing leadership for an evidence-informed practice task force may be
the project. appointed to select and address a clinical practice
issue (e.g., pain management) or surveying a
Selection of a Topic panel of experts may be used to prioritize areas for
The first step in carrying out an evidence-informed evidence-informed practice. Criteria to consider
practice project is to select a topic. Ideas for when a topic is selected are outlined in Box 20-1.
evidence-informed practice come from several Table 20-1 is a helpful chart for selecting a topic.
sources categorized as problem- and knowledge-
focused triggers. Proble ocused triggers are Helpful Hint
research ideas identified by staff through quality Regardless of which method is used to select an
improvement, risk surveillance, benchmarking evidence-informed practice topic, it is critical that the
data, financial data, or recurrent clinical prob- staff members who will implement the potential prac-
tice changes are involved in selecting the topic and view
lems. For example, the increased incidence of it as contributing significantly to the quality of care.
Clostridium dif cile on a long-term care unit,
resulting in increased morbidity, is a problem-
BOX 20-1
focused trigger because it raises concern among
hospital staff. SELECTION CRITERIA FOR AN EVIDENCE-
INFORMED PRACTICE PROJECT
no ledge ocused triggers are research
ideas generated when staff members read research, 1. The priority of this topic for nursing and for the
organization
listen to scientific papers at research conferences, 2. The magnitude of the problem (small, medium,
or encounter evidence-informed practice guide- large)
lines published by federal agencies or specialty 3. Applicability to several or few clinical areas
4. Likelihood of the change to improve quality of care,
organizations. Examples of such triggers include decrease length of stay, contain costs, or improve
ideas about pain management, assessing place- patient satisfaction
ment of nasogastric and nasointestinal tubes, and 5. Potential problems associated with the topic and
capability to diffuse them
use of saline to maintain patency of arterial lines. 6. Availability of baseline quality improvement or risk
Sometimes topics arise from a combination of data that will be helpful during evaluation
problem- and knowledge-focused triggers, such 7. Multidisciplinary nature of the topic and ability to
as the length of bed rest time after femoral artery create collaborative relationships to effect the
needed changes
catheterization. In selecting a topic, nurses must 8. Interest and commitment of staff to the potential
consider how the topic fits with organization, topic
department, and unit priorities in order to garner 9. Availability of a sound body of evidence, preferably
research evidence
support from leaders within the organization and
 CHAPTER 20 Developing an Evidence-Informed Practice 487

TABLE 20-1
TOOL TO USE IN SELECTING A TOPIC* FOR EVIDENCE-INFORMED PRACTICE
RATING ITEM TOPIC A TOPIC B TOPIC C

Priority for nursing (1 = low; 5 = high) ! ! !

Priority for organization (1 = low; 5 = high) ! ! !
Magnitude of the problem (1 = small; 5 = large) ! ! !
Applicability (1 = narrow; 5 = broad) ! ! !
= low; were
Likelihood to improve quality of care (1 Rights 5 = high) !
not granted to include this table ! !
Likelihood to decrease length of stay/contain costs (1 in =electronic
low; 5 = media.
high) ! ! !
Please
Likelihood to improve satisfaction (1 = low; 5 = high)refer to the printed publication. ! ! !
Body of science (1 = little; 5 = multiple studies) ! ! !
Total ! ! !
Modified from Titler, M. G. (2002). Toolkit for promoting evidence-based practice. Iowa City: University of Iowa Hospitals and Clinics, Department of
Nursing Services and Patient Care.
*Each topic should be rated with regard to the scoring criteria and on a 1-to-5 scale. The topic or topics receiving the higher scores should be
considered for selection.

stakeholders may already be members of the team.

Forming a Team Others may not be team members but are key
A team is responsible for development, imple- individuals within the organization or unit who
mentation, and evaluation of the evidence- can adversely or positively in uence the adoption
informed practice. The team or group may be an of the evidence-informed practice. Examples of
existing committee such as the quality improve- key stakeholders are nurse managers, nurse educa-
ment committee, the practice council, or the tors, researchers, nursing supervisors, chairs of
research committee. A task force approach also committees or councils that must approve system
may be used, in which a group is appointed to changes (e.g., policy/procedure revisions changes
address a specific practice issue and use research in documentation forms), and patients/families.
findings or other evidence to improve practice. uestions to consider in identification of key
The composition of the team is directed by the stakeholders include the following:
topic selected and should include interested How are decisions made in the practice
stakeholders in the delivery of care. For example, areas in which the evidence-informed prac-
a team working on evidence-informed pain man- tice will be implemented
agement should be interdisciplinary and include What types of system changes will be
pharmacists, nurses, physicians, and psycholo- needed
gists. In contrast, a team working on the evidence- Who is involved in decision making
informed practice of bathing might include a Who is likely to lead and champion
nurse expert in skin care, assistive nursing per- implementation of the evidence-informed
sonnel, and staff nurses. practice
In addition to forming a team, key stakeholders Who can in uence the decision to proceed
who can facilitate the evidence-informed practice with implementation of an evidence-
project or put up barriers against successful imple- informed practice
mentation should be identified. A stakeholder is a What type of cooperation is needed from
key individual or group of individuals who are which stakeholders to be successful
directly or indirectly affected by the implementa- Failure to involve or keep supportive stakehold-
tion of the evidence-informed practice. Some ers informed may place the success of the
488 PART SIX Application of Research: Evidence-Informed Practice

evidence-informed practice project at risk because An important early task for the evidence-
they are unable to anticipate or defend the ratio- informed practice team is to formulate the
nale for changing practice, particularly with resis- evidence-informed practice question. This helps
tant (nonsupportive) stakeholders who have a set boundaries around the project and assists in
great deal of in uence among their peer group. retrieval of the evidence. A clearly defined ques-
se Figure 20-4 to think about the status of key tion should specify the types of people/patients,
stakeholders and to strategize about interventions interventions or exposures, outcomes, and rele-
to engage various types of stakeholders for your vant study designs (Higgins Green, 2011). For
evidence-informed practice project. types of people, the team should specify the

Rights were not granted to include this figure

in electronic media.
Please refer to the printed publication.

FIGURE 20-4 Characteristics of stakeholders (resistors and facilitators).

Redrawn from Titler, M. G. (2002). Toolkit for promoting evidence-based practice. Iowa City:
University of Iowa Hospitals and Clinics, Department of Nursing Services and Patient Care.
 CHAPTER 20 Developing an Evidence-Informed Practice 489

diseases or conditions of interest, the patient pop- TABLE 20-2

ulation (e.g., age, gender, educational status), and PICOT: COMPONENTS OF AN ANSWERABLE,
the setting. For example, if the topic for the SEARCHABLE QUESTION
evidence-informed practice project is pain, the
Patient population/disease: The patient population or
team needs to specify the type of pain (e.g., acute, disease of interest; for example:
persistent, cancer), the age of the population (e.g., • Age
children, neonates, adults, older adults), and the • Gender
• Ethnicity
setting (e.g., inpatient, outpatient, ambulatory • With certain disorder (e.g., hepatitis)
care, home care, primary care). For intervention, Intervention or issue of interest: The intervention or
the types of interventions of interest to the project range of interventions of interest, for example:
• Therapy
and the comparison interventions (e.g., standard • Exposure to disease
care, alternative treatments) need to be specified. • Prognostic factor A
In the example of pain, the interventions of inter- • Risk behaviour (e.g., smoking)
Comparison intervention or issue of interest: What you
est might include pharmacological treatment,
want to compare the intervention or issue against,
analgesic administration methods (e.g., patient- for example:
controlled analgesia, epidural, intravenous), pain • Alternative therapy, placebo, or no intervention/
assessment, nonpharmacological treatment, and therapy
• NoRights were not granted to include this table
disease
patient/family education regarding self-care pain • Prognostic factorin electronic
B media.
management. For outcomes, the team should • Absence of risk
Please referfactor
to the(e.g., nonsmoking)
printed publication.
select outcomes of primary importance and con- Outcome: Outcome of interest, for example:
• Outcome expected from therapy (e.g., pressure
sider the type of outcome data that will be needed ulcers)
for decision making (e.g., benefits, harm, cost). • Risk of disease
Outcomes that may be interesting but of little • Accuracy of diagnosis
• Rate of occurrence of adverse outcome (e.g.,
importance to the project should be excluded.
death)
Finally, it is important to consider the types of Time: The time involved to demonstrate an outcome,
study designs that are likely to provide reliable for example:
data to answer the question, and the team must • The time it takes for the intervention to achieve
the outcome
search for the highest level of evidence available. • The time over which populations are observed for
A similar type of approach to formulating the the outcome (e.g., quality of life) to occur, given a
practice question is PICOT: patient, population, certain condition (e.g., prostate cancer)
or problem intervention/treatment comparison From Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
intervention/treatment outcomes and time practice in nursing and healthcare: A guide to best practice (2nd ed.,
p. 30, Table 2-1). New York: Wolters Kluwer.
frame (Melnyk Fineout-Overholt, 2011). This
approach is illustrated in Table 20-2.
making available evidence-informed practice
guidelines. It is important that these guidelines
Evidence Retrieval are accessed as part of the literature retrieval
Once a topic is selected, relevant research and process.
related literature must be retrieved and should In October 2007, the Agency for Healthcare
include clinical studies, meta-analyses, integra- Research and uality (AHR ) announced the
tive literature reviews, and existing evidence- third award of 5-year contracts for evidence-
informed practice guidelines. As more evidence informed practice centres (EPC-III) to 14 evidence-
is available to guide practice, professional orga- informed practice centres to continue and expand
nizations and federal agencies are developing and the work performed by the previous group of
490 PART SIX Application of Research: Evidence-Informed Practice

evidence-informed practice centres. Three Cana- The plan for dissemination of the guidelines
dian universities are among the 14: McMaster is threefold. First, the RNAO requested propos-
niversity, niversity of Alberta, and niversity als from interested and eligible health care orga-
of Ottawa. Contacts and additional information nizations to work in collaboration to plan,
about the current participating evidence-informed implement, and evaluate nursing best practice
practice centres are available at http://www.ahrq. guideline and disseminate knowledge from dem-
gov/clinic/epc/epcenters.htm. onstrated experiences with guidelines. Second,
The AHR also sponsors a National Guideline the Best Practice Champions Network was estab-
Clearinghouse in which abstracts of evidence- lished to prepare nurses to disseminate the best
informed practice guidelines are set forth on a practice guidelines in their practices throughout
Web site (http://www.guideline.gov). Other pro- Ontario. Third, the RNAO sponsored 10 demon-
fessional organizations that publish evidence- stration projects for colleges and universities to
informed practice guidelines are the American integrate best practice guidelines into nursing
Pain Society (http://www.ampainsoc.org) On- curricula.
cology Nursing Society (http://www.ons.org) The guidelines program now also offers
American Association of Critical-Care Nurses Advanced Clinical/Practice Fellowship (ACPF)
(http://www.aacn.org) Registered Nurses’ Asso- for nurses or health care organization. It is
ciation of Ontario (RNAO http://www.rnao.org) designed to provide registered nurses with a
National Institute for Health and Clinical Excel- focused self-directed learning experience to
lence (http://www.nice.org.uk) Association of develop clinical, leadership or best practices
Women’s Health, Obstetric and Neonatal Nurses guideline implementation knowledge and skills,
(http://www.awhonn.org) and the American Tho- with support from a mentor(s), the organization
racic Society (http://www.thoracic.org). Current where the registered nurse is employed, and the
best evidence from specific studies of clinical RNAO. This initiative is aimed at developing and
problems can be found in an increasing number promoting nursing knowledge and expertise, and
of electronic databases such as the Cochrane improving patient care and outcomes in Ontario
Library (http://www.thecochranelibrary.com/ (RNAO, 2011, p. 1). More than 365 ACPF fel-
view/0/index.html), the Centre for Health Evi- lowships have been funded in Ontario.
dence (http://www.cche.net), the Joanna Briggs The RNAO best practice guideline is readily
Institute (http://www.joannabriggs.edu.au), and available to nurses through an application for
the American College of Physicians (http://www. personal electronics devices (e.g., BlackBerry,
acponline.org) (Straus, Richardson, Glasziou, iPhone, and Android). Details are provided at the
Haynes, 2005). Web site at: http://www.rnao.org/Page.asp
In 1999, the RNAO initiated the Nursing Best PageID 924 ContentID 3282.
Practice Guidelines Project to develop practice Another electronic database, Evidence-
guidelines for nurses providing patient care. The Based Medicine Reviews (EBMR) from Ovid
project had published 32 completed guidelines as Technologies (http://www.ovid.com/site/catalog/
of January 2012, with several additional guide- DataBase/904.jsp), combines several electronic
lines are under development. Each best practice databases, including the Cochrane Database of
guideline is developed in phases: planning, devel- Systematic Reviews, Cochrane Database of
opment, implementation, evaluation, and dis- Methodology Reviews (CDMR), and MEDLINE,
semination. The topics covered by these guidelines plus links to more than 200 full-text journals.
range from smoking cessation to screening for EBMR links these databases to one another if a
delirium, dementia, and depression in older study on a topic of interest is found on MEDLINE
adults. and also has been included in a systematic review
 CHAPTER 20 Developing an Evidence-Informed Practice 491

in the Cochrane Library, the review also can be methodology/index.html). The Grading of Rec-
readily and easily accessed (Straus et al., 2005). ommendations Assessment, Development, and
In using these sources, it is important to iden- Evaluation (GRADE) Working Group, initiated
tify key search terms and to use the expertise of in 2000, is an informal collaboration of individ-
health science librarians in locating publications uals interested in addressing grading schema
relevant to the project. Additional information in health care (http://www.gradeworkinggroup.
about locating the evidence is in Chapter 5. org). In setting forth practice recommendations,
Once the literature is located, it is helpful to the GRADE system first rates the quality of the
classify the articles as clinical (nonresearch), inte- evidence as high, moderate, low, or very low and
grative research reviews, theory articles, research then grades the strength of the evidence as strong
articles, synthesis reports, meta-analyses, and or weak (GRADE Working Group, 2004 Guyatt,
evidence-informed practice guidelines. Before Oxman, Kunz, Jaeschke, Helfand, Liberati,
you read and critique the research, it is useful to Schunemann, 2008a Guyatt, Oxman, Kunz, Vist,
read theoretical and clinical articles to have a Falck-Ytter, Alonso-Cuello, 2008b Table
broad view of the nature of the topic and related 20-3). Their methods are available on their Web
concepts and to then review existing evidence- site with grading software (GRADEpro) avail-
informed practice guidelines. It is helpful to read able. The National Guidelines Clearinghouse
articles in the following order: classifies submitted guidelines according to
1. Clinical articles, to understand the state of methods used by developers to accomplish two
the practice goals: (1) to assess the quality and strength of the
2. Theory articles, to understand the various evidence through expert consensus (committee
theoretical perspectives and concepts that or expert panel method), through subjective
may be encountered when you critique review, through weighting according to a rating
studies scheme provided by the developers, or through
3. Systematic review articles and synthesis weighting according to a rating scheme not pro-
reports, to understand the state of the vided by the developers and (2) to formulate
science recommendations through various types of expert
4. Evidence-informed practice guidelines and consensus (e.g., expert manual method nominal
evidence reports group technique, consensus development confer-
5. Research articles, including meta-analyses ence) and balance sheets.
The RNAO (2011) guidelines for best prac-
Schemas for Grading the Evidence tices are based on scientific evidence after a thor-
There is no consensus among professional orga- ough review of the literature. Each of the studies
nizations or across health care disciplines regard- is rated to determine whether it should be included
ing the best system to use for denoting the type in the guideline. The rating system used for the
and quality of evidence or for grading schemas to level of evidence and the grades of recommenda-
denote the strength of the body of evidence tion are illustrated in Table 20-3.
(Atkins, Briss, Eccles, Flottorp, Guyatt, Harbour, Before critiquing research articles, reading
The GRADE Working Group, 2005 Guyatt, relevant literature, and reviewing evidence-
Oxman, Vist, Kunz, Falck-Ytter, Alonso-Coello, informed practice guidelines, an organization or
Schunemann, 2008c). For example, the Scot- group responsible for the review must agree on
tish Intercollegiate Guidelines Network has methods for noting the type of research, rating
an extensive method detailed on their Web site the quality of individual articles, and grading
for appraising research and setting forth guide- the strength of the body of evidence. sers must
line recommendations (http://www.sign.ac.uk/ evaluate which systems are most appropriate
492 PART SIX Application of Research: Evidence-Informed Practice

TABLE 20-3
EXAMPLES OF EVIDENCE-INFORMED PRACTICE RATING SYSTEMS
GRADE WORKING GROUP (GRADE
WORKING GROUP, 2004; GUYATT ET AL., REGISTERED NURSES ASSOCIATION HARRIS ET AL., 2001 (U.S. PREVENTIVE
2008A) OF ONTARIO (2011) SERVICES TASK FORCE, 2008)

STRENGTH OF EVIDENCE/QUALITY OF LEVELS OF CERTAINTY REGARDING NET

THE EVIDENCE LEVELS OF EVIDENCE BENEFIT
High: Further research is very unlikely to Ia: Evidence obtained from High: The available evidence usually
change our confidence in the estimate meta-analysis or systematic includes consistent results from
of effect. Scientific evidence provided review of randomized well-designed, well-conducted studies in
by well-designed, well-conducted, controlled trials. representative primary care populations.
controlled trials (randomized and Ib: Evidence obtained from at These studies assess the effects of the
nonrandomized) with statistically least one randomized preventive service on health outcomes.
significant results that consistently controlled trial. This conclusion is therefore unlikely to
support the guideline IIa: Evidence obtained from at be strongly affected by the results of
recommendation. least one well-designed future studies.
Moderate: Further research is likely to controlled study without Moderate: The available evidence is
have an important impact on our randomization. sufficient to determine the effects of the
confidence in the estimate of effect IIb: Evidence obtained from at preventive service on health outcomes,
and may change the estimate. least one other type of but confidence in the estimate is
Low: Further research is very likely to well-designed constrained by such factors as the
have an important impact on our quasiexperimental study. following:
confidence in the estimate of effect III: Evidence obtained from • The number, size, or quality of
and is likely to change the estimate. well-designed individual studies
Very Low: Any estimate of effect is very nonexperimental • Inconsistency of findings across
uncertain. descriptive studies, such as individual studies
Note: The type of evidence is first ranked comparative studies, • Limited generalizability of findings to
as follows: correlation studies and case routine primary care practice.
Randomized trial = high. studies. • Lack of coherence in the chain of
Observational study = low. IV: Evidence obtained from evidence.
Any other evidence = very low. expert committee reports As more information becomes available,
Limitations in study quality, important or opinions and/or clinical the magnitude or direction of the
inconsistency of results, uncertainty experiences of respected observed effect could change, and this
about the directness of the evidence, authorities. change may be large enough to alter the
imprecise or sparse data, and high conclusion.
probability of reporting bias can lower Low: The available evidence is insufficient
the grade of evidence. Expert opinion to assess effects on health outcomes.
that supports the guideline Evidence is insufficient because of one or
recommendation because the more of the following:
available scientific evidence did not • The limited number or size of studies.
present consistent results or because • Important flaws in study design or
controlled trials were lacking. Grade methods.
of evidence can be increased if there • Inconsistency of findings across
is (1) strong evidence of association— individual studies.
significant relative risk of >2 (<0.5) • Gaps in the chain of evidence.
based on consistent evidence from • Findings not generalizable to routine
two or more observational studies, primary care practice.
with no plausible confounders (1); • Lack of information on important
(2) very strong evidence of association health outcomes.
—significant relative risk of >5 (<0.2) More information may allow estimation of
based on direct evidence with no effects on health outcomes.
major threats to validity (2);
(3) evidence of a dose response
gradient (1); and (4) all plausible
confounders would have reduced the
effect (1).
 CHAPTER 20 Developing an Evidence-Informed Practice 493

TABLE 20-3
EXAMPLES OF EVIDENCE-INFORMED PRACTICE RATING SYSTEMS—cont’d
GRADE WORKING GROUP (GRADE
WORKING GROUP, 2004; GUYATT ET AL., REGISTERED NURSES ASSOCIATION HARRIS ET AL., 2001 (U.S. PREVENTIVE
2008A) OF ONTARIO (2011) SERVICES TASK FORCE, 2008)
GRADES OF
STRENGTH OF RECOMMENDATIONS RECOMMENDATION GRADES OF RECOMMENDATION
Strong: Confident that the desirable A: There is good evidence to A: The USPSTF recommends the service.
effects of adherence to a recommend the clinical There is high certainty that the net
recommendation outweigh the preventive action. benefit is substantial. Practice: Offer or
undesirable effects. B: There is fair evidence to provide this service.
Weak: The desirable effects of adherence recommend the clinical B: The USPSTF recommends the service.
to a recommendation probably preventive action. There is high certainty that the net
outweigh the undesirable effects, but C: The existing evidence is benefit is moderate or there is
the developers are less confident. conflicting and does not moderate certainty that the net benefit
Note: Strength of recommendation is allow making a is moderate to substantial. Practice:
determined by the balance between recommendation for or Offer or provide this service.
desirable and undesirable against use of the clinical C: The USPSTF recommends against
consequences of alternative preventive action; however, routinely providing the service. There
management strategies, quality of other factors may influence may be considerations that support
evidence, variability in values and decision making. providing the service in an individual
preferences, and resource use. D: There is fair evidence to patient. There is at least moderate
recommend against the certainty that the net benefit is small.
clinical preventive action. Practice: Offer or provide this service
E: There is good evidence to only if other considerations support the
recommend against the offering or providing the service in an
clinical preventive action. individual patient.
I: There is insufficient D: The USPSTF recommends against the
evidence (in quantity and/ service. There is moderate or high
or quality) to make certainty that the service has no net
recommendations, however benefit or that the harms outweigh the
other factors may influence benefits. Practice: Discourage the use of
decision-making. this service.
I: The USPSTF concludes that the current
evidence is insufficient to assess the
balance of benefits and harms of the
service. Evidence is lacking, of poor
quality, or conflicting, and the balance
of benefits and harms cannot be
determined. Practice: Read the clinical
considerations section of USPSTF
Recommendation Statement. If the
service is offered, patients should
understand the uncertainty about the
balance of benefits and harms.
From Registered Nurses Association of Ontario (2011) rating system described by Canadian Task Force on Preventive Health Care. (CTFPHC). (1997).
Quick tables by strength of evidence. Available at http://www.canadiantaskforce.ca.
USPSTF, U.S. Preventive Services Task Force.
494 PART SIX Application of Research: Evidence-Informed Practice

for the task being undertaken, the length of time and refinements may be needed. Although
to complete each instrument, and its ease of use information in well-developed, national, evidence-
(West, King, Carey, Lohr, McKoy, Sutton, Lux, informed practice guidelines is a helpful reference,
2002). It is also important to decide how the it is usually necessary to localize the guideline
strength of the evidence will be re ected in the through the use of institution-specific, evidence-
guideline. informed policies, procedures, or standards before
the guideline is applied within a specific setting. A
Critique of Evidence-Informed useful tool for critiquing clinical practice guide-
Practice Guidelines lines is the AGREE II tool (available at http://
As the number of evidence-informed practice www.agreetrust.org/).
guidelines proliferate, it becomes increasingly
important that nurses critique these guidelines Critique of Research
with regard to the methods used for formulating Critique of each study should involve the same
them and consider how they might be used in methodology, and the critique process should be
their practice. Critical areas that should be a shared responsibility. It is helpful, however, to
assessed when evidence-informed practice guide- have one individual provide leadership for the
lines are critiqued include (1) date of publication project and design strategies for completing cri-
or release (2) authors of the guideline (3) tiques. A group approach to critiques is recom-
endorsement of the guideline (4) a clear purpose mended because it distributes the workload, helps
of what the guideline covers and patient groups those responsible for implementing the changes
for which it was designed (5) types of evidence to understand the scientific base for the change in
(research, nonresearch) used in formulating the practice, arms nurses with citations and research-
guideline (6) types of research included in for- based sound bites to use in effecting practice
mulating the guideline (e.g., We considered only changes with peers and other disciplines, and pro-
randomized and other prospective controlled vides novices an environment to learn how to
trials in determining efficacy of therapeutic inter- critique and apply research findings. Methods to
ventions ) (7) a description of the methods used make the critique process fun and interesting
in grading the evidence (8) search terms and include the following:
retrieval methods used to acquire research and sing a journal club to discuss critiques
nonresearch evidence used in the guideline (9) performed by each member of the group
well-referenced statements regarding practice Pairing a novice and expert to do critiques
(10) comprehensive reference list (11) review of Eliciting assistance from students who may
the guideline by experts and (12) whether the be interested in the topic and want experi-
guideline has been used or tested in practice and, ence performing critiques
if so, with what types of patients and in what Making a class project of critique and syn-
types of settings. thesis of research for a given topic
Evidence in or ed ractice guidelines are Several resources are available to assist with
principles that help the researcher better under- the critique process, including the Evidence-
stand the evidence base of certain practices, Such Based Medicine and accompanying compact disc
guidelines, formulated through the use of rigorous (Straus et al., 2005) and Evidence-Based Nursing:
methods, provide a useful starting point for nurses A Guide to Clinical Practice (DiCenso, Guyatt,
to understand the evidence base of certain prac- Ciliska, 2005). If you wish to start your own
tices. However, more research may have become journal club, refer to Silversides (2011) for practi-
available since the publication of the guideline, cal advice and further references.
 CHAPTER 20 Developing an Evidence-Informed Practice 495

and strength of evidence used to support the prac-

Helpful Hint tice need to be clearly delineated. Box 20-2 is a
Keep critique processes simple, and encourage useful tool to assist with this activity.
participation by staff members who are providing direct
patient care. The following are examples of practice recom-
mendation statements:
Small, informal group health education
Synthesis of the Research classes, delivered in the antenatal period,
Once studies are critiqued, a decision is made have a better impact on breastfeeding initia-
regarding use of each study in the synthesis of the tion rates than breastfeeding literature alone
evidence for application in clinical practice. or combined with formal, noninteractive
Factors that should be considered for inclusion of methods of teaching. (Strength of recom-
studies in the synthesis of findings are overall mendation = B RNAO, 2003, p. 46)
scientific merit of the study type (e.g., age, Older people who have recurrent falls
gender, pathological condition) of participants should be offered long-term exercise and
enrolled in the study and their similarity to the balance training. (Strength of recommenda-
patient population to which the findings will be tion = B American Geriatrics Society,
applied and relevance of the study to the topic of British Geriatrics Society, American
question. For example, if the practice area is pre- Academy of Orthopaedic Surgeons,
vention of deep venous thrombosis in patients Panel on Falls Prevention, 2001).
after surgery, a descriptive study with a heteroge- Every patient should be screened to identify
neous population of medical patients is not appro- those most likely to be affected by asthma.
priate for inclusion in the synthesis of findings. As part of the basic respiratory assessment,
To synthesize the findings from research cri- nurses should ask every patient two
tiques, it is helpful to use a summary table (Table questions:
20-4) in which critical information from studies 1. Have you ever been told by a physician
can be documented. Essential information to that you have asthma
include in such summary is the following: 2. Have you ever used a puffer or inhaler
Study purpose or asthma medication for breathing prob-
Research questions/hypotheses lems (Strength of recommendation = D
The variables studied RNAO, 2004, p. 30).
A description of the study sample and sett- Apply dressings that maintain a moist wound
ing environment. Examples of moist dressings
The type of research design include, but are not limited to, hydrogels,
The methods used to measure each vari- hydrocolloids, saline moistened gauze, and
able transparent film dressings. The ulcer bed
Detailed description of the independent should be kept continuously moist. (Evi-
variable/intervention tested dence Grade = B Colwell, Foreman,
The study findings Trotter, 1992 Folkedahl, Frantz, Goode,
2002 Fowler Goupil, 1984 Gorse
Setting Forth Evidence-Informed Messner, 1987 Kurzuk-Howard, Simpson,
Practice Recommendations Palmieri, 1985 Neill, Conforti, Kedas,
On the basis of the critique of evidence-informed 1989 Oleske, Smith, White, Pottage,
practice guidelines and synthesis of research, rec- Donovan, 1986 Saydak, 1990 Sebern,
ommendations for practice are set forth. The type 1986 akellis Chrischilles, 1992).
 496

TABLE 20-4
EXAMPLE OF A SUMMARY TABLE FOR RESEARCH CRITIQUES
PURPOSE INDEPENDENT DEPENDENT SUMMARY
AND VARIABLES VARIABLES OVERALL STATEMENTS
RESEARCH RESEARCH AND AND STATISTICAL GENERAL GENERAL QUALITY FOR
CITATION QUESTION DESIGN SAMPLE MEASURES MEASURES TESTS RESULTS IMPLICATIONS STRENGTHS WEAKNESSES OF STUDY* PRACTICE
PART SIX Application of Research: Evidence-Informed Practice

*Use a consistent rating system (e.g., good, fair, poor).

 CHAPTER 20 Developing an Evidence-Informed Practice 497

BOX 20-2 Relevance of evidence for practice

CONSISTENCY OF EVIDENCE FROM CRITIQUED Consistency in findings across studies,
RESEARCH AND APPRAISALS OF EVIDENCE- guidelines, or both
INFORMED PRACTICE GUIDELINES A significant number of studies, evidence-
1. Are studies replicated with consistent results? informed practice guidelines, or both in
2. Are the studies well designed? which sample characteristics are similar to
3. Are recommendations consistent among systematic those to which the findings will be applied
reviews, evidence-informed practice guidelines, and
critiqued research? Consistency among evidence from research
4. Are risks to the patient identified from evidence- and other nonresearch evidence
informed practice recommendations? Feasibility for use in practice
5. Are benefits to the patient identified?
6. Have cost analysis studies been conducted with The risk/benefit ratio (risk of harm poten-
regard to the recommended action, intervention, or tial benefit for the patient)
treatment? It is recommended that practice changes be based
7. Are summary recommendations about
on knowledge/evidence derived from several
assessments, actions, and interventions or
treatments available from the research, systematic sources (e.g., several research studies) that dem-
reviews, and evidence-informed guidelines with an onstrate consistent findings.
assigned evidence grade? Synthesis of study findings and other evidence
8. Is one of the following examples of grading the
evidence used? may result in supporting current practice, making
a. Evidence from well-designed meta-analysis or minor practice modifications, undertaking major
other systematic reviews practice changes, or developing a new area of
b. Evidence from well-designed controlled trials,
both randomized and nonrandomized, with practice. For example, a project on gauze versus
results that consistently support a specific action transparent dressings did not result in a practice
(e.g., assessment), intervention, or treatment change because the results of the studies reviewed
c. Evidence from observational studies (e.g.,
supported current practice (Pettit Kraus, 1995).
correlational descriptive studies) or controlled
trials with inconsistent results In comparison, Madsen and colleagues (2005)
d. Evidence from expert opinion or multiple cases used a combination of research findings and
expert consultation to derive a guideline that
Modified from Titler, M. G. (2002). Toolkit for promoting resulted in a change in practice for assessing
evidence-based practice. Iowa City: University of Iowa Hospitals bowel motility after abdominal surgery in an
and Clinics, Department of Nursing Services and Patient Care.
adult inpatient population. Madsen and col-
leagues’ project resulted in (1) omitting bowel
Helpful Hint sound assessment as a marker of return of gastro-
Use of a summary form helps identify common- intestinal motility and (2) using return of atus,
alities across several studies with regard to study find- first bowel movement, and absence of abdominal
ings and the types of patients to which study findings distention as primary indicators of return of bowel
can be applied. It also helps in synthesizing the overall
strengths and weakness of the studies as a group. motility after abdominal surgery in adults.

Development of Evidence-
Decision to Change Practice Informed Practice
After studies are critiqued and synthesized and The next step is to put in writing the evidence
evidence-informed practices are set forth, the base of the practice (Haber, Feldman, Penney,
next step is to decide whether findings are appro- Carter, Bidwell-Cerone, Rose Hott, 1994) the
priate for use in practice. The following criteria grading schema that has been agreed upon should
should be considered in making these decisions: be used. When results of the critique and
498 PART SIX Application of Research: Evidence-Informed Practice

synthesis of evidence support current practice or BOX 20-3

suggest a change in practice, a written evidence- KEY QUESTIONS FOR FOCUS GROUPS
informed practice standard (e.g., policy, proce-
1. What is needed by (nurses, physicians) to use the
dure, guideline) is warranted. This is necessary so evidence-informed practice with patients in units
that professionals in the organization (1) know (specify unit)?
that the practices are based on evidence and (2) 2. In your opinion, how will this standard improve
patient care in your unit/practice?
know which type of evidence (e.g., randomized 3. What modifications would you suggest in the
controlled trial, expert opinion) was used in evidence-informed practice standard before you
developing the evidence-informed standard. would use it in your practice?
4. What content in the evidence-informed practice
Several different formats can be used to docu-
standard is unclear? Needs revision?
ment evidence-informed practice changes. The 5. What would you change about the format of the
format chosen is in uenced by what the docu- evidence-informed practice standard?
ment is and how it will be used. Written evidence- 6. What part of this evidence-informed practice
change do you view as most challenging?
informed practices should be part of the 7. Any other suggestions?
organizational policy and procedure manual and
should include linkages to the references for the
parts of the policy and procedure that are based
on research and other types of evidence. support, and system changes. Rogers’s (2003)
Clinicians (e.g., nurses, physicians, pharma- seminal work on diffusion of innovations is
cists) who adopt evidence-informed practices are extremely useful for selecting strategies for pro-
in uenced by the perceived participation they moting adoption of evidence-informed practices.
have had in developing and reviewing the proto- Other investigators describing barriers to and
col (Titler, 2008). It is imperative that once the strategies for adoption of evidence-informed
evidence-informed practice standard is written, practices have used Rogers’s (2003) model
key stakeholders have an opportunity to review it (Gravel et al., 2006 Scott et al., 2008 Thomp-
and provide feedback to the person or persons son, Estabrooks, Scott-Findlay, Moore, Wallin,
responsible for writing it. Focus groups can 2007).
provide discussion about the evidence-informed According to this model, adoption of innova-
standard and identify key areas that may be tions, such as evidence-informed practices, is
potentially troublesome during the implementa- in uenced by the nature of the innovation (e.g.,
tion phase. Key questions that can be used in the the type and strength of evidence the clinical
focus groups are listed in Box 20-3. topic) and the manner in which it is communi-
cated (disseminated) to members (nurses) of a
Helpful Hint social system (organization, nursing profession
Use a consistent approach to writing evidence- Rogers, 2003 Titler Everett, 2001). Strategies
informed practice standards and referencing the research for promoting adoption of evidence-informed
and related literature.
practices must address these areas within a
context of participative, planned change
Implementing the Practice Change (Figure 20-5).
If a practice change is warranted, the next steps
are to make the evidence-informed changes in Nature of the Innovation/Evidence-
practice. This step goes beyond writing a policy Informed Practice
or procedure that is evidence informed it requires Characteristics of an innovation or evidence-
interaction among direct care providers to cham- informed practice topic that affect adoption
pion and foster evidence adoption, leadership include the relative advantage of the evidence-
 CHAPTER 20 Developing an Evidence-Informed Practice 499

Social
system

Communication
Characteristics of Communication Extent of
the EBP process adoption

Users of the
EBP

FIGURE 20-5 Implementation model.

Redrawn from Rogers, E. M. (1995). Diffusion of innovations. New York: Free Press; and from
Titler, M. G., & Everett, L. Q. (2001). Translating research into practice: Considerations for critical
care investigators. Critical Care Nursing Clinics of North America, 13(4), 587–604.

informed practice (e.g., effectiveness, relevance determines the rate and extent of adoption (Green-
to the task, social prestige) the compatibility halgh et al., 2004 Rogers, 2003).
with values, norms, work, and perceived needs of Studies suggest that clinical systems, com-
users and complexity of the evidence-informed puterized decision support, and prompts/quick
practice topic (Rogers, 2003). For example, reference guides that support practice (e.g.,
evidence-informed practice topics that are per- decision-making algorithms paper reminders)
ceived by users as relatively simple (e.g., in u- have a positive effect on aligning practices with
enza vaccines for older adults) are more easily the evidence base (Doebbeling et al., 2006
adopted in less time than those that are more Shojania Grimshaw, 2005 Titler, 2006).
complex (e.g., acute pain management for hospi- Computerized knowledge management has
talized older adults). consistently demonstrated significant improve-
Strategies to promote adoption of evidence- ments in provider performance and patient out-
informed practices related to characteristics of the comes (Wensing, Wollersheim, Grol, 2006).
topic include practitioner review and reinven- Feldman and associates (2005), using a just-in-
tion of the evidence-informed practice guideline time e-mail reminder in home health care, dem-
to fit the local context, use of quick reference onstrated (1) improvements in evidence-informed
guides and decision aids, and use of clinical care and outcomes for patients with heart failure
reminders (Doebbeling, Chou, Tierney, and (2) reduced pain intensity for cancer patients
2006). (McDonald, Pezzin, Feldman, Murtaugh, Peng,
An important principle to remember for plan- 2005). There is still much to learn about the
ning implementation of an evidence-informed best manner of deploying evidence-informed
practice is that the attributes of the evidence- information through electronic clinical informa-
informed practice topic as perceived by users and tion systems to support evidence-informed care.
stakeholders (e.g., ease of use, valued part of An example of a quick reference guide is shown
practice) are neither stable features nor sure in Figure 20-6.
determinants of their adoption. Rather, it is the
interaction among the characteristics of the Methods of Communication
evidence-informed practice topic, the intended Interpersonal communication channels, methods
users, and a particular context of practice that of communication, and in uence among social
Use this quick reference guide to help in the assessment of pain:
• Before clients undergo medical procedures or surgeries that can cause pain
• When clients are experiencing pain from recent surgeries, medical procedures, trauma, or other acute illness

General principles for assessing pain in older adults:

• Verify sensory ability (Can the person see you? Hear you?).
• Allow time to respond.
• Repeat questions/instructions as necessary.
• Use printed materials with large type and dark lines.

Yes No
Ask client if in
pain, discomfort,
Rate pain intensity: or aching. Validate absence of pain
Choose a scale that works best
using observable signs
with this client and continue to Uninterpretable Or not alert
use the same scale. If alert and response
oriented, use 0–10 scale. If
unsuccessful, try Verbal Look for observable signs Observation
Descriptors or Faces. agrees with
Look for possible causes of pain client’s
response?
No
No Yes
Is the client If available, get additional
able to self- information from family, significant Reflect back Reflect back
report pain? others: usual pain behaviours, to clarify to confirm
history of pain. incongruence, “no pain,”
e.g., say “But e.g., say
Yes you look “So, right now,
uncomfortable you are not
What is the to me.” having any
level of pain? discomfort?”

Severe Moderate Mild No pain

7–10 on a 0–10 scale 4–6 on a 0–10 scale or 1–3 on a 0–10 scale or
or client appears to client appears to be client does not appear
be in significant distress in significant distress to be in distress

Do a rapid assessment: Do a rapid assessment: Do a complete Document absence of pain.

location, quality, duration location, quality, duration pain assessment Tell client to let nurse
know if pain develops.

Document pain Go to treatment. See Establish regular

intensity in flow Quick Reference Guide: assessment and
chart, noting the “Management of Acute Pain documentation
scale you used in Older Adults” of pain

FIGURE 20-6 Quick reference guide: assessment of acute pain in older adults.
Redrawn from Harris, R. P., Helfan, M., Woolf, S. H., Lohr, K. N., Mulrow, C. D., Teutsch, S. M.,
& Atkins, D. (2001). Current methods of the U.S. Preventive Services Task Force: A review of
the process. American Journal of Preventive Medicine 20(3S), 21–35; and from Herr, K., Titler,
M., Sorofman, B., Ardery, G., Schmitt, M., & Young, D. (2000). Evidence-based guideline: Acute
pain management in the elderly. In From book to bedside: Acute pain Management in the elderly
[Grant No. 1 R01 HS10482-01]. Iowa City: University of Iowa.
 CHAPTER 20 Developing an Evidence-Informed Practice 501

networks of users affect adoption of evidence- and competency testing are helpful in education
informed practices (Rogers, 2003). se of mass of staff.
media and of consultation with opinion leaders, Several studies have demonstrated that opinion
change champions, and experts, along with edu- leaders are effective in changing behaviours of
cation, are among strategies tested to promote use health care practitioners (Cullen, 2005 Green-
of evidence-informed practices. Interactive edu- halgh et al., 2004 Irwin Ozer, 2004), espe-
cation, used in combination with other practice- cially in combination with educational outreach
reinforcing strategies, has more positive effects or performance feedback. O inion leaders are
on improving evidence-informed practice than from the local peer group, viewed as a respected
does education alone (Irwin Ozer, 2004 Loeb, source of in uence, considered by associates to
Brazil, McGeer, Stevenson, Walter, Lohfeld, be technically competent, and trusted to judge the
outman, 2004). fit between the innovation and the local situation.
It is important that staff know the scientific They have a wide sphere of in uence across
basis for the changes in practice and the improve- several microsystems/units and use the innova-
ments in quality of care that are anticipated from tion, in uence peers, and alter group norms
the change. This information must be dissemi- (Rogers, 2003). The key characteristic of an
nated to staff creatively through various educa- opinion leader is that he or she is trusted to evalu-
tional strategies. A staff in-service session may ate new information in the context of group
not be the most effective method, nor might it norms. To do this, an opinion leader must be
reach the majority of the staff. Although it is considered by associates not only to be techni-
unrealistic for all staff to have participated in the cally competent but also a full and dedicated
critique process or to have read all studies used member of the local group (Rogers, 2003). Social
to develop the evidence-informed practice, it is interactions such as hallway chats, one-on-one
important that they know the myths and realities discussions, and addressing questions are impor-
of the practice. Education of staff also must tant yet often overlooked components of transla-
include ensuring that they are competent in the tion (Rogers, 2003). Thus, discussions about the
skills necessary to carry out the new practice. For evidence-informed practices between local
example, if a pain assessment tool is being imple- opinion leaders and members of their peer group
mented to assess pain in cognitively impaired is necessary for translating research into practice.
older patients in the long-term care setting, it is If the evidence-informed practice that is being
essential that caregivers have the knowledge and implemented is interdisciplinary in nature,
skill to use the tool in their practice setting. discipline-specific opinion leaders should be used
One method of communicating information to to promote the change in practice (Rogers, 2003).
staff is through use of colourful posters that iden- Role expectations of an opinion leader are in
tify myths and realities or describe the essence of Box 20-4.
the change in practice. Visibly identifying person- Change champions are also helpful for imple-
nel who have learned the information and are menting innovations (Rogers, 2003 Titler et al.,
using the evidence-informed practice (e.g., 2006). They are practitioners within the local
buttons, ribbons, pins) stimulates interest in group setting (e.g., clinic, patient care unit) who
others who may not have internalized the change. are expert clinicians, are passionate about the
As a result, the new learner may begin asking innovation, are committed to improving quality
questions about the practice and be more open to of care, and have a positive working relationship
learning. Other educational strategies such as with other health professionals (Rogers, 2003).
train-the-trainer programs, Webinars, podcasts, They circulate information, encourage peers to
502 PART SIX Application of Research: Evidence-Informed Practice

BOX 20-4 in their unit/microsystem (Nelson, Batalden,

ROLE EXPECTATIONS OF AN OPINION LEADER Huber, Moher, Godfrey, Headrick, Wasson,
2002). Core group members represent various
1. Be or become an expert in the evidence-informed
practice. shifts and days of the week and become knowl-
2. Provide organizational or unit leadership for edgeable about the scientific basis for the prac-
adopting the evidence-informed practice. tice the change champion educates and assists
3. Implement various strategies to educate peers
about the evidence-informed practice. them in using practices that are aligned with the
4. Work with peers, other disciplines, and leadership evidence. Each member of the core group, in turn,
staff to incorporate key information about the takes the responsibility for imparting evidence-
evidence-informed practice into organizational/unit
informed information and effecting practice
standards, policies, procedures, and documentation
systems. change with several (e.g., two to three) of their
5. Promote initial and ongoing use of the evidence- peers. They assist the change champion and
informed practice by peers. opinion leader with disseminating the evidence-
Modified from Titler, M. G., Herr, K., Everett, L. Q., Marsh, J. L., informed information to other staff, reinforce the
Schilling, M., Brooks, J., et al. (2006). Book to bedside: Promoting practice change on a daily basis, and provide
and sustaining EBPs in elders (Final Progress Report to AHRQ, Grant
No. 2R01 HS010482-04). Iowa City: University of Iowa College of positive feedback to staff members who align
Nursing.
their practice with the evidence base (Titler,
2006). sing a core group approach in conjunc-
tion with a change champion results in a critical
mass of practitioners promoting adoption of the
adopt the innovation, arrange demonstrations, evidence-informed practice (Rogers, 2003).
and orient staff to the innovation. The change Clinical nurse specialists can provide one-on-
champion believes in an idea is undaunted by one consultation to staff regarding use of the
insults and rebuffs and, above all, is persistent. evidence-informed practice with specific patients,
Because nurses prefer interpersonal contact assist staff in troubleshooting issues in applica-
and communication with colleagues rather than tion of the practice, and provide feedback on
Internet or traditional sources of practice know- provider performance regarding use of the
ledge (Estabrooks, Chong, Brigidear, Profetto- evidence-informed practices. Studies have dem-
McGrath, 2005), it is imperative that one or two onstrated that when clinical nurse specialists act
change champions be identified for each patient as facilitators of change, adherence to the
care unit or clinic where the change is being made evidence-informed practice is promoted (Stetler,
so that evidence-informed practices can be Legro, Rycroft-Malone, Bowman, Curran,
enacted by direct care providers (Titler et al., Guihan, Wallace, 2006 Titler et al., 2009).
2006). Staff nurses are some of the best change
agents for evidence-informed practice. Confer- Evaluation
encing with opinion leaders and change champi- Evaluation—the process of determining the
ons periodically during implementation is helpful value of data—provides an opportunity to collect
in addressing questions and providing guidance and analyze data with regard to use of a new
as needed (Titler et al., 2006). evidence-informed practice and then to modify
sing a core group in conjunction with change the practice as necessary. It is important that the
champions is also helpful for implementing the evidence-informed change be evaluated, both in
practice change. A core group is a select group of the area where it is piloted and when the practice
practitioners with the mutual goal of disseminat- is changed in additional patient care areas. The
ing information regarding a practice change and importance of the evaluation cannot be overem-
facilitating the change in practice by other staff phasized it provides information for performance
 CHAPTER 20 Developing an Evidence-Informed Practice 503

gap assessment, audit, and feedback, and it practice in care delivery and whether they are
provides information necessary to determine implementing the practice as noted in the written
whether the evidence-informed practice should evidence-informed practice standard. Evaluation
be retained, modified, or eliminated. of the process also should include notes about (1)
A desired outcome achieved in a more con- barriers that staff encounter in carrying out the
trolled environment, when a researcher is imple- practice (e.g., lack of information, skills, or nec-
menting a study protocol for a homogeneous essary equipment), (2) differences in opinions
group of patients (conduct of research), may not among health care professionals, and (3) diffi-
be achieved when the practice is implemented in culty in carrying out the steps of the practice as
the natural clinical setting, by several caregivers, originally designed (e.g., shutting off tube feed-
or to a more heterogeneous patient population. ings 1 hour before aspirating contents for check-
Steps of the evaluation process are summarized ing placement of nasointestinal tubes). Process
in Box 20-5. data can be collected from staff reports or patient
Evaluation should include both process and self-reports, or both from medical record audits
outcome measures. The process component or from observation of clinical practice. Exam-
focuses on how the evidence-informed practice ples of process and outcome questions are shown
change is being implemented. It is important in Table 20-5.
to know whether staff members are using the Outcome data are an equally important part of
evaluation. The purpose of outcome evaluation is
BOX 20-5 to assess whether the patient, staff, or fiscal out-
STEPS OF EVALUATION FOR EVIDENCE- comes expected, or a combination of these, are
INFORMED PRACTICE achieved. Therefore, it is important that baseline
1. Identify process and outcome variables of interest. data be used for a preintervention/postintervention
Examples: comparison (Cullen, 2005). Outcome measures
Process variable: For patients older than 65 years,
a Braden scale will be completed on should be measured before the change in practice
admission. is implemented, after implementation, and every 6
Outcome variable: Presence/absence of to 12 months thereafter. Findings must be pro-
nosocomial pressure ulcer; if present,
determine stage as I, II, III, or IV.
vided to clinicians to reinforce the effects of the
2. Determine methods and frequency of data change in practice and to ensure that they are
collection. incorporated into quality improvement programs.
Example: When process and outcome data are collected
Process variable: Chart audit of all patients older
than 65 years, 1 day a month for evaluation of evidence-informed practice
Outcome variable: Assessment of all patients change, it is important that the data-collection
older than 65 years, 1 day a month tools are user-friendly, short, concise, and easy to
3. Determine baseline and follow-up sample sizes.
4. Design data-collection forms. complete and that they have content validity. The
Example: evaluation process includes planned feedback to
Process variable: chart audit abstraction form staff members who are making the change. The
Outcome variable: pressure ulcer assessment form
5. Establish content validity of data-collection forms.
feedback includes verbal or written (or both)
6. Train data-collectors. appreciation for the work and visual demonstra-
7. Assess interrater reliability of data-collectors. tion of progress in implementation and improve-
8. Collect data at specified intervals. ment in patient outcomes. The key to effective
9. Provide “on-site” feedback to staff regarding the
progress in achieving the practice change. evaluation is to ensure that the evidence-informed
10. Provide feedback of analyzed data to staff. change in practice is warranted (e.g., will improve
11. Use data to assist staff in modifying or integrating quality of care) and that the intervention does not
the evidence-informed practice change.
bring harm to patients. For example, when the
504 PART SIX Application of Research: Evidence-Informed Practice

TABLE 20-5
EXAMPLES OF EVALUATION MEASURES
NEITHER AGREE
EXAMPLE OF PROCESS QUESTIONS STRONGLY DISAGREE DISAGREE NOR DISAGREE AGREE STRONGLY AGREE

1. I feel well prepared to use the 1 2 3 4 5

Bracken Scale with older patients.
2. Malnutrition increases patient risk 1 2 3 4 5
for pressure ulcer development.
EXAMPLE OF OUTCOME QUESTION

Patient: On a scale of 0 (no pain) to _____________ (Pain intensity)

10 (worst possible pain), how
much pain have you experienced
over the past 24 hours?

practice for assessing return of bowel motility nursing departments and the health care
after abdominal surgery in adults was changed, it system
was important to inform staff that using other Recognition for and rewarding of evidence-
markers for return of bowel motility, rather than informed practice behaviours
bowel sound assessment, did not result in
increased paralytic ileus or bowel obstruction Helpful Hint
(Madsen et al., 2005). Include patient outcome measures (e.g., pressure
ulcer prevalence) and cost (e.g., cost savings, cost avoid-
CREATING A CULTURE OF EVIDENCE- ance) in evaluation.
INFORMED PRACTICE
se of research evidence to guide clinical and The first building block involves ensuring that
operational decisions is a necessity in health care the mission and vision statements of the health
delivery (Cullen Titler, 2004). Chief nurse care system and nursing services re ect a com-
executives and their leadership staff set the stage mitment to the provision of evidence-informed
and culture for evidence-informed practice in health care for example: The vision of the
their settings. How this is done varies, but essen- department of nursing services and patient care is
tial components are necessary for evidence- to be an international exemplar of using evidence
informed practices (both the process and product) to guide clinical and operational decision-
to be an integral part of the organization. making. For evidence-informed practices to be
Providing this leadership is a continuous manifested in everyday work, it is necessary to
process that involves four major building blocks incorporate into the organization’s or depart-
(Figure 20-7): ment’s strategic plan specific action statements
Incorporating evidence-informed practice that promote and foster evidence-informed prac-
terminology into the mission, vision, strate- tices. Such actions might include offering an
gic plan, and philosophy of care delivery annual evidence-informed practice staff nurse
Establishing explicit performance expecta- internship program integrating educational
tions about evidence-informed practice for content about evidence-informed practice into
staff at all levels of the organization orientation of new staff monitoring and acting on
Integrating the work of evidence-informed the results of key indicators for selected evidence-
practice into the governance structure of informed practices (e.g., acute pain management,
 CHAPTER 20 Developing an Evidence-Informed Practice 505

Leadership

Recognition
and rewards

Resources
and governance

Performance expectations

Strategic plan
Vision Mission
Philosophy of care

FIGURE 20-7 Four major building blocks in the process of creating a culture of evidence-
informed practice.

prevention of pressure ulcers, prevention of falls) evidence-informed practices. Performance ex-

and initiating two or three new evidence-informed pectations for nurse managers include creating
practices per year that are triggered by opera- a culture that fosters interdisciplinary quality
tional data, quality improvement data, or both. improvement informed by evidence. Clinical
Equally important to a mission and vision that nurse specialists’ performance expectations
embrace evidence-informed practice is clarity include leading a team, finding the evidence, and
about (1) the definition and meaning of evidence- synthesizing the evidence for practice. Clinical
informed practice (some departments actually nurse specialists assist staff with focusing their
articulate a definition), (2) the organizational clinical question about improving practice,
process or model of evidence-informed practice, finding and evaluating the research evidence, and
and (3) a philosophy of care that embraces clini- manoeuvring through the committee structures to
cal inquiry and questioning of the status quo. implement and sustain the changes in practice.
The second building block involves develop- The ability of clinical nurse specialists to meet
ing and using performance expectations regard- these performance criteria is an essential part of
ing evidence-informed practices. For example, their annual performance appraisals. Similarly,
evidence-informed practice performance expec- nurse managers set the tone, value, and work
tations for staff nurses should include critical culture for the microsystems they lead.
thinking, continual questioning of practice, par- Staff migrate to microsystems that foster pro-
ticipating in making evidence-informed practice fessional growth, professional nursing practice,
changes, serving as leaders of change in their site data-informed decision making, and innovative
of care delivery, and participating in evaluating practices all characteristics of cultures that
evidence-informed changes in practice. promote adoption of evidence-informed nursing
The chief nurse executive sets the tone for practices. Nurse managers also foster evidence-
evidence-informed practice and explicates role informed practices in their units by allocation of
expectations of other nurse leaders within the resources, which is an important element for staff
organization regarding the knowledge, skills, and nurse participation in new evidence-informed
behaviours necessary to promote adoption of practice projects. Consequently, institutions who
506 PART SIX Application of Research: Evidence-Informed Practice

hire, retain, and value (through performance care, strategic plan, and performance criteria
appraisals) nurse managers and clinical nurse incorporate language about the value and com-
specialists skilled in evidence-informed practice mitment of the organization to evidence-informed
are more likely to observe development of clini- practice.
cal innovations and adoption of evidence- The third building block involves integrating
informed practices in the multiple units and sites evidence-informed practice into the governance
of care delivery for which they are responsible. of the health care system and ensuring that
Enactment of evidence-informed practice resources are available to assist staff with this
behaviours by the chief nurse executive includes work. One question frequently asked is Where
providing resources for evidence-informed prac- should the work of evidence-informed practice
tice such as easy access to evidence-informed reside The short answer is Everywhere,
practice Web sites, retaining personnel with because evidence-informed practice saves health
expertise in evidence-informed practice, support- care dollars and improves patient outcomes
ing programs that develop a critical mass of staff (Brooks et al., 2009 Titler et al., 2009). More
nurses with expertise in evidence-informed prac- explicitly, to sustain a vision of providing
tice (e.g., an evidence-informed practice staff evidence-informed health care, the work and
nurse internship program Cullen Titler, 2004), accountability for evidence-informed practice
and providing access to assistance with analysis must be integrated into the governance structure.
of data and transforming data into information. This includes interdisciplinary collaboration
Chief nurse executives also enact the value of across departments and services, as well as
evidence-informed practice by using information coordination within discipline-specific areas of
from evaluations of existing and new clinical practice.
programs in operational decisions, and by reward- Evidence-informed changes in practice must
ing and recognizing direct care providers who be coordinated with professional policy and pro-
make evidence-informed practice a reality in their cedure committees in order for the evidence to be
daily work. sing evidence in administrative re ected in practice standards. Documentation
decisions is another behaviour modelled by chief systems, whether electronic or manual, must
nurse executives who value evidence-informed support the evidence-informed practices through
practice. reminder systems, decision-support algorithms,
One example of an evidence-informed admin- and easy-to-use documentation forms. Too often,
istrative practice involves assessing the work personnel providing direct care are expected to
culture of nurses that contribute to job satisfac- change practices without full modification of
tion and retention and then using this information, the documentation systems that capture and rein-
along with research evidence, to create adminis- force the desired changes. Although the pri-
trative interventions that decrease turnover. mary responsibility for tracking and promoting
Because it is difficult to support multiple evidence-informed practice may reside in a spe-
evidence-informed practice changes simultane- cific department or program (e.g., research, edu-
ously, chief nurse executives committed to cation, quality improvement), evidence-informed
evidence-informed practice lead discussion practice must be viewed and valued as essential
and decision making regarding priority setting work at all levels of the organization and within
for areas of evidence-informed practice (e.g., the committees/councils that govern the health
skin care, pain). Last, but most important, it care system.
is the chief nurse executive’s responsibility to The fourth building block involves recognition
ensure that the mission, vision, philosophy of for and rewarding of evidence-informed practice
 CHAPTER 20 Developing an Evidence-Informed Practice 507

behaviours. Such recognition can range from (1) commonly used terms (Graham et al., 2006). This
submitting staff projects and names to national is evidenced by differing definitions and the inter-
and international professional organizations that changing of terms that, in fact, may represent
have programs for recognition of excellence in different concepts to different people. Adding to
evidence-informed practice to (2) recognizing the confusion is that terminology may vary
specific staff members in their unit at the shift depending on the country in which the research
change for the care they provide that is informed was conducted (Adams Titler, 2010). Graham
by evidence. Other recognition activities include and colleagues (2006) reported identifying
an annual recognition day with a display of 29 terms in nine countries that refer to some
posters of the evidence-informed practice work aspect of translating research findings into prac-
occurring in each unit recognition in a weekly or tice. For example, researchers in Canada may use
monthly internal communication postings on the terms research utili ation, no ledge-to-
Web sites and broadcasting the stellar accom- action, no ledge transfer, or no ledge transla-
plishments in the local, regional, and national tion interchangeably, whereas researchers in the
media. Some organizations integrate evidence- nited States, the nited Kingdom, and Europe
informed practice expectations into the clinical may be more likely to use the term implementa-
ladder system, and others provide staff release tion or research translation to express similar
time from direct patient care to do the work concepts (Graham et al., 2006 Graham Logan,
of evidence-informed practice. Recognition 2004 Titler Everett, 2001). The goals of the
by peers, as well as senior administrators, is Canadian Institutes of Health Research (CIHR)
important. are not only to support the development of new
knowledge through research but also to ensure
TRANSLATION SCIENCE that the knowledge is translated into practice.
Translation science mentioned previously in this Knowledge translation is defined as a dynamic
chapter, is the investigation of strategies to increase and iterative process that includes synthesis, dis-
the rate and extent of adoption and sustainability of semination, exchange and ethically sound appli-
evidence-informed practice by individuals and cation of knowledge to improve the health of
organizations to improve clinical and operational Canadians, provide more effective health services
decision making (Eccles Mittman, 2006 Titler, and products and strengthen the health care
Everett, Adams, 2007). It includes research to system (CIHR, 2009).
(1) understand context variables that in uence Building this body of research knowledge
adoption of evidence-informed practices and (2) mandates development in many areas theoretical
test the effectiveness of interventions to promote developments are needed to provide frameworks
and sustain use of evidence-informed health care and predictive theories for creating generalizable
practices. Translation science denotes both the research such as how to change individual and
systematic investigation of methods, interven- organizational behaviour. Methodological devel-
tions, and variables that in uence adoption of opments are also required, as well as exploratory
evidence-informed health care practices, as well as studies aimed at understanding the experiential
the organized body of knowledge gained through and organizational learning that accompanies
such research (Eccles Mittman, 2006 Ruben- implementation. Rigorous evaluations are needed
stein Pugh, 2006 Sussman, Valente, Rohrbach, to evaluate the effectiveness and efficiency of
Skara, Pentz, 2006 Titler et al., 2007). implementation interventions (Adams Titler,
Because translation research is a young 2010). Partnerships are needed to encourage
science, there are no standardized definitions of communication among researchers, theorists, and
508 PART SIX Application of Research: Evidence-Informed Practice

implementers and to understand what types of nursing research and to justify the allocation of
knowledge are needed and how that knowledge resources and personnel to create the necessary
can best be developed (Adams Titler, 2010 infrastructures. The second challenge is to reduce the
time gap between when knowledge is developed and
Dawson, 2004 Gold Taylor, 2007 Tripp- when it is used. The third challenge is to create
Reimer Doebbeling, 2004). agency infrastructures that will support the
transformation of nursing practice from a ritual base to
FUTURE DIRECTIONS a research base. Efforts to address these challenges
se of research across health care systems for require partnerships among practitioners, researchers,
administrators, and disciplines.
improving the quality of care is essential. As pro-
fessionals continue to understand the science of Education of nurses must include knowledge
nursing and synthesize this science for applica- and skills in the use of research evidence in prac-
tion in practice, it will become increasingly nec- tice. Nurses are increasingly being held account-
essary to test and understand how to best promote able for practices informed by scientific evidence.
the use of this science in daily practice. Thus, nurses must communicate and integrate
Ross-Kerr and Wood (2011, p. 136) detailed into their profession the expectation that all
some of the challenges in the future of the research nurses have a professional responsibility to read
process: and use research in their practice and to commu-
Facilitating research utilization in nursing practice is
nicate with nurse scientists the many and varied
not an easy task, but neither is it an impossible clinical problems for which a scientific basis for
dream. Some activities that might facilitate the practice does not yet exist.
research process include the following:

• Funding key research positions

• Creating institutional infrastructures, including
resources needed to access summarized evidence
• Encouraging and assisting nurses to attend
practice-based research workshops and conferences
CRITICAL THINKING CHALLENGES
• Providing continuing education courses to assist ■ Discuss the differences among nursing research,
nurses in critiquing research
research utilization, and evidence-informed
• Creating a reward system within the agency for
research utilization in practice
practice. Support your discussion with examples.
• Promoting collaborative efforts between agency ■ Why would it be important to use an evidence-
personnel and other health-care agencies or informed practice model, such as the Iowa
educational institutions Model of Evidence-Based Practice to Promote
• Establishing a means for nurses to access research Quality Care, to guide a practice project focused
reports relevant to their practice on justifying and implementing a change in
• Promoting demonstration projects that illustrate clinical practice?
the cost-effectiveness of changing from a ■ You are a staff nurse working on a cardiac
traditional, intuitively based practice to one that is step-down unit. Many of your colleagues do not
research-based understand evidence-informed practice. How
• Ensuring that research utilization is a role
would you help them to understand how
expectation for all nursing positions (management,
educational, and clinical) and that it is reinforced in
evidence-informed practice is relevant to
job descriptions, agency policy, and the institution’s providing optimal care to this patient
philosophy of nursing. population?
■ What barriers do you see to applying evidence-
Diagnosing barriers, and planning and implementing informed practice in your clinical setting? Discuss
strategies to overcome them, are challenges. The first strategies to use in overcoming these barriers.
challenge is to demonstrate the cost-effectiveness of
 CHAPTER 20 Developing an Evidence-Informed Practice 509

American Geriatrics Society, British Geriatrics Society,

KEY POINTS American Academy of Orthopaedic Surgeons, and
Panel on Falls Prevention. (2001). Guideline for the
• According to the Iowa Model of Evidence-Based prevention of falls in older persons. ournal of the
Practice to Promote Quality Care, the steps of American eriatrics Society, , 664-672.
evidence-informed practice are as follows: selecting Atkins, D., Briss, P. A., Eccles, M., Flottorp, S., Guyatt,
a topic, forming a team, retrieving the evidence, G. H., Harbour, R. T., . . . The GRADE Working
grading the evidence, developing an evidence- Group. (2005). Systems for grading the quality of
informed practice standard, implementing the evidence and the strength of recommendations II:
evidence-informed practice, and evaluating the Pilot study of a new system. BMC ealth Services
effect on staff, patient, and fiscal outcomes. Research, (1), 25.
• Adoption of evidence-informed practice standards Bootsmiller, B. J., Yankey, J. W., Flach, S. D., Ward,
requires education and dissemination to staff and
M. M., Vaughn, T. E., Welke, K. F., Doebbeling,
use of change strategies, such as communication
with opinion leaders, change champions, a core B. N. (2004). Classifying the effectiveness of
group, and consultants. Veterans Affairs guideline implementation
• It is important to evaluate the change. Evaluation approaches. American ournal of Medical uality,
provides data for performance gap assessment, (6), 248-254.
audit, and feedback, and provides information Brooks, J. M., Titler, M. G., Ardery, G., Herr, K.
necessary to determine whether the practice (2009). Effect of evidence-based acute pain manage-
should be retained. ment practices on inpatient costs. ealth Services
• Evaluation includes both process and outcome Research, (1), 245-263.
measures. Canadian Institutes of Health Research (2009). About
• It is important for organizations to create a culture
knowledge translation. Retrieved from http://www.
of evidence-informed practice. Creating this culture
requires an interactive process. To provide this canadiantaskforce.ca
culture, organizations need to provide access to Canadian Task Force on Preventative Health Care
information, access to individuals who have skills (CTFPHC). (1997). uic tables by strength of
necessary for evidence-informed practice, and a evidence. Retrieved from http://www.ctfphc.org
written and verbal commitment to evidence- Clancy, C. M., Slutsky, J. R., Patton, L. T. (2004).
informed practice in the organization’s operations. AHR moves research to translation and implemen-
• The terms research utilization and evidence- tation. ealth Services Research, (5), xv-xxiii.
informed practice are sometimes used Colwell, J. C., Foreman, M. D., Trotter, J. P. (1992).
interchangeably. These terms, although related, are A comparison of the efficacy and cost-effectiveness
not one and the same. Research utilization is the
of two methods of managing pressure ulcers.
process of using research findings to improve
practice. Evidence-informed practice is a broad term Decubitus, (4), 28-36.
that not only encompasses use of research findings Cronenwett, L. R. (1995). Effective methods for
but also use of other types of evidence such as disseminating research findings to nurses in
case reports and expert opinion in deciding the practice. Nursing Clinics of North America, ,
evidence base that informs practice. 429-438.
• There are two forms of evidence use: conceptual Cullen, L. (2005). Evidence-based practice: Strategies
and decision driven. for nursing leaders. In D. Huber (Ed.), eadership
• There are several models of evidence-informed and nursing care management (3rd ed., pp. 461-478).
practice. A key feature of all models is the Philadelphia: Elsevier.
judicious review and synthesis of research and
Cullen, L., Titler, M. G. (2004). Promoting evidence-
other types of evidence to develop an evidence-
informed practice standard. based practice: An internship for staff nurses.
orldvie s on Evidence-Based Practice, (4),
215-223.
Davey, P., Brown, E., Fenelon, L., Finch, R., Gould, I.,
Hartman, G., . . . Wiffen, P. J. (2005). Interven-
REFERENCES tions to improve antibiotic prescribing practices for
Adams, S. L., Titler, M. G. (2010). Building a hospital inpatients. Cochrane Database of Systematic
learning collaborative. orldvie s on Evidence-Based Revie s, (4), CD003543. doi: 003510.001002/
Nursing, (3), 165-173. 14651858.CD14003543.pub14651852
510 PART SIX Application of Research: Evidence-Informed Practice

Dawson, J. D. (2004). uantitative analytical methods GRADE Working Group. (2004). Grading quality of
in translation research. orldvie s on Evidence- evidence and strength of recommendations. British
Based Nursing, (S1), S60-S64. Medical ournal, , 1490-1494.
DiCenso, A., Guyatt, G., Ciliska, D. (2005). Graham, I. D., Logan, J., Harrison, M. B., Straus, S. E.,
Evidence-based nursing: A guide to clinical practice. Tetroe, M. A., Caswell, W., Robinson, N. (2006).
St. Louis: Elsevier. Lost in knowledge translation: Time for a map
Doebbeling, B. N., Chou, A. F., Tierney, W. M. ournal of Continuing Education in the ealth
(2006). Priorities and strategies for the implementa- Professions, (1), 13-24.
tion of integrated informatics and communications Graham, K., Logan, J. (2004). sing the Ottawa
technology to improve evidence-based practice. model of research use to implement a skin care
ournal of eneral Internal Medicine, (S2), program. ournal of Nursing Care uality, (1),
S50-S57. 18-24.
Eccles, M. P., Mittman, B. S. (2006). Welcome to Gravel, K., L gar , F., Graham, I. D. (2006). Barriers
implementation science. Implementation Science, and facilitators to implementing shared decision-
, 1. making in clinical practice: a systematic review of
Estabrooks, C. A. (2004). Thoughts on evidence-based health professionals’ perceptions. Implementation
nursing and its science: A Canadian perspective. Science, , 16.
orldvie s on Evidence-Based Nursing, (2), 88-91. Greenhalgh, T., Robert, G., MacFarlane, F. B., Bate, P.,
Estabrooks, C. A., Chong, H., Brigidear, K., Kyriakdiou, O. (2004). Diffusion of innovations in
Profetto-McGrath, J. (2005). Profiling Canadian health service organisations: A systematic literature
nurses’ preferred knowledge sources for clinical review. The Milban uarterly, (4). Retrieved from
practice. Canadian ournal of Nursing Research, http://www.milbank.org/quarterly/8204feat.html
(2), 119-140. Grol, R. P., Bosch, M. C., Hulscher, M. E., Eccles,
Estabrooks, C. A., Derksen, L., Winther, C., Lavis, J. L., M. P., Wensing, M. (2007). Planning and studying
Scott, S. D., Wallin, L., Profetto-McGrath, J. improvement in patient care: The use of theoretical
(2008). The intellectual structure and substance of the perspectives. The Milban uarterly, (1),
knowledge utilization field: a longitudinal author 93-138.
co-citation analysis, 1945-2004. Implementation Guyatt, G. H., Oxman, A. D., Kunz, R., Jaeschke, R.,
Science, , 49. doi:10.1186/1748-5908-3-49 Helfand, M., Liberati, A., . . . Schunemann, H. J.
Feldman, P. H., Murtaugh, C. M., Pezzin, L. E., (2008a). Rating quality of evidence and strength of
McDonald, L. E., Peng, T. R. (2005). Just-in-time recommendations: What is quality of evidence and
evidence-based e-mail reminders in home health why is it important to clinicians British Medical
care: impact on patient outcomes. ealth Services ournal, (7651), 995-998. doi:10.1136/
Research, (3), 865-885. bmj.39490.551019.BE
Folkedahl, B., Frantz, R., Goode, C. (2002). Guyatt, G. H., Oxman, A. D., Kunz, R., Vist, G. E.,
Evidence-based protocol: Treatment of pressure Falck-Ytter, Y., Schunemann, H. J. (2008b). Rating
ulcers (Series Ed., Titler, M. G.). Iowa City: quality of evidence and strength of recommendations:
niversity of Iowa College of Nursing, Gerontologi- Incorporating considerations of resources use into
cal Nursing Interventions Research Center, Research grading recommendations. British Medical ournal,
Dissemination Core (P30 NR03979 PI: (7654), 1170-1173. doi:10.1136/
T. Tripp-Reimer). bmj.39504.506319.80
Fowler, E., Goupil, D. L. (1984). Comparison of the Guyatt, G. H., Oxman, A. D., Vist, G., Kunz, R.,
wet-to-dry dressing and a copolymer starch in the Falck-Ytter, Y., Alonso-Coello, P., Schunemann,
management of debrided pressure sores. ournal of H. J. (2008c). Rating quality of evidence and strength
Enterostomal Therapy, (1), 22-25. of recommendations GRADE: An emerging consen-
Gold, M., Taylor, E. F. (2007). Moving research into sus on rating quality of evidence and strength of
practice: Lessons from the S Agency for Healthcare recommendations. British Medical ournal,
Research and uality’s IDSRN program. Implementa- (7650), 924-926. doi: 10.1136/bmj.39489.470347.
tion Science, , 9. AD
Gorse, G. J., Messner, R. L. (1987). Improved Haber, J., Feldman, H., Penney, N., Carter, E., Bidwell-
pressure sore healing with hydrocolloid dressings. Cerone, S., Rose Hott, J. (1994). Shaping nursing
Archives of Dermatology, (6), 766-771. practice through research-based protocols. ournal
 CHAPTER 20 Developing an Evidence-Informed Practice 511

of the Ne or State Nurses Association, (3), sounds Report of an evidence-based practice project.
3-8. American ournal of Nursing, , 40-49.
Haller, K. B., Reynolds, M. A., Horsley, J. O. (1979). McDonald, M. V., Pezzin, L. E., Feldman, P. H.,
Developing research-based innovation protocols: Murtaugh, C. M., Peng, T. R. (2005). Can
Process, criteria, and issues. Research in Nursing just-in-time, evidence-based reminders improve
ealth, (2), 45-51. pain management among home health care nurses and
Harris, R. P., Helfan, M., Woolf, S. H., Lohr, K. N., their patients ournal of Pain Symptom Manage-
Mulrow, C. D., Teutsch, S. M., Atkins, D. (2001). ment, (5), 474-488.
Current methods of the S Preventive Services Task Melnyk, B. M., Fineout-Overholt, E. (2011).
Force: A review of the process. American ournal of Evidence-based practice in nursing and healthcare:
Preventative Medicine, (3S), 21-35. A guide to best practice (2nd ed). New York: Wolters
Higgins, J. P. T., Green, S. (Eds.). (2011). Cochrane Kluwer.
handboo for systematic revie s of interventions Neill, K. M., Conforti, C., Kedas, A., et al. (1989).
updated March 2011 . Retrieved from http:// Pressure sore response to a new hydrocolloid
www.cochrane.org/sites/default/files/uploads/ dressing. ounds, (3), 173-185.
Handbook4.2.6Sep2006.pdf Nelson, E. C., Batalden, P. B., Huber, T. P., Moher, J. J.,
Horsley, J. A., Crane, J., Crabtree, M. K., Wood, Godfrey, M. M., Headrick, L. A., Wasson, J. H.
D. K. (1983). sing research to improve nursing (2002). Microsystems in health care: Earning from
practice: A guide. New York: Grune Stratton. high performing front-line clinical units. ournal of
ICEBeRG. (2006). Designing theoretically informed uality Improvement, (9), 472-493.
implementation interventions: The Improved Clinical Nightingale, F. (1858). Notes on matters affecting the
Effectiveness through Behavioural Research Group. health, ef ciency, and hospital administration of the
Implementation Science, , 4. British Army. London, England: Harrison and Sons.
Irwin, C., Ozer, E. M. (2004). Implementing Nightingale, F. (1859). A contribution to the sanitary
adolescent preventive guidelines inal progress history of the British Army during the late ar ith
report to A R (Grant No. 18HS11095). San Russia. London, England: John W. Parker and Sons.
Francisco: niversity of California, San Francisco, Nightingale, F. (1863a). Notes on hospitals. London,
Division of Adolescent Medicine. England: Longman, Green, Roberts, and Green.
Kirchhoff, K. T. (2004). State of the science of Nightingale, F. (1863b). bservation on the evidence
translational research: From demonstration projects to contained in the statistical reports submitted by her
intervention testing. orldvie s on Evidence-Based to the Royal Commission on the Sanitary State of the
Nursing, (Suppl. 1), S6-S12. Army in India. London, England: Edward Stanford.
Kurzuk-Howard, G., Simpson, L., Palmieri, A. Oleske, D. M., Smith, . P., White, P., Pottage, J.,
(1985). Decubitus ulcer care: A comparative Donovan, M. I. (1986). A randomized clinical trial of
study. estern ournal of Nursing Research, (1), two dressing methods for the treatment of low-grade
58-79. pressure ulcers. ournal of Enterostomal Therapy,
Leape, L. L. (2005). Where the rubber meets the road. (3), 90-98.
In Advances in patient safety: rom research to Pettit, D. M., Kraus, V. (1995). The use of gauze
implementation, Vol : Implementation issues versus transparent dressing for peripheral intravenous
(AHR pub no 05-0021-3). Rockville, MD: Agency catheter sites. Nursing Clinics of North America, ,
for Healthcare Research and uality. 495-506.
Loeb, M., Brazil, K., McGeer, A., Stevenson, K., Registered Nurses Association of Ontario (2003).
Walter, S. D., Lohfeld, L., . . . outman, D. (2004). Breastfeeding best practice guidelines for nurses.
ptimi ing antibiotic use in long term care inal Retrieved from www.rnao.org/Storage/11/564 BPG
progress report to A R (Project No. Breastfeeding.pdf
2R18HS011113-03). Hamilton, ON: McMaster Registered Nurses Association of Ontario (2004). Adult
niversity. asthma care guidelines for nurses: Promoting control
Logan, J., Graham, I. (1998). Toward a comprehen- of asthma. Retrieved from http://www.rnao.
sive interdisciplinary model of health care research org/.../2206 Asthma Guideline and Supplement -
use. Science Communication, (2), 229. FIN
Madsen, D., Sebolt, T., Cullen, L., Mueller, T., Registered Nurses Association of Ontario. (2011).
Richardson, C., et al. (2005). Why listen to bowel Implementation of clinical practice guidelines.
512 PART SIX Application of Research: Evidence-Informed Practice

Retrieved from http://www.rnao.org/Page.asp PageID Health Administration. Implementation Science,

924 ContentID 823 , 23.
Rogers, E. (1995). Diffusion of innovations. New York: Stetler, C. B., Mc uenn, L., Demkis, J., Mittman, B.
Free Press. S. (2008). An organizational framework and strategic
Rogers, E. M. (2003). Diffusion of innovations (5th ed.). implementation for system-level change to enhance
New York: Free Press. research-based practice: ERI Series. Implementa-
Ross-Kerr, J. C., Wood, M. J. (2011). Canadian tion Science, , 30. doi:10.1186/1748-5908-3-30
nursing: Issues and perspectives (5th ed). Toronto: Straus, S. E., Richardson, W. S., Glasziou, P.,
Elsevier. Haynes, R. B. (2005). Evidence based medicine: o
Rubenstein, L. V., Pugh, J. A. (2006). Strategies for to practice and teach EBM (3rd ed.). Philadelphia:
promoting organizational and practice change by Churchill Livingstone.
advancing implementation research. ournal of Sussman, S., Valente, T. W., Rohrbach, L. A., Skara, S.,
eneral Internal Medicine, , S58-S64. Pentz, M. A. (2006). Translation in the health
Rycroft-Malone, J., Kitson, A., Harvey, G., McCormack, professions: Converting science into action. Evalua-
B., Seers, K., Titchen, A., Eastabrooks, C. (2002). tion and the ealth Professions, (1), 7-32.
Ingredients for change: Revisiting a conceptual Thompson, D. S., Estabrooks, C. A., Scott-Findlay, S.,
framework. uality Safety ealth Care, , Moore, K., Wallin, L. (2007). Interventions aimed
174-180. at increasing research use in nursing: A systematic
Saydak, S. J. (1990). A pilot test of two methods for the review. Implementation Science, , 15. doi:
treatment of pressure ulcers. ournal of Enterostomal 10.1186/1748-5908-2-15
Therapy, (3), 139-142. Titler, M. G. (1993). Critical analysis of research
Scott, S. D., Plotnikoff, R. C., Karunamuni, N., Bize, utilization (R ): An historical perspective. American
R., Rodgers, W. (2008). Factors in uencing the ournal of Critical Care, (3), 264.
adoption of an innovation: An examination of the Titler, M. G. (2002). Tool it for promoting evidence-
uptake of the Canadian Heart Health Kit (HHK). based practice Iowa City: niversity of Iowa
Implementation Science, , 41. Hospitals and Clinics, Department of Nursing
doi:10.1186/1748-5908-3-41 Services and Patient Care.
Sebern, M. D. (1986). Pressure ulcer management in Titler, M. G. (2006). Developing an evidence-based
home health care: Efficacy and cost effectiveness and practice. In G. LoBiondo-Wood J. Haber (Eds.),
moisture vapor permeable dressing. Archives of Nursing research (6th ed.). St. Louis: Elsevier
Physical Medicine and Rehabilitation, (10), Mosby.
726-729. Titler, M. G. (2008). The evidence for evidence-based
Shojania, K. G., Grimshaw, J. M. (2005). Evidence- practice implementation. In R. Hughes (Ed.), Patient
based quality improvement: The state of the science. safety and uality an evidence-based handboo for
ealth Affairs, (1), 138-150. nurses. Rockville, MD: Agency for Healthcare
Silversides, A. (2011). Journal clubs: A forum for Research and uality.
discussion and professional development. Canadian Titler, M. G., Everett, L. . (2001). Translating
Nurse, (2), 18-23. research into practice: Considerations for critical care
Smith, J. L., Williams, J. W., Owen, R. R., Rubenstein, investigators. Critical Care Nursing Clinics of North
L. V., Chaney, E. (2008). Developing a national America, (4), 587-604.
dissemination plan for collaborative care for depres- Titler, M. G., Everett, L. ., Adams, S. (2007).
sion: ERI Series. Implementation Science, , 59. Implications for implementation science. Nursing
doi:10.1186/1748-5908-3-59 Research, (4S), S53-S59.
Squires, J. E., Moralejo, D., LeFort, S. M. (2007). Titler, M. G., Herr, K., Brooks, J. M., ian-Jin, .,
Exploring the role of organizational policies and Ardery, A., Schilling, J., . . . Clarke, W. R. (2009).
procedures in promoting research utilization in A translating research into practice intervention
registered nurses. Implementation Science, , 17. improves management of acute pain in older hip
Stetler, C. A., Legro, M. W., Rycroft-Malone, J., fracture patients, ealth Services Research, (1),
Bowman, C., Curran, G., Guihan, M., . . . Wallace, 264-287.
C. M. (2006). Role of external facilitation in Titler, M. G., Herr, K., Everett, L. ., Marsh, J. L.,
implementation of research findings: A qualitative Schilling, M., Brooks, J., et al. (2006). Boo to
evaluation of facilitation experiences in the Veterans bedside: Promoting sustaining EBPs in elders
 CHAPTER 20 Developing an Evidence-Informed Practice 513

(Final Progress Report to AHR , Grant No. 2R01 Wensing, M., Wollersheim, H., Grol, R. (2006).
HS010482-04). Iowa City: niversity of Iowa Organizational interventions to implement improve-
College of Nursing. ments in patient care: A structured review of review.
Titler, M.G., Kleiber, C., Steelman, V., Goode, C., Implementation Science, (2), doi:10.1186/
Rakel, B., Barry-Walker, J., . . . Buckwalter, K. 1748-5908-1-2
(1994). Infusing research into practice to promote West, S., King, V., Carey, T. S., Lohr, K. N., McKoy,
quality care. Nursing Research, (5), 307-313. N., Sutton, S. F., Lux, L. (2002). Systems to rate
Titler, M. G., Kleiber, C., Steelman, V. J., Rakel, B. A., the strength of scienti c evidence. (Evidence Report/
Budreau, G., Everett, L. ., . . . Goode, C. J. Technology Assessment No. 47, prepared by the
(2001). The Iowa Model of Evidence-Based Practice Research Triangle Institute- niversity of North
to Promote uality Care. Critical Care Nursing Carolina Evidence-Based Practice Center under
Clinics of North America, (4), 497-509. Contract No. 290-97-0011, AHR Publication No.
Tripp-Reimer, T., Doebbeling, B. N. (2004). 02-E016.) Rockville, MD: Agency for Healthcare
ualitative perspectives in translational research. Research and uality.
orldvie s on Evidence-Based Nursing, (S1), akellis, G. C., Chrischilles, E. A. (1992). Hydrocol-
S65-S72. loid versus saline gauze dressings in treating pressure
.S. Preventive Services Task Force. (2008). ulcers: A cost-effectiveness analysis. Archives of
S Preventive Services Tas orce grade Physiotherapy and Medical Rehabilitation, (5),
de nitions. Retrieved from http://www. 463-469.
uspreventiveservicestaskforce.org/uspstf/grades.htm
Ward, M. M., Evans, T. C., Spies, A. J., Roberts, L.,
FOR FURTHER STUDY
Wakefield, D. (2006). National uality Forum 30 safe
practices: priority and progress in Iowa hospitals. Go to Evolve at http://evolve.elsevier.com/
American ournal of Medical uality, (2), Canada/LoBiondo/Research for Audio Glossary, how-to
101-108. instructions for Writing Proposals for Funding, and
Weiss, C. H. (1980). Knowledge creep and decision additional research articles for practice in reviewing and
accretion. Science Communication, (3), 381-404. critiquing.
AP PENDIX A

Understanding Nursing on an Acute Stroke Unit:

Perceptions of Space, Time and Interprofessional
Practice
Cydnee C. Seneviratne, PhD, RN • Post-Doctoral Fellow and Instructor • Faculty of Nursing • University of Calgary •
Calgary, Alberta

Charles M. Mather, PhD • Assistant Professor • Department of Anthropology • University of Calgary • Calgary, Alberta
Karen L. Then, PhD, RN, ACNP • Professor • Faculty of Nursing • University of Calgary • Calgary, Alberta

TITLE. Understanding nursing on an acute stroke unit: an interprofessional atmosphere is fundamental for collab-
perceptions of space, time and interprofessional practice. orative stroke practice, despite working in a multiprofes-
sional environment.
AIM. This paper is a report of a study conducted to
uncover nurses’ perceptions of the contexts of caring for CONCLUSION. Understanding how care providers conceive
acute stroke survivors. of and respond to space, time and interprofessionalism has
the potential to improve acute stroke care. Future research
BACKGROUND. Nurses coordinate and organize care and focusing on nurses and other professionals as members of
continue the rehabilitative role of physiotherapists, occupa- interprofessional teams could help inform stroke care to
tional therapists and social workers during evenings and at enhance poststroke outcomes.
weekends. Healthcare professionals view the nursing role
as essential, but are uncertain about its nature. KEYWORDS: acute stroke unit, ethnography, interprofes-
sional practice, nursing, perceptions, space, time
METHOD. Ethnographic fieldwork was carried out in
2006 on a stroke unit in Canada. Interviews with nine INTRODUCTION
healthcare professionals, including nurses, complemented Stroke can be a devastating and physically debili-
observations of 20 healthcare professionals during patient
tating cardio/neurovascular disease (Hickey
care, team meetings and daily interactions. Analysis
methods included ethnographic coding of field notes and 2003). It interrupts life, arrests previously cher-
interview transcripts. ished activities, and decreases overall quality of
life for survivors and their families (Canadian
FINDINGS. Three local domains frame how nurses under- Stroke Network, 2007). According to the World
stand challenges in organizing stroke care: 1) space, 2) time Health Organization (2004), heart attack and
and 3) interprofessional practice. Structural factors force stroke are leading causes of death in the world
nurses to work in exceptionally close quarters. Time con- and approximately 15 million people worldwide
straints compel them to find novel ways of providing care.
survive stroke annually.
Moreover, sharing of information with other members of
the team enhances relationships and improves “interpro- Dedicated stroke units are part of a widespread
fessional collaboration”. The nurses believed that effort to ameliorate the impact of stroke. Gener-
ally, these units house comprehensive stroke pro-
From the ournal of Advanced Nursing, 2009 65(9):1872- grammes which include interdisciplinary teams
1881. © 2009 Blackwell Publishing Ltd. of caregivers including nurses, pharmacists,
514
APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice 515

physicians, nurse practitioners (NP), social other professions (O’Connor 1993, 2000a). This
workers, occupational and physical therapists, emerging role is patchy in that nursing ability
and speech therapists. The rationale behind the in some areas (e.g. feeding and continence care)
programmes is that the needs of individual is advancing while other areas (e.g. mobility and
patients require caregivers with varied expertise exercise therapies) are lagging behind (Perry
(Teasell et al 2007). Stroke units yield clear ben- et al 2004). Without knowledge and skills in
efits to patients (Hill 2002, Stroke nit Trialists’ acute stroke care, and accepting rehabilitation as
Collaboration (S TC) 2007 Indredavik et al a normal part of nursing, Henderson’s vision is
1991, Jorgensen et al 1995a, 1995b, Kalra et al unattainable (Myco 1984).
1993, Kalra 1994). Patients who receive inpatient The role of nurses in acute stroke rehabilitation
stroke care vs. care from a conventional or general is unclear (O’Connor 1993, 2000a, 2000b, Forster
medical ward stand a better chance of surviving, Young 1996, Kirkevold 1997, Burton 1999,
and living independently at home 1 year post- 2000, Elliott 1999, Thorn 2000). Researchers
stroke (S TC 2007). Patients on stroke units have identified nurses as managers or coordina-
experience greater improvement in functional tors (O’Connor 1993, Burton 1999, 2000), clini-
outcome and quality of life, and a decreased cal specialists (Elliott 1999), community
length of stay (Cifu Stewart 1999). integrators (Forster Young 1996) and caregiv-
Researchers have proven the effectiveness of ers who perform interpretive, consoling, conserv-
interdisciplinary stroke care (Cifu Stewart ing and integrative tasks (Kirkevold 1997).
1999 S TC 2007). This study looks beyond Kirkevold (1997) describes four unique functions
outcome measures to the daily interactions and in rehabilitation of stroke survivors that nurses
beliefs that characterize comprehensive pro- perform but fails to operationalize these functions
grammes, with a particular focus on the role of (O’Connor 2000a).
nurses. Nurses have long believed that they play Nurses vary in their attitudes and perceptions
an essential role in stroke care, but they remain of their role in stroke care. In an early qualitative
uncertain about the nature of their contributions study, Waters and Luker (1996) found that nurses
(Gibbon 1993, Gibbon Little 1995, Waters thought that they were good at basic care, ranging
Luker 1996, Burton 2000). sing an ethnographic from ensuring that patients were clean and dressed
approach, we examined nurses’ perspectives on prior to medical assessments to ensuring patients
the contribution they make to the care of stroke were physically ready prior to therapy sessions,
survivors in acute settings. Part of our concern but the sic considered that they had little time
was how nurses see the social connections they for rehabilitative care. Burton (2000) discovered
have with other members of the interdisciplinary that nurses provided care, facilitated personal
team, and what sort of values they hold toward recovery, and managed multidisciplinary care
their practice. teams, and that these roles suggested that they
could provide focused 24-hour coordinated stroke
BACKGROUND rehabilitation. Perry et al (2004) agreed with
Henderson (1980) offered a vision of nurses with Burton (2000) and suggested that nurses must
a leading role in acute care and rehabilitation, move beyond their traditional role of providing
including working with patients to relearn activi- basic care and become active participants in acute
ties, movement, and continence and nutritional and rehabilitative care.
care. In contrast to this vision, nurses are moving Observational studies of nurses in acute and
toward a managerial or understudy role that coor- rehabilitative care settings includes work by
dinates rehabilitative tasks under the guidance of Pound and Ebrahim (2000) showing that nurses
516 APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice

on a general medical unit and a stroke unit pro- observation and interviewing. Ethnographers
vided impersonal, standardized care, considered record their observations and interviews in field
rehabilitation secondary to nursing practice and notes and other media, including audio and visual
did not regularly consult with therapists. In con- formats. Post-fieldwork, researchers analyse and
trast, nurses on an elder care unit valued and interpret the records to uncover dominant themes
promoted patient independence, and frequently or understandings among members of the culture
consulted therapists to encourage optimal reha- (Spradley 1979, 1980, Aamodt 1991). In this
bilitation. The authors concluded that optimal study, ethnography provided a means of explor-
stroke care requires engaging nurses in rehabilita- ing how stroke unit nurses organized and coordi-
tion, increasing training in rehabilitation and nated care.
compassionate care. Booth et al (2001) com-
pared interventions by nurses with those by occu- Participants and Setting
pational therapists (OT) and found that OT used The study took place on an18-bed acute stroke
patient prompting and facilitation while nurses unit located in a large tertiary medical centre in
favoured supervision. Variation in care practice Canada. As part of a greater health region, the
was because of different intervention and assess- stroke unit provided specialized interventions,
ment styles and lack of preparation and education management and investigative care during acute
in stroke rehabilitation on the par sic of the and sub-acute stroke phases. It had 18 beds, two
nurses. located beside a nursing station and 16 in four-
According to Bukowski et al (1986), neuro- bed rooms with connecting corridors leading to
science nurses could implement rehabilitation a nursing station. Located on a different oor
therapy over the 24 hour period, support treat- from the neurosciences department, the stroke
ments recommended by physiotherapists (PT), unit shared space with another general neuro-
and ensure that patients and families learn therapy logical unit.
techniques to continue rehabilitation at home. Staff on the unit included registered nurses
Nurses play an essential role in acute and reha- (RN), licensed practical nurses (LPN), patient
bilitative stroke care (Gibbon 1993, Gibbon care attendants (PCA), NP, PT, OT, speech thera-
Little 1995, Waters Luker 1996, Burton 2000) pists and physicians. The staffing ratio was one
but the broader social construction of stroke reha- RN or LPN to every four patients. We used pur-
bilitation and care providers’ perceptions toward posive sampling (Morse Field 1995, Hammer-
this construction remains unclear. sley Atkinson 2007) to locate participants, and
excluded individuals who could not read, write,
THE STUDY or speak English. In total, we followed ten RN,
two LPN, one PCA, one NP, two PT, a PT and
Aim three physicians (n = 20), nine of whom we for-
The aim of the study was to uncover nurses’ per- mally interviewed. Participants ranged in age
ceptions of the contexts of caring for acute stroke from 24 to 52 years, 15 were female and five were
survivors. male. Nurses were the study focus but we also
interviewed four other professionals to help con-
Methodology textualize interprofessional perspectives.
Ethnography is a qualitative research approach
(Spradley 1979, 1980). In anthropology, ethno- Data Collection
graphy is a tool for describing cultures, and Fieldwork took place from February to Novem-
the chief methods ethnographers employ are ber of 2006. Observations averaged 2–3 hours on
APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice 517

3 days/week. The fieldworker (CS) made obser-

vations during every type of shift. The fieldwork Reflexivity
began with 3 months of general observation on Ethnographers change in the process of conduct-
the unit with the fieldworker watching at the ing research and spending an extensive period
charting desk, nursing station and walking in the learning cultural domains and categories, and
hall. Notes from observations were transcribed from building relationships with study partici-
via computer. The fieldworker clarified gaps in pants (Davies 1999, Hammersley Atkinson
field notes by returning to the field site and 2007). Indeed, if an ethnographer’s ideas, beliefs
making more focused observations driven by and values did not change it might indicate that
informant comments. To explore work practices, she failed in understanding the culture she was
we interviewed participants (Emerson et al studying. To keep track of the sorts of changes
1995) using grand tour questions such as Can they undergo and the impact of these changes on
you walk me through your typical day and their fieldwork, ethnographers employ re ex-
asking for examples of nursing practice vs. that ive techniques. In this study, the fieldworker
of other professions. used asides , integrative memos and research
journals (Spradley 1979, 1980, Emerson et al
Ethical Considerations 1995) to make explicit her presuppositions and
The local health research ethics board granted insider relationships, and to maintain awareness
ethical approval for the study. We used several of her social position within the culture. Some
information sessions to introduce the study to staff knew the researcher as a nursing instructor
members of the stroke unit. During these or a previous employee. Asides and journaling
sessions and interviews we informed stroke helped her denote prior relationships, and high-
unit members about patient confidentiality. We light instances when she might have had built-
sought written consent only from observation and in biases or made priori judgments. Re exive
interview participants and verbal assent from techniques helped control for potential biases
patients when participants were providing from having an insider perspective, which we did
direct care. not want to lose because it was essential for
gaining access to the study setting, building
Data Analysis rapport and building and maintaining trust with
Analysis of field notes focused on identifying staff.
central domains, specific domain components and
related work typologies (Spradley 1979, 1980, FINDINGS
Hammersley Atkinson 2007). We discovered
three main themes (domains) and further identi- Space
fied theme components (componential analysis). Participants described space as a challenge to
Finally, we broke work activities into types on the patient care. Three themes predominated in
basis of relationships between nurses, and the data: nursing in a submarine , nursing
between nurses and other professionals (typologi- too close and nursing in a state of code
cal analysis). Through an ongoing process of burgundy .
reading field notes, transcripts and then returning
to the field setting for further observations, we Nursing in a Submarine
crosschecked our findings and were assured that Staff used the term submarine to refer to the
our study domains, components and related sub- unit. Command centres are located near the front
categories were culturally salient. of submarines, with missile rooms at the rear. The
518 APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice

nursing station was located near the entrance of hall. Caring for patients in hallways compounded
the unit, while the majority of four-bed rooms having to work too close . Nurses disapproved
were located at the middle and far end. According of nursing under a state of code burgundy:
to one nurse:
[We] feel badly for the lack of privacy for that patient
Our submarine . . . it’s just a more condensed unit. in the hallway. I mean even I had to perform an
But the thing that most bothers me is it’s not centred. intimate procedure, a urinary catheter insertion in the
If you have patients in the last room . . . at the other hall, and I hated doing it.
end you are not in close proximity to anything or
anybody—you’re alone. That drives me crazy because Ironically, despite disapproving of these condi-
the nursing station is so far away.
tions, nurses often faced criticism from patients’
Participants connected their feelings of claustro- families for the conditions:
phobia on the unit with being on a submarine. The I think it really stressed [us] out, because [we] were
unit lacked work space and storage space: taking the brunt of family complaints. You try telling
the patient that you are just following procedure. This
I’m too claustrophobic on this unit. It’s like I am closed is a region issue and we are obligated to do what the
in . . . if you look down the hall from the nursing region tells us, but we don’t agree.
station you feel like the walls and curtains are closing
in around you. It is so narrow. I feel constricted A state of code burgundy means increased work-
because I cannot do my work in cramped space. I loads and the ethical challenge of hallway care .
bump into other people all the time.
Time
Nursing Too Close Participants’ talked about time in three major
Limited space made it difficult to move, to use ways: lack of time , preserving time and time
and relocate equipment, transfer patients, docu- with and without space . Each concept denotes
ment nursing care and interact with colleagues. limitations and challenges to providing care.
Limited space required alternative work strate-
gies to ensure that one did not get in the way of Lack of Time
one’s colleagues. For example, nurses unlocked Participants complained that care errors, missed
the wheels on beds in rooms near the nursing therapy or treatment appointments, and awkward
station, wheeled them to an open space, and then patient transfers occurred because of lack of time
transferred patients to stretchers. One nurse said: to plan. Lack of time also compromised nurses’
This unit is not set up for us to nurse or do rehab. It is
wellbeing. They associated work-related injuries
designed so that we are constantly bumping bums, to the pressures of needing to work quickly to
literally bum to bum . . . when we transfer patients. complete all the work before the end of a shift:
We are bashing into one another when we feed
patients and when we provide any kind of nursing We are always injuring ourselves because we rush
care. around. There is just not enough time for us to do
things properly with our patients . . . So if things get
The layout of the unit caused nurses to bump into missed so be it.
each other, and put patients in situations where
staff could not ensure appropriate care, privacy or Lack of time hindered working on patient
confidentiality. rehabilitation. Participants knew that correct
positioning and transferring of patients assists in
Nursing Under a State of “Code Burgundy” stroke rehabilitation, but they believed that they
A code burgundy signals a lack of beds. The lacked the time for patients to move and position
unit included an over-capacity bed in a shared independently:
APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice 519

It is easier to take over for patients, dressing them or therapy in the main therapy department. A PT
brushing their teeth rather than helping them do the commented:
tasks. It is a matter of accomplishing what is required
for patients in a specific window of time. I know that it is much easier for [physiotherapists] to
do transfers in the main therapy department because
Nurses organized their time according to what the setup is ideal. There is so much of a space conflict
they believed they were physically capable of on the unit that it’s really hard sometimes to set
accomplishing during the work day. When patient things up optimally, so we would rather work with
patients down in the gym without the nurses.
acuity was high there were time constraints and
rehabilitation was not a priority: One nurse said it was hard to do rehabilitation on
the unit because of limited time and space:
The patients need time for us to let them do what
they can and . . . for themselves. But, that requires a There are limitations on what we can do in the time
whole lot of time and effort, which the nurses don’t and space allotted. To be able to come in and have
have. So I am sure some of [the patients] are the time to do all of those extra things . . . like assist
frustrated because they realize that and they aren’t patients in their room with feeding or mobilization
able to do as much as they would like to do . . . and all of those things . . . that would be great, but it
Some days you just can’t wait, you have to get it doesn’t happen.
done and move on.
Time with and with out adequate space affected
nurses’ participation in bedside rounds. nit
Preserving Time policy stated that nurses should attend and review
To preserve time nurses coordinated their work patients’ neurological status, vital signs and
and cared for patients as a team. They met and changes in condition. Rounds occurred at 09.00
identified tasks they could do more efficiently hours, when nurses were preparing patients for
working together. Alternately, individuals who therapy appointments and tests, and/or were pro-
had fewer patients volunteered to pick-up viding acute medical care. For nurses attendance
patients from colleagues who had a heavier at rounds was not a priority:
patient load. A more implicit approach to preserv- Then there is the issue of doing rounds on the unit
ing time developed as a consequence of familiar- with the docs. I really do not like the idea because
ity. As one nurse related: your time is so compressed. You have so much going
on during the day and to just repeat what the [charge
For me team nursing is knowing who you work with. I nurse] already knows is . . . well, just repetitive.
don’t know whether it’s the people I normally work
One reason to miss bedside rounds was a lack
with, but we just know each others [sic] rhythms. It is
a matter of not talking about what we should do but of space in the four-bed rooms. Field observa-
just knowing each other. tions revealed that the stroke neurologist, stroke
residents, one or both NP and the charge nurse
Participants organized their work to meet time- attended bedside rounds. Gathered around each
lines and prescribed schedules. They prepoured bed, the group discussed patient status. Accord-
medications, and arrived early to complete stroke ing to the nurses there was never enough room
assessments, vital signs and morning care prior to any way , and attendance did not give them new
the start of their shift. patient information and was thus a waste of
time .
Time With and Without Space
Time and space were evolving and intercon- Interprofessional Practice
nected concepts. As result of a lack of space, Participant descriptions of interprofessional
patients received physiotherapy and occupational practice included two main components: rela-
520 APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice

tionships between stroke professionals and relatively minor. It’s all about excellent nursing care
communication/collaboration . Each compo- and rehabilitation. So the team, by accident of history
nent highlights how participants understood and hierarchy, is led by a physician but we have a NP,
all the nurses, the physiotherapist, and social work
interactions on the stroke unit and interprofes- . . . Everybody is involved in care including home care
sional practice more generally. planning, etc.

Relationships Between Stroke Professionals Another participant commented:

Nurses’ understanding of stroke care differed
We needed something different on our stroke unit.
from other professionals and this affected how Our unit was not designed to be exclusionary in any
they developed and maintained relationships on way. It was intended to be interprofessional.
the stroke team in three major ways. First, working
together and sharing experiences were the corner- The stroke unit was supposed to be an inter-
stones of relationships between nurses. disciplinary place where staff felt comfortable
Second, relationships between nurses and ther- sharing information. It would work in a non-
apists developed around perceptions of how best hierarchical way, with each professional’s opinion
to address patient needs. Nurses felt that thera- being valued and forming part of the plan for
pists’ failure to recognize and acknowledge the medical and rehabilitative care.
role of nurses in rehabilitation can lead to resent- Stroke nurses thought that collaboration was
ment and half-hearted attempts at rehabilitation. important, and that it involved open communica-
One nurse lamented: tion with physicians and NP. They felt that they
We are not recognized for the mobility things we do
had positive relationships with the two NP
or in any concerns we have about our patients. So, because these professionals were approachable
sometimes we don’t work hard at it. The physios are and accessible during acute and non-acute situa-
only concerned that we get the patients ready for their tions. One nurse asserted:
rehab times in the gym. So we do that for them and
then concentrate on our patients’ medical needs. Having both NP has helped as sort of go between to
advocate to the doctors for the things that we need.
Nurses felt that therapists had a narrow view of The NP are more available to come see the patient in
nursing practice. In the course of their profes- an emergent or urgent situation. That helps because
sional interactions, nurses were responsible for you can get the ball rolling for whatever procedures
patients’ medical needs while therapists were that need to be done.
responsible for rehabilitation.
Third, nurses saw their relationships with Communication and Collaboration
stroke physicians and NP in terms of interprofes- Participants claimed that communication and col-
sionalism. Participants claimed that in its original laboration were critical to the success of the unit.
state the stroke team was a matter of interprofes- Regular communication ensured everyone under-
sional practice. Nurses and physicians candidly stood how other members envisioned patient
discussed patients and collaborated. One of the care. Weekly stroke rounds provided opportuni-
stroke physicians summarized his view of the ties to discuss patients, collaborate on care and
situation: share alternate plans or possibilities. Attendance
did not guarantee participation, and inclusion in
It should be interprofessional, ideally. In general,
discussions was predicated on who led rounds.
stroke units are interdisciplinary. Only a small part of
stroke unit care is the physician roles. So during most One stroke team member commented:
of stroke care, beyond the acute phase, when you When I lead rounds I like to spend more time on the
have somebody settled, the physician’s role is functional and the social end of things than the
APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice 521

medical. I want to make sure the team feels like Weinberg (2003) notes that nurses have long
they’re involved and valued. faced a lack of resources to complete daily tasks
This individual thought that it was important to safely and effectively, interact with patients, or
include all team members, and was concerned attend in-service education sessions. Notwith-
that some did not feel valued or believe that they standing the apparent universality of these prob-
were in uencing progression and discharge plans lems, in our view it is not advisable to dismiss or
for their patients. devalue the concerns of those who work in care-
Nurses were the only team members who did giving environments. Our investigation presented
not regularly attend stroke unit rounds because an opportunity to re-open dialogue about the
they did not have enough time to leave their importance of institutional organization and
required duties and attend an hour-long stroke structure regarding appropriate space and use of
round’. One of the physicians claimed that the time related to stroke care (Peszczynski et al
attendance of nurses at the stroke rounds would 1972, lrich et al 2004).
have provided more information for both doctors A substantial body of literature supports the
and nurses: view that organized stroke unit care improves
I think that maybe if we could arrange the time for
stroke outcomes (Indredavik et al 1991, Kalra
once a week in the [stroke unit] rounds for the nurses et al 1993, Kalra 1994, Jorgensen et al 1995a,
to attend. I think that might be beneficial in the long 1995b, Hill 2002, S TC 2007). What the litera-
run for all the staff because we can learn so much ture fails to address is the importance of adequate
from each other especially about stuff we don’t have work space for providing this care. Our partici-
time to find out.
pants perceived lack of space as a constant chal-
The attendance of nurses at stroke rounds would lenge to providing care. Rather than describe
have reinforced the notion that the unit was what they did regarding stroke and rehabilitative
interprofessional. care, nurses talked about what they were forced
to do because of inadequate spaces and insuffi-
DISCUSSION cient time. They did not take the time to assist
patients to wash, dress and practise mobilization.
Study Limitations They complained about inadequate physical
Ethnography is not a science of generalization. space for medication delivery, charting and inter-
Ethnographic findings come from certain indi- actions with patients. These comments are con-
viduals, situations or single cases from a particu- cerning because they show that nurses did not
lar context and a particular time (Hammersley (and probably cannot) make rehabilitation and
Atkinson 2007). Our findings are not necessarily patient autonomy (Burton 1999, 2000) priorities
indicative of what happens on all stroke units and in their acute stroke care.
thus the study is not about how stroke units are, How conceptions of time affect work practices
but rather about how a stroke unit can be. Although in stroke care have not been explored in the lit-
we accept limitations where representativeness is erature. In the past, stroke physicians adopted
concerned, we also believe that we have found watchful waiting for patient recovery. Throm-
common issues in stroke care in particular, and bolytic therapy changed stroke care in the 1980s,
medical care more broadly. providing a means of treating a class of acute
cases of ischaemic stroke. The window of oppor-
Space and Time tunity for thrombolysis is three hours after
Space limitations and time constraints are the back- symptom onset. Members of the stroke team now
drop of clinical care throughout North America. use phrases such as time is brain as a reminder
522 APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice

that the longer it takes for intervention, the greater positioned to assist in the development and evalu-
the resulting neurological deficit (Barber et al ation of appropriate staffing levels on rehabilita-
2005). Team members have a second term, door tion units through documentation of how they
to needle time , which refers to the time between spend their time at work.
a patient’s arrival at the hospital and the start of
thrombolytic therapy (Hill et al 2000). Temporal Interprofessional Practice
metaphors denote boundaries and different The nurses in our study discussed how their role
dimensions within the work space (Gell 1996, in the interprofessional team developed out of
Bluedorn 2002, Hearn Michelson 2006, day-to-day working relationships. The stroke
Patmore 2006). Thrombolysis made stroke an team is multidisciplinary because each team
acute event, and thereby helped radically alter the member works independently and reports assess-
practice of stroke care. Armed with a novel inter- ments and interventions mainly in team meetings.
vention, stroke teams had to work within the Nevertheless, participants used the term interpro-
boundaries of a narrow therapeutic window. Both fessional to refer to the stroke team. This suggests
the acknowledgment sic that organized stroke a lack of clarity about the type of work relation-
units improves outcomes and the temporal ships and team interactions that exist on the unit.
demands of thrombolysis motivated the adoption In multidisciplinary teams individuals work sepa-
of coordinated interprofessional care teams and rately and come together to share information,
units such as the stroke unit in this study. while interdisciplinary teams members collabo-
Although the ideal acute stroke care team rate to create care plans as they jointly assess and
includes a focus on rehabilitation, the nurses in treat patients (Ovretveit 1997, Sorrells-Jones
this study chose not to consistently spend time 1997, Payne 2000, Pollard et al 2005). Health-
walking with their patients or taking time to assist care professionals commonly use the terms syn-
with dressing and grooming (all important reha- onymously, although in the case of acute stroke
bilitation activities). They believed that if they care interprofessional and interdisciplinary teams
had the time, they would perform rehabilitative are the gold standard (Canadian Stroke Network
care. A recent study (V h kangas et al 2008) and the Heart and Stroke Foundation of Canada:
showed that nurses who incorporated rehabilita- Canadian Stroke Strategy 2006 S TC, 2007
tion into their daily care increased the amount of Teasell et al 2007).
time working with patients to maximize patient Despite desiring to work interprofessionally,
independence. The question is whether our stroke team members found it difficult to communi-
unit nurses were aware that a re-evaluation of cate and collaborate consistently. Participants
their time from focusing on nursing tasks to facil- explained that only some nurses wanted to attend
itating rehabilitation might increase or preserve rounds and perceived that only some members of
time spent with their patients. Organizing and the stroke team valued nursing attendance. These
implementing an education session for the stroke findings are consistent with literature exploring
nurses about facilitation of care, as established by team members’ perceptions of nurse attendance
Booth et al (2005), could increase their use of at unit rounds or team meetings (Cott 1998, Mil-
facilitative interventions in rehabilitation. They ligan et al 1999). According to Cott (1998),
could have advocated for change by documenting nurses do not regularly attend team meetings
their lac of time concerns and by requesting except through a representative such as a charge
more staff through evaluation of patient acuity nurse. The exclusion of nurses may be as a result
levels. A recent American study (Neatherlin of lack of interest in attending team meetings
Prater 2003) illustrates that nurses are well or to how a culture typically organizes team
APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice 523

meetings. In our study, nurses said that whether Stroke nurses worldwide must embrace pro-
or not they felt welcome depended on which phy- fessional development and attend education ses-
sicians were present. ltimately, nurses stopped sions regarding the use of facilitative interventions
attending because of time constraints and their in rehabilitation. We see no reason that nurses
perception that the charge nurse could provide cannot take on a leadership role as rehabilitation
requisite information on their behalf. practitioners who promote working with rather
than doing for their stroke survivors. Further-
CONCLUSION more, nurses ought to become advocates for work
Nurses are an undervalued and underutilized spaces and temporal environments appropriate
resource in rehabilitation. Our study shows that for patients admitted to acute stroke units.
in some cases nurses hold themselves back from
incorporating rehabilitation principles, and that ACKNOWLEDGEMENT
they believe this occurs because of real world Many thanks go to Dr Kathryn King for assis-
structural and temporal work issues. An embed- tance with manuscript preparation and to the late
ded cultural belief exists that nurses only have Dr Marlene Reimer, mentor and colleague.
time for basic care and that rehabilitative care
requires expert knowledge usually held by PT and FUNDING
OT. However, nurses do not work in isolation and Dr Cydnee Seneviratne received funding for this
have the capacity to work with other profession- doctoral research from the Canadian Association
als outside traditional boundaries. of Neuroscience Nurses research fund and
from the F T RE Program for Cardiovascular
WHAT IS ALREADY KNOWN ABOUT THIS TOPIC Nurse Scientists, a CIHR Strategic Training
• Stroke is a devastating neurovascular disease that Fellowship.
affects over 15 million people worldwide annually.
• Organized stroke units decrease overall mortality
and average length of stay, improve quality of life, CONFLICT OF INTEREST
independence and likelihood of living at home No con ict of interest has been declared by the
1 year poststroke.
• Nurses are important and essential members of
authors.
interprofessional stroke teams as they work with
and care for patients 24 hours a day. AUTHOR CONTRIBUTIONS
WHAT THIS PAPER ADDS CS, SM and KLT were responsible for the study
• Limited work space and lack of time to care for conception and design CS performed the data
patients are important issues for neuroscience collection CS, CM and KLT performed the data
nurses. analysis CS and CM were responsible for the
• Interprofessional practice is a key factor that
requires re-evaluation in acute stroke care. drafting of the manuscript CS and CM made
• Nurses should assume a leadership role as critical revisions to the paper for important intel-
rehabilitation practitioners who promote “working lectual content CS and KLT obtained funding
with” rather than “doing for” their patients.
CS provided administrative, technical or material
IMPLICATIONS FOR PRACTICE AND/OR POLICY support KLT supervised the study.
• Providing education sessions for stroke nurses
about facilitation of care could increase nursing use
of facilitative interventions in rehabilitation.
• Nurses ought to become advocates for change by REFERENCES
documenting their space and time concerns and Aamodt A.M. (1991) Ethnography and epistemology:
by requesting workspaces and temporal
Generating nursing knowledge. In ualitative
environments appropriate for stroke survivors.
Nursing Research: A Contemporary Dialogue, 2nd
524 APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice

edn (Morse J.M., ed.), Sage, Newbury Park, Emerson R.M., Fretz R.I. Shaw L.L. (1995) riting
pp. 40-53. Ethnographic ield Notes. niversity of Chicago
Barber P.A., Hill M.D., Eliasziw M., Demchuk A.M., Press, Chicago and London.
Pexman J.H.W., Hudon M.E., Tomanek A., Frayne R. Forster A. Young J. (1996) Specialist nurse support
Buchan A. (2005) Imaging of the brain in acute for patients with stroke in the community:
ischemic stroke: comparison of computed tomogra- A randomised controlled trial. British Medical
phy and magnetic resonance diffusion-weighted ournal (7047), 1642-1646.
imaging. ournal of Neurology Neurosurgery and Gell A. (1996) The Anthropology of Time: Cultural
Psychiatry , 1528-1533. Constructions of Temporal Maps and Images. Berg,
Bluedorn A.C. (2002) The uman rgani ation of Oxford.
Time: Temporal Realities and E perience. Stanford Gibbon B. (1993) Implications for nurses in approaches
niversity Press, Stanford, CA. to the management of stroke rehabilitation: a review
Booth J., Davidson I., Winstanley J. Waters K. (2001) of the literature. International ournal of Nursing
Observing washing and dressing of stroke patients: Studies , 133-141.
nursing intervention compared with occupational Gibbon B. Little V. (1995) Improving stroke care
therapists. What is the difference ournal of through action research. ournal of Clinical Nursing
Advanced Nursing , 98-105. , 93-100.
Booth J., Hillier V.F., Waters K.R. Davidson I. (2005) Hammersley M. Atkinson P. (2007) Ethnography:
Effects of a stroke rehabilitation education pro- Principles in Practice, 3rd edn. Routledge, New
gramme for nurses. ournal of Advanced Nursing , York.
465-473. Hearn M. Michelson G. (2006) Time. In Rethin ing
Bukowski L., Bonavolonta M., Keehn M.T. or : Time, Space, and Discourse (Hearn M.
Morgan K.A. (1986) Interdisciplinary roles in Michelson G., eds), Cambridge niversity Press,
stroke care. Nursing Clinics of North America , Cambridge.
359-374. Henderson V. (1980) Preserving—the essence of nursing
Burton C. (1999) An exploration of the stroke coordina- in a technological age. ournal of Advanced Nursing
tor role. ournal of Clinical Nursing , 535-541. , 245-260.
Burton C.R. (2000) A description of the nursing role in Hickey J.V. (2003) The Clinical Practice of Neurologi-
stroke rehabilitation. ournal of Advanced Nursing cal and Neurosurgical Nursing, 5th edn. Lippincott
, 174-181. Williams Wilkins, Philadelphia.
Canadian Stroke Network (2007) About Stro e. Hill M.D. (2002) Stroke units in Canada. Canadian
Retrieved from http://www.canadianstrokenetwork. Medical Association ournal (6), 649-650.
ca/eng/about/aboutstroke.php on 29 September Hill M.D., Barber P.A., Demchuk A.M., Sevick R.J.,
2007. Newcommon N.J., Green T. Buchan A. (2000)
Canadian Stroke Network and the Heart and Stroke Building a brain attack team to administer
Foundation of Canada: Canadian Stroke Strategy thrombolytic therapy for acute ischemic stroke.
(2006). Canadian Best Practice Recommendations for Canadian Medical Association ournal (11),
Stro e Care. Canadian Stroke Network and the Heart 1589-1593.
and Stroke Foundation of Canada, Ottawa. Indredavik B., Bakke F., Solberg R., Rokseth R.,
Cifu D. . Stewart D.G. (1999) Factors affecting Haheim L.L. Holme I. (1991) Benefit of a stroke
functional outcome after stroke: A critical review of unit: a randomized controlled trial. Stro e ,
rehabilitation interventions. Archives of Physical 1026-1031.
Medicine and Rehabilitation , S35-S39. Jorgensen H.S., Nakayama H., Raaschou H.O.,
Cott C. (1998) Structure and meaning in multidisci- Vive-Larson J., Stoier M. Olsen T.S. (1995a)
plinary teamwork. Sociology of ealth and Illness , Outcome and time course of recovery in stroke.
848-873. Part I: Outcome. The Copenhagen Stroke Study.
Davies C.A. (1999) Re e ive Ethnography: A uide Archives of Physical Medicine and Rehabilitation ,
to Researching Selves and thers. Routledge, 399-405.
London. Jorgensen H.S., Nakayama H., Raaschou H.O.,
Elliott A. (1999) The specialist nurse in rehabilitation. Vive-Larson J., Stoier M. Olsen T.S. (1995b)
In Rehabilitation Nursing (Davis S. O’Connor S., Outcome and time course of recovery in stroke.
eds), Edinburgh, Bailliere Tindall, pp. 231-243. Part II: outcome. The Copenhagen Stroke Study.
APPENDIX A Understanding Nursing on an Acute Stroke Unit: Perceptions of Space, Time and Interprofessional Practice 525

Archives of Physical Medicine and Rehabilitation , Rothberg J.S. Waggonerm R.W. (1972) Stroke
406-412. rehabilitation. Stro e , 375-407.
Kalra L. (1994) The in uence of stroke unit rehabilita- Pollard K., Sellman D. Senior B. (2005) The need for
tion on functional recovery from stroke. Stro e , interprofessional working. In Interprofessional
821-825. or ing in ealth and Social Care: Professional
Kalra L., Dale P. Crome P. (1993) Improving stroke Perspectives (Barrett G., Sellman D. Thomas J.,
rehabilitation: a controlled study. Stro e , eds), Palgrave MacMillan, New York, NY,
1462-1467. pp. 7-17.
Kirkevold M. (1997) The role of nursing in the Pound P. Ebrahim S. (2000) Rhetoric and reality in
rehabilitation of acute stroke patients: toward a stroke patient care. Social Science and Medicine ,
unified theoretical perspective. Advances in Nursing 1437-1446.
Science , 55-64. Sorrells-Jones J. (1997) The challenge of making it real:
Milligan R.A., Gilroy J., Katz K.S., Rodan M.F. interdisciplinary practice in a seamless organiza-
Subramanian K.N. (1999) Developing a shared tion. Nursing Administration uarterly , 20-30.
language: interdisciplinary communication among Spradley J. (1979) The Ethnographic Intervie . Holt,
diverse health care professionals. olistic Nurse Rinehart and Winston, New York.
Practitioner , 47-53. Spradley J. (1980) Participant bservation. Holt,
Morse J.M. Field P.A. (1995) ualitative Research Rinehart and Winston, New York.
Methods for ealth Professionals, 2nd edn. Sage Stroke nit Trialists’ Collaboration (2007) Organized
Publications, Thousand Oaks. inpatient (stroke unit) care for stroke. Cochrane
Myco F. (1984) Stroke and its rehabilitation: The Database of Systematic Revie s (4) CD000197.
perceived role. ournal of Advanced Nursing , Teasell R., Foley N., Bhogal S.K. Speechley M.
429-439. (2007) Evidenced based revie of stro e rehabilita-
Neatherlin J.S. Prater L. (2003) Nursing time and tion: the ef cacy of stro e rehabilitation. Retrieved
work in an acute rehabilitation setting. Rehabilitation from http://www.ebrsr.com/modules/module5.pdf on
Nursing , 186-190. 207. 29 January 2008.
O’Connor S.E. (1993) Nursing and rehabilitation: the Thorn S. (2000) Neurological rehabilitation nursing: a
interventions of nurses in stroke patient care. ournal review of the research. ournal of Advanced Nursing
of Clinical Nursing , 29-34. , 1029-1038.
O’Connor S.E. (2000a) Mode of care delivery in stroke lrich R., uan ., imring C., Joseph A.
rehabilitation nursing: a development of Kirkevold’s Choudhary R. (2004) The role of the physical
unified theoretical perspective of the role of the environment in the hospital of the st century: A
nurse. Clinical Effectiveness in Nursing , 180-188. once-in-a-lifetime opportunity. Report to The Center
O’Connor S.E. (2000b) Nursing interventions in stroke for Health Design for the Designing the 21st Century
rehabilitation: a study of nurses’ views of their Project. Retrieved from http://www.rwjf.org/files/
pattern of care in stroke units. Rehabilitation Nursing publications/other/RoleofthePhysicalEnvironment.pdf
, 224-230. on 1 June 2008.
Ovretveit J. (1997) How to describe interprofessional V h kangas P., Noro A. Finne-Soveri H. (2008)
working. In Interprofessional or ing for ealth and Daily rehabilitation nursing increases the nursing time
Social Care (Ovretveit J., Mathias P. Thompson T., spent on residents. International ournal of Nursing
eds), Macmillan, Basingstoke, pp. 9-33. Practice , 157-164.
Patmore G. (2006) Time and work. In Rethin ing or : Waters K.R. Luker K.A. (1996) Staff perspectives on
Time, Space, and Discourse (Hearn M. Michelson the role of the nurse in rehabilitation wards for
G., eds), Cambridge niversity Press, Cambridge, elderly people. ournal of Clinical Nursing ,
pp. 21-38. 105-114.
Payne M. (2000) Team or in Multiprofessional Care. Weinberg D.B. (2003) Code reen: Money-Driven
Macmillan Press, Basingstoke. ospitals and the Dismantling of Nursing. Cornell
Perry L., Brooks W. Hamilton S. (2004) Exploring niversity Press, Ithaca, NY.
nurses’ perspectives of stroke care. Nursing Standard World Health Organization (2004) The atlas of heart
, 33-38. disease and stro e. Retrieved from http://www.who.
Peszczynski M., Benson F., Collins M.M., Darley F.L., int/cardiovascular diseases/resources/atlas/en/ on 14
Diller L., Greenhouse A.H., Katzen F.P., Lake L.F., January 2007.
AP PENDIX B

Decisional Involvement of Senior Nurse Leaders in

Canadian Acute Care Hospitals
Carol A. Wong, RN, PhD • Associate Professor • Arthur Labatt Family School of Nursing • Faculty of Health Sciences •
The University of Western Ontario • London, Ontario

Heather Laschinger, RN, PhD • Professor and Associate Director Nursing Research • Arthur Labatt Family School of
Nursing • Faculty of Health Sciences • The University of Western Ontario • London, Ontario

Greta G. Cummings, RN, PhD • Associate Professor • Faculty of Nursing • University of Alberta • CIHR New
Investigator and AHFMR Population Health Investigator • Edmonton, Alberta

Leslie Vincent, MSc(A), RN • Senior Vice President Patient Services and Chief Nursing Executive • Mount Sinai Hospital •
Toronto, Ontario

Patty O’Connor, MSc(A), CHE • Associate Director of Nursing • Neurosciences • McGill University Health Centre •
Montreal, Quebec

AIM The aim of the present study was to describe the CONCLUSIONS Senior nurse leaders contribute to organi-
scope and degree of involvement of senior nurse leaders zational processes in healthcare organizations that are
(SNLs)in executive level decisions in acute care organiza- important for nurses and patients,through their participa-
tions across Canada. tion in decision-making at the senior team level.

BACKGROUND Significant changes in SNL roles including IMPLICATIONS FOR NURSING MANAGEMENT Findings
expansion of decisionmaking responsibilities have occurred may be useful to current and future SNLs learning to shape
but little is known about the patterns of SNL the nature and content of information shared with CEOs
decision-making. particularly in the area of professional practice issues.

METHODS Data were collected by mailed survey from 63 KEYWORDS acute care,decision-making,hospitals,influ-
SNLs and 49 chief executive officers (CEOs)in 66 healthcare ence,leadership,nurse executive
organizations in 10 Canadian provinces. Regression analyses
were used to examine whether timing,breadth of content INTRODUCTION
expertise and the number of decision activities predicted
SNL decision-making influence and quality of decisions. Healthcare restructuring in the 1990s in Canada
and the nites States contributed to significant
RESULTS Breadth of content expertise and number of changes in senior nurse leader (SNL) roles includ-
decision activities with which the SNL was involved were ing expansion of their decision-making responsi-
significant predictors of decision influence explaining 22% bilities (Murray et al 1998, Mass et al 2006,
of the variance in influence. Overall,CEOs rated SNL Smith et al 2006). In some organizations nurse
involvement in decision-making higher than the SNL. executives were added to senior executive teams
and in others, their scope of participation in orga-
From the ournal of Nursing Management, 2010 nizational decisions related to budget, strategic
18:122-133. © 2010 Blackwell Publishing Ltd. planning, quality of care and a host of challenging
526
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 527

organizational issues greatly increased. Similarly, of SNL participation in decision-making (PDM)

healthcare reform in the nited Kingdom and at the senior executive level of healthcare organi-
other European countries created role changes zations and in particular, the consequences of
and new opportunities for nurse leaders in health organizational changes on nurse executive
care organizations (Fedoruk 2000, Filkins 2003, decision-making. The present study was part of a
Kirk 2008). national survey to describe the profile of nursing
New governance structures and organizational leadership structures in Canada and to assess rela-
models radically changed disciplinary leadership tionships among personal and structural factors,
structures, particularly in nursing (Havens 1998, processes and outcomes pertaining to nurse
2001, Baumann et al 2001, Shannon French leaders’ work post restructuring (Laschinger
2005, Sharp et al 2006). Specifically, the pro- et al 2008).
gramme management structure and regionalized Our aim was to describe SNL decision-making
healthcare systems were implemented across processes in terms of the scope and degree of
Canada (Leatt et al 1994, Smith et al 2006). In their involvement in strategic and tactical deci-
regionalization, responsibility for a wide span of sions at the executive management level in orga-
health services, frequently spanning community, nizations across Canada. We define the SNL as
long-term care and acute care services, were the nurse who holds the most senior nursing lead-
organized under one governing body with con- ership position in the organization with direct
solidation of authority which was previously dis- responsibility for nursing. While titles such as
tributed among many organizations. A goal of this chief nurse executive, nurse executive or senior
restructuring was greater integration of services nurse leader are often used interchangeably, for
with reduced duplication and overlap. Within clarity we use the SNL term throughout this
programme management structures in hospitals, paper. Specifically, we examined whether timing,
distinct professional departments were elimi- breadth of content expertise and the number of
nated, services were organized around popula- decision activities predicted SNL perceptions
tions of patients and care was provided by of their decision-making in uence and the quality
multidisciplinary teams. Some claim that these of management decisions made by the senior
changes provided opportunities for nurse leaders team. We also included a description of how
to demonstrate their leadership skills and play a chief executive officers (CEOs) perceive SNL
greater role in decision-making within new inter- decision-making.
disciplinary and more broad-based programme
structures. Others argued that these changes LITERATURE REVIEW
diminished authority and communication links
between senior nurse leaders and other nursing SNL Role Changes with Restructuring
personnel and deprived nurses of disciplinary Substantial changes in the healthcare system have
leadership representation at the policy making contributed to new role expectations, higher
level (Shamian Lightstone 1997, Clifford 1998, knowledge requirements and increased responsi-
Canadian Nursing Advisory Committee 2002). bility and accountability for nurse leaders includ-
Several sources attest to the dissatisfaction that ing: quality and effective coordination of patient
nurses experienced with the reduction in nurse services managing many clinical areas with a
leaders at all levels of organizations with restruc- broadened span of control operating merged
turing (Baumann et al 2001, Canadian Nursing facilities and decentralized structures and
Advisory Committee 2002). However, surpris- decision-making in finance, human resources and
ingly little is actually known about the patterns quality and safety of patient services across the
528 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals

continuum of care (Duffield et al 2001, Klein- The importance of SNLs participation in orga-
man 2003, penieks 2003, Anthony et al 2005, nizational decision-making is acknowledged
Arnold et al 2006). In many cases restructuring in the literature (Fedoruk 2000, Clancy 2003).
changes provided opportunities for nurse leaders However, most of the empirical evidence in this
to demonstrate their leadership skills and play a area has focused on participation of physicians
greater role in decision-making within the new and registered nurses (RNs) in strategic decision-
multidisciplinary programme structures (Clancy making (Ashmos McDaniel 1991, Havens
2003, Thorman 2004, Kirk 2008). As an inte- Laschinger 1997, Ashmos et al 1998, Anderson
grated member of the senior leadership team, the McDaniel 1999). Ashmos and McDaniel
SNL has the opportunity to in uence team (1991) determined that the greater the intensity
members by ensuring that patient care and nursing (number and type of decision activities) of par-
practice perspectives are voiced when decisions ticipation in decision-making (PDM) by profes-
are being made that affect organizational direc- sionals, the more likely that they will be perceived
tions, quality management and resource use. The as having an in uence on decisions. tilizing a
recent Sharp et al (2006) findings on the effects survey method that included questions on the
of service line (similar to programme) manage- timing, breadth and intensity of participation in
ment implementation in .S. Veterans Health decision-making to capture overall decision
Administration sic hospitals supported many of involvement, they examined the effect of clinical,
the positive aspects of SNL role changes described professional and middle manager participation on
previously. However, SNLs in pure service line hospital performance (Ashmos et al 1998). The
organizations without a discipline-based nursing participation of medical and other clinical profes-
service reported decreased direct supervision of sionals (e.g. nurses) in organizational strategic
nurses and challenges in achieving consistency in decisions was associated with reduced hospital
quality of nursing care. costs while there was no such effect for middle
manager participation. Adding decision in uence
SNL Role in Organizational to the Ashmos et al (1998) measure, Anderson
Decision-Making and McDaniel (1998) showed that administrators
Decision-making research in nursing has focused in nursing homes perceived greater in uence of
primarily on the study of clinical nurses (Orovio- RNs in decisions when RNs were more involved
goicoechea 1996, Thompson 1999, Lauri et al in decision activities. They also showed that
2001). Of the published literature on nurse leader increased RN participation in decision-making
decision-making, there is little coherence in was associated with improved resident outcomes
topics such as, risk propensity (Smith Fried- in nursing homes (Anderson McDaniel 1999).
land 1998), ethical decision-making (Fonville Thus, these studies suggest that decision-making
2002, Berggren Severinsson 2003), manager involvement can be measured and that there is a
role in facilitating staff participation in decision- connection between involvement in decision
making (Krairiksh 2000), middle manager activities, perceived in uence over decisions and
involvement in organizational strategic decision- organizational outcomes.
making (Ashmos et al 1998) and personality A few studies focused on the integration of
type and decision-making styles (Freund 1988). nurse leaders in executive level organizational
Only a small body of research focused on SNLs’ decision-making during the healthcare restructur-
organizational decision-making in uence (Wangs- ing era of the 1990s. In a survey of 115 SNLs in
ness 1991, Havens 1998, Banaszak-Holl et al Pennsylvania acute care hospitals, Wangsness
1999, Dwore et al 2000). (1991) determined that most participants had
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 529

considerable decision-making authority at the Langton 2003). More specifically, Mintzberg

departmental but very little at the organizational et al (1976) emphasized that a decision in orga-
level. Havens (1998) studied the involvement of nizations is a specific commitment to action
the nurse executive in .S. hospital governance (usually a commitment of resources)’ (p. 246).
and policy making in 1990 and again in 1996, Decision-making has been described as a complex
reporting little change in SNL involvement at cognitive process that involves critical thinking,
senior executive levels over that time. Another judgement, evaluation and memory (Oetjen et al
study documenting the integration of SNLs into 2008). Generally, decision-making occurs as a
executive-level administration of 53 acute care response to problems or opportunities and thus,
hospitals in tah, found that 80 of the SNLs decision-makers must address a gap between the
perceived they were more involved in hospital current state and some desired future state.
activities and decisions (Dwore et al 2000). The theoretical framework (Figure B-1) for
There is evidence that the scope and intensity this study was adapted from the work of Ashmos
of SNL involvement in strategic decision-making and McDaniel (1991) and Anderson and
is related to their perceptions of in uence in the McDaniel (1998) who developed an approach to
organization (Banaszak-Holl et al 1999, Wells examining health professionals’ participation in
et al 1999). Banaszak-Holl et al (1999) exam- strategic decision-making in healthcare organiza-
ined the role of SNLs in organizational decision- tions using information processing and complex-
making within Veteran’s Affairs Medical Centres ity science theories (Anderson McDaniel 1998,
(VAMCs) in the nited States during the mid- Ashmos et al 1998). According to information
1990s. Restructuring led to the inclusion of SNLs processing theory, increased participation of mul-
on the executive management team as an equal tiple stakeholders increases both the amount of
collaborator with the medical chief of staff and information and the ability to deal with it (Ashmos
other senior directors. All SNLs and their man- et al 1998). Participation is a mechanism for the
agement team colleagues were surveyed in 84 exchange of information’ in decision-making pro-
VAMCs using an adapted form of the Ashmos cesses (Ashmos McDaniel 1991, p. 386) and
and McDaniel (1991) decision-making instru- this capacity can be altered by changing the par-
ment. Senior nurse leaders perceived they brought ticipants but also by changing the . . . timing,
greater breadth of expertise to decisions and par- scope, and formalization of the process’ (p. 388).
ticipated in more decision activities than non- Bringing participants into a decision process
nurse colleagues’ perceptions of the SNL role. early expands the capacity of the organization to
There were no significant differences in how process information as does widening the scope
SNLs and their colleagues rated SNL in uence of participation in various decision activities such
over decisions or quality of decisions made by as raising issues, clarifying problems and gener-
management teams, except that SNLs consis- ating alternatives. From the complexity perspec-
tently ranked decision quality lower than their tive, diversity and expansion of those involved in
non-nurse colleagues. Banaszak-Holl et al (1999) organizational decision-making increases internal
did not analyse the effect of decision timing, complexity and also enhances an organization’s
breadth or number of activities on in uence or ability to create meaning through increased use
perceived quality of decisions. of connections’ among stakeholders (Anderson
McDaniel 1999, p. 8). Such connections also
Theoretical Framework enable interactions among people that may
Simply stated, a decision is defined as a choice broaden and deepen interpretations of events and
made from two or more alternatives (Robbins actions. Leaders can increase the speed or degree
530 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals

CNEs as Sources of
Information

Scope of
Participation:

• Timing of
Decision Context— Involvement
Decisions in these • Breadth of Content
Areas: Expertise
Planning
Decision Quality of
Operational &
Influence Decisions
Management
Clinical CNEs as Processors of
Resource Information
Professional Practice
Intensity of
Participation:
Decision Activities
• Raising the Issue
• Clarifying the Issue
• Generating Alternatives
• Evaluating Alternatives
• Choosing Alternatives

FIGURE B-1 Participation in organizational decision making.

of information ow by increasing the number of 1999). The greater the scope and intensity of SNL
people involved and by expanding the number of PDM, the greater the likelihood that they and
decision activities in which they are involved others perceive them as having an in uence on
using both formal and informal mechanisms for decisions. Last, we propose that decision-making
interaction. in uence is related to the quality of final manage-
In our framework, PDM by SNLs in executive ment decisions reached.
management teams is viewed as creating new
organizational connections and mechanisms for Hypotheses
exchanging information and enriching interpreta- The scope (timing and breadth) and intensity
tion of issues that ultimately in uence the quality (number of decision activities) of SNL partici-
of management decisions. The scope of decision- pation in executive decision-making processes
making is enhanced by involving SNLs at the positively predicts the degree of SNL decision
beginning of decision-making stages (timing) and in uence.
the breadth of content expertise is expanded by SNL decision in uence positively predicts per-
their clinical and professional knowledge. The ceived quality of operational management
intensity of PDM is a function of the number and decisions.
range of decision-making activities involving
SNLs. Any decision-making process entails METHODS
several different fundamental information pro-
cessing actions from raising issues, clarifying Sample and Data Collection Procedures
problems, generating and evaluating solutions to To obtain a comprehensive description of nurs-
making a final choice (Anderson McDaniel ing management structures in Canada, Academic
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 531

Health Centres (AHC) and community hospitals operational management decisions that deal
(CH) in 10 provinces were selected as data with the day-to-day operation of the orga-
sources. An AHC is a health care facility that nization (excluding direct patient care)
participates in medical research and in teaching clinical care decisions are policy or admin-
undergraduate and graduate medical students. All istrative issues related to the provision of
of the AHCs that focused on acute care in each direct patient care
province were selected to participate in this study. resource decisions address fiscal issues
For each of the AHCs, a CH with more than 100 (including budgeting, revenues and spend-
beds was randomly selected from a complete list ing) and human resources and
of CHs in the Health Authority/geographic region professional practice decisions relate to
of each AHC. Ethics approvals were obtained standards of nursing practice, discipline,
from a university ethics review board as well as education and research issues (Figure B-2).
from specific organizations in the study. Data The last decisional area, professional practice,
were collected from SNLs and CEOs in 28 aca- was added to the Banaszak-Holl et al (1999)
demic health centres and 38 community hospitals scale because the clinical decision type did not
in 10 Canadian provinces. All SNLs (n = 66) and fully represent the range of decisions relevant to
CEOs (n = 66) were surveyed by mail. Of the nursing as we believe that many organizational
original 132 surveys to SNLS and CEOs, 112 decisions made at the top management level
surveys were returned, for an overall response affect the professional practice of nurses. For
rate of 84.8 (Table B-1). each of these five areas of decision-making, we
asked survey participants to rate the SNL’s
Instrument involvement in senior organizational level
We used the participation in Strategic Decision- decision-making over the past 6 months accord-
Making Scale (Banaszak-Holl et al 1999) to ing to the scope of participation (consists of
measure SNL decision-making processes. This timing and the breadth of content expertise),
Likert-type instrument was adapted from a survey intensity of participation (the number of decision
developed by Ashmos et al (1998). The SNLs’ activities and the mechanisms used), the SNL’s
participation in decision-making processes was in uence over decision and the quality of man-
evaluated according to five different types of stra- agement decisions made.
tegic decisions commonly considered by the Timing of decisional involvement was defined
executive level of the organization: as the time point a SNL most often became
planning decisions defined as the formal involved in the decision-making process
process of developing organizational goals (1 = beginning of process 5 = end of the
and strategies process). No precise timeframes (e.g. days,
weeks or months) for decisions are included in
the instruments as there is considerable variation
TABLE B-1 in the timeframes for organizational level deci-
COMPARISON OF RESPONSE RATES sions. For example, strategic planning decisions
TOTAL AHC CH may take months whereas some clinical deci-
n % n % n % sions may be required within days. Breadth was
SNL 63 95.5 30 100.0 33 91.7 defined as the scope of expertise the SNL most
CEO 49 74.2 23 82.1 26 68.4 often offered in the decision-making process (1
AHC, academic health centre; CEO, chief executive officer; CH,
= narrow/single area of expertise 5 = broad/
community hospital; SNL, senior nurse leader. many different areas of expertise). Number of
532 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals

Section 3: Please respond to the questions below in reference to each of the following
types of decisions.

PLANNING DECISIONS – The formal process of developing organizational goals and

methodologies for achieving those goals. Activities may include addressing critical
issues, significant events and major trends, both internal and external, that will impact
the overall direction of the hospital. Examples: Development of a 5-year plan, addition of
a cardiac surgery program, conversion from inpatient to outpatient surgery.

OPERATIONAL MANAGEMENT DECISIONS – Significant decisions dealing

with the day-to-day operational management of the hospital; excludes items involving
direct patient care. Examples: Review/approval of major projects (e.g. interior design
renovations), communications deficits/morale issues, review/approval of space for
program expansion within the existing facility.

CLINICAL DECISIONS – Policy or administrative issues related to the provision of

direct patient care. Examples: Plans for addressing a backlog of surgical cases, individual
patient care problems not solvable at the unit/service level, clinic waiting time.

RESOURCE DECISIONS – Administrative and policy issues dealing with fiscal (e.g.
income, budgeting, expenditure of monies) and human (e.g. employee management)
resources. Examples: Review/approval of hospital and service-wide budgets, total
number and distribution of FTEs, nurse shortage in the ICU.

PROFESSIONAL PRACTICE DECISIONS – Policy or administrative issues

relating to decisions about standards of nursing practice, disciplinary issues, quality of
patient care, staffing, evidence-based practice, education, and research.

FIGURE B-2 Excerpt from SNL decision-making survey.

decision activities was defined as the total count decisions was measured by rating their level of
of decision activities in which the SNL was agreement (1 = do not agree 5 = strongly agree)
involved including raising issues, clarifying with six items pertaining to: compatibility of
problems, generating alternatives, evaluating decisions with existing constraints and policy,
alternatives and choosing options. In addition, advantageous timing of decisions, appropriate
SNLs were asked to identify the most frequently use of information, balance of risks and rewards
used mechanisms through which the SNL is and whether decisions created con ict of interest
involved in the decision-making process 1 = (reverse coded item). We evaluated quality of
meeting with top management 2 = established management decisions for all decision areas
standing committees 3 = task forces/ad hoc rather than by each of the five decision areas.
committees 4 = meetings with top management Banaszak-Holl et al (1999) reported Cron-
team members (excluding CEO) 5 = informal bach alphas only for quality of decisions across
meetings with CEO . Decision mechanisms were the original four decisions areas and these ranged
used for descriptive purposes only and not from 0.74 to 0.81. In our study, the Cronbach
included in the regression analysis. Perceptions alpha for the six quality scale items was
of SNL in uence over final decisions in the five 0.68. The alphas for timing, breadth, number of
decision areas were measured according to: 1 = decision activities, mechanisms and in uence
no in uence and 5 = great deal of in uence. Per- subscales ranged from 0.83 to 0.90 (see Table
ceived quality of operational management B-5).
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 533

All SNLs had a nursing background with univer-

Data Analyses sity degrees in nursing primarily at master’s level
uantitative survey data were analysed using (57.6 ). In addition, a large percentage of SNLs
SPSS 16.0 for Windows (SPSS Inc., Chicago, IL, had earned non-nursing degrees (42 ). Twenty-
SA 2005). one per cent of SNLs had obtained general or
Descriptive statistics were analysed by level of business-related masters degrees whereas others
management (SNL and CEO) and correlation had obtained Masters in Health Administration
analyses were used to examine relationships (MHA) (14 ), PhD (8 ) and Masters in Educa-
among the decision-making variables in our tion (6 ). Thirty-one per cent of SNLs completed
model. Paired t-tests were applied to assess dif- the annual 3-week nurse executive development
ferences in mean scores for SNLs and CEOs. programme sponsored by the Wharton School of
Finally, hierarchical multiple regression was used Business ( niversity of Pennsylvania) and the
to test study hypotheses using SNL mean scale Johnson Johnson Family of companies. Gradu-
scores of the decision-making variables. ates of this programme receive the designation as
a Wharton Fellow.
RESULTS Most CEOs surveyed (n = 49) were male
(75.5 ). The average age of these senior execu-
Demographics tives was 52 years and on average had been in
Demographics of the sample are illustrated in their present position for 5.3 years. The CEOs
Table B-2. A notable finding was the high average most commonly reported their highest education
age (50.4 years) of SNLs, suggesting the urgent as a masters degree: MHA (36.7 ) and Masters
need for succession planning to ensure the future in Business Administration (MBA) (20.4 ).
of nursing leadership. Overall, current SNLs were Other common responses were a combination of
very experienced individuals as over 80 of an MD and an MBA (18.2 ), other Masters
senior nurse leaders had at least 15 years of man- degrees (6.1 ) and a combination of an MBA
agement experience and all had at least 5 years. and an MHA (4.1 ). One respondent reported

TABLE B-2
DEMOGRAPHICS
SENIOR NURSE LEADER CEO
M SD M SD

Age 50.4 4.8 51.97 6.01

Years management experience 20.8 6.7
Years in role 3.7 3.3 5.32 4.10
n % n %

GENDER
Male 0 0 37 75.5
Female 63 100 11 22.4

HIGHEST LEVEL 2.1

Diploma 2 3.4 1 10.4
Baccalaureate 19 32.2 5 8
Masters 34 57.6 40 3.3
PhD 4 6.8
Other 2 4.2
534 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals

having a Masters of Science in Nursing (MScN) TABLE B-3

as the highest earned degree and one reported an MEANS AND STANDARD DEVIATIONS OF
MD as their highest earned degree. DECISIONAL INVOLVEMENT SUBSCALES
SNL (n = 63) CEO (n = 49)
DECISIONAL INVOLVEMENT M(SD) M(SD)
Descriptive Results
TIMING*
SNL Perceptions of Decisional Involvement Planning 1.49 (1.10) 1.43 (0.89)
Operational 2.25 (1.34) 1.70 (1.13)
Means and standard deviations for SNL and Clinical 2.23 (1.33) 1.74 (1.16)
CEO ratings of the items and subscales of the Resource 1.85 (1.17) 1.52 (0.86)
decision-making instrument are included in Prof practice 1.83 (1.95) 1.52 (1.17)
Average all decision types 1.93 (0.95) 1.58 (0.85)
Table B-3. SNLs were involved near the begin-
ning of the decision-making process for all deci- BREADTH OF CONTENT EXPERTISE
sion areas. The earliest involvement was reported Planning 4.16 (1.16) 4.11 (0.89)
for planning decisions whereas the latest involve- Operational 3.89 (1.17) 4.21 (0.88)
Clinical 4.03 (1.15) 4.19 (0.92)
ment of the SNL was reported for operational Resource 4.15 (1.08) 4.17 (0.79)
and clinical decisions. SNLs reported offering a Prof practice 4.46 (1.03) 4.28 (1.06)
broad range of expertise in executive decisions Average all decision types 4.08 (0.98) 4.19 (0.76)
with professional practice decisions rated the NUMBER OF DECISION ACTIVITIES
broadest range of expertise and operational deci- Planning 4.25 (1.36) 3.96 (1.56)
sions received the lowest mean for range of Operational 3.62 (1.50) 3.98 (1.55)
Clinical 3.40 (1.56) 3.83 (1.66)
expertise offered. The highest mean number of Resource 4.07 (1.37) 3.91 (1.64)
decision activities was for professional practice Prof practice 4.15 (1.41) 4.00 (1.62)
and the lowest was reported for clinical Average all decision types 3.88 (1.19) 3.94 (1.54)
decisions. DECISION INFLUENCE
Not surprisingly, the types of decision domains Planning 4.15 (1.08) 4.26 (0.74)
(planning, operational, clinical, resources or pro- Operational 4.03 (1.14) 4.36 (0.71)
fessional practice issues) require different deci- Clinical 4.05 (1.06) 4.66 (0.60)
Resource 4.21 (0.91) 4.28 (0.65)
sion mechanisms (Table B-4). Both CEOs and Prof practice 4.69 (0.83) 4.72 (0.54)
SNLs rated these forums quite similarly. Plan- Average all decision types 4.23 (0.76) 4.46 (0.52)
ning, operational and resource matters were pri-
QUALITY OF DECISIONS
marily dealt with in meetings with the top Compatible with constraints 4.11 (0.87)
management team (as the decision mechanism), Timing for max coverage 3.89 (0.98)
with clinical and professional practice matters Optimal information 3.54 (0.85)
Conflict of interest 3.48 (1.25)
included to lesser degrees in these forums. Stand- Balance risk and reward 3.70 (0.76)
ing committees were used most frequently for Basis for implementation 3.97 (0.71)
professional practice, operational and clinical Average all decision types 3.80 (0.59)
decisions and task forces were used most often *Note: Low score = early involvement.
for clinical decisions. In terms of in uence (Table
B-3), SNLs reported having a large amount of
in uence over final decisions across all types of decisions was rated as moderate and the highest
decisions. The highest level of in uence was agreement was for decisions compatible with
reported for decisions regarding professional existing constraints, policies, etc.’ and the lowest
practice and the lowest in uence was reported agreement was for implementing decisions
for operational decisions. Overall quality of caused a con ict of interest’.
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 535

TABLE B-4
FREQUENCY, MEANS AND STANDARD DEVIATIONS OF DECISION MECHANISMS
SNL CEO
DECISION AREAS AND MECHANISMS M(SD) n (%) M(SD) n (%)

Planning 1.61 (1.35) 1.37 (1.0)

Meeting with top management 47 (74.6) 39 (79.6)
Standing committees 3 (4.8) 2 (4.1)
Task forces 1 (1.6) 2 (4.1)
Informal meetings with management 3 (4.8) 1 (2.0)
Informal meetings with CEO 6 (9.5) 2 (4.1)
Operational 1.77 (1.11) 1.74 (1.16)
Meeting with top management 32 (50.) 27 (55.1)
Standing committees 15 (23.0) 12 (24.5)
Task forces 6 (9.5) 2 (4.1)
Informal meetings with management 4 (6.3) 2 (4.1)
Informal meetings with CEO 2 (3.2) 3 (6.1)
Clinical 2.33 (1.23) 2.07 (1.06)
Meeting with top management 18 (28.6) 16 (32.7)
Standing committees 16 (25.4) 17 (34.7)
Task forces 16 (25.4) 9 (18.4)
Informal meetings with management 6 (9.5) 2 (4.1)
Informal meetings with CEO 4 (6.3) 2 (4.1)
Resource 1.90 (1/42) 1.57 (1.11)
Meeting with top management 39 (61.9) 33 (67.3)
Standing committees 7 (11.1) 7 (14.3)
Task forces 4 (6.3) 1 (2.0)
Informal meetings with management 4 (6.3) 3 (6.1)
Informal meetings with CEO 7 (11.1) 2 (4.1)
Professional practice 2.15 (1.67) 1.91 (0.99)
Meeting with top management 18 (28.6) 16 (32.7)
Standing committees 30 (47.6) 24 (49.0)
Task forces 4 (6.3) 2 (4.1)
Informal meetings with management 4 (6.3) 2 (4.1)
Informal meetings with CEO 5 (7.9) 2 (4.1)
Overall Mean 1.95 (0.98) 1.73 (0.89)

CEO Perceptions of the SNL ties was for professional practice decisions and
Decisional Involvement the lowest was reported for clinical decisions. As
CEOs described the SNLs as being involved near for decision mechanisms, CEOs reported the
the beginning of the decision-making process for SNLs in their organization were more likely to
all decision areas. The earliest involvement was utilize meetings with top management, especially
reported for planning decisions whereas the latest for planning, operational and resource decisions.
involvement of the SNL was reported for clinical SNLs were more likely to use standing commit-
decisions. CEOs reported that SNLs offered a tees for professional practice decisions and were
broad range of expertise in top-level decisions. least likely to utilize informal meetings with man-
Professional practice decisions rated the broadest agement or CEO and task forces. In terms of
range of expertise and planning decisions received in uence over final decisions, CEOs reported that
the lowest mean for range of expertise offered. SNLs had a large amount of in uence over final
The highest mean for number of decision activi- decisions that were reached across all types of
536 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals

decisions. The highest level of in uence was 0.312, t = 2.524, P = 0.015) and number of deci-
reported for decisions regarding professional sion activities (β = 0.364, t = 2.997, P = 0.004)
practice and the lowest level of in uence for plan- were significant predictors of in uence and thus,
ning decisions. Ratings of overall quality of man- the first hypothesis was partially supported. The
agement decisions were not included in the CEO second hypothesis was tested by entering mean
survey. Results of paired t-tests to compare SNL decision-making in uence as the predictor with
and CEO means of the four decision variables mean quality of management decisions as the
(timing, breadth of involvement, activities and dependent variable. Decision-making in uence
in uence) showed no statistically significant was not a significant predictor of the quality of
differences. management decisions, thus the second hypoth-
esis was not supported.
Correlations Among Major
Study Variables
Only SNL breadth of content expertise (r = 0.324, DISCUSSION
P < 0.01) and of decision-making activities (r = In the present study, we developed a model of
0.356, P < 0.01) were significantly correlated SNL participation in organizational decision-
with decision-making in uence and also with making based on the work of Ashmos et al
quality of management decisions (r = 0.219, P < (1998), Anderson and McDaniel (1998) and
0.05 for breadth r = 0.242, P < 0.05 for number Banaszak-Holl et al (1999). We found partial
of activities) (Table B-5). support for our hypothesis that SNL timing of
involvement, breadth of content expertise and
Test of Hypotheses number of decision activities in which they are
For hypothesis one, the predictor variables, SNL involved predicted their perception of in uence
mean scale scores for timing of involvement, in organizational decisions. Timing was not a sig-
breadth of expertise and number of decision nificant predictor whereas SNLs reported early
activities, were entered hierarchically with SNL involvement in most decision types. Although
mean scale decision-making in uence as the there is some indication in the healthcare litera-
dependent variable. Twenty-two per cent of the ture that increased involvement in organizational
variance in SNL decision-making in uence was decision-making by physicians and registered
explained by timing of involvement, breadth of nurses was associated with outcomes such as
content expertise and number of decision activi- lower costs in hospitals (Ashmos et al 1998) and
ties (R2 = 0.223, (3, 53) = 5.06, P = 0.004). improved resident outcomes in nursing homes
However, only breadth of content expertise (β = (Anderson McDaniel 1999), we found that

TABLE B-5
MEANS, STANDARD DEVIATIONS AND CORRELATIONS AMONG SNL TIMING, BREADTH OF CONTENT
EXPERTISE, NUMBER OF DECISION ACTIVITIES, DECISION INFLUENCE AND QUALITY OF DECISIONS
VARIABLE n M(SD) α 1 2 3 4

Timing of involvement 63 1.93 (0.95) 0.84 —

Breadth of content expertise 63 4.08 (0.98) 0.89 0.133 —
Number of decision activities 57 3.88 (1.19) 0.90 0.064 −0.022 —
Decision-making influence 63 4.23 (0.76) 0.83 0.022 0.324** 0.356** —
Quality of decisions 62 3.80 (0.59) 0.68 0.133 0.219* 0.242* 0.195
*P < 0.05, one-tailed. **P < 0.01, one-tailed.
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 537

SNL decision involvement (timing, breadth and CEOs do not differentiate professional
number of activities) and in uence in decision- practice decisions from operational ones to
making did not predict their perceived quality of the same degree as SNLs.
organizational decisions. Some differences in perceptions between
There were significant positive, albeit small, CEOs and SNLs were related to number of deci-
correlations between breadth of expertise and sion activities: CEOs reported more SNL activi-
number of activities and quality of decisions. ties for clinical and operational decisions whereas
Small sample size may have been an issue in why the SNLs reported a higher number of activities
we did not find a significant relationship between than the CEO for resource, planning and profes-
in uence and quality. Post hoc power analysis sional practice decisions. CEOs also rated SNL
showed power was only 0.33. Even although decision in uence consistently higher than the
SNLs perceived a significant in uence over deci- SNL did for all decision types. These positive
sions, there were likely many other factors outside comparative findings may indicate that CEOs
of the control of the SNLs and even the senior have considerable confidence in the decision-
leadership team such as government directives, making role of their respective SNLs. Wells
economic constraints or community reactions et al (1999) reported a similar finding in the
that ultimately in uenced the quality of deci- Banaszak-Holl et al (1999) study: directors’
sions. The measure of decision quality was a (equivalent to the CEO role) perceptions of SNL
subjective measure and included only SNL involvement were similar to SNLs’ self-ratings
ratings. Interestingly, Banaszak-Holl et al (1999) and often were more positive than SNLs were
reported that SNLs rated decision quality signifi- about their participation in decision-making.
cantly lower than did non-nurse members of the CEO perceptions of SNL decision-making may
senior leadership team in all decision categories provide some insights for current and future
but no means were reported so we could not SNLs learning to shape the nature and content
compare them with our decision quality findings. of information shared with CEOs, particularly in
nfortunately, we could not compare our results the area of professional practice issues and
to their findings as they did not report quality item decision-making.
or scale means. Concurrence between CEOs and SNLs about
In general, CEO ratings of SNL involvement the SNL role in executive decision-making is
in decision-making were higher than the SNL self both reassuring and critical to ensuring his/her
ratings. In all decision areas, CEOs rated SNL effectiveness. An essential responsibility of the
involvement earlier in the decision process than SNL is to effectively assess, plan, forecast and
SNLs reported. CEOs ratings of SNLs breadth of execute decisions based on the needs within
content expertise also were higher than SNL self- nursing, clinical departments and the patient pop-
ratings in four out of the five decision areas ulations served. But to be successful at this
SNLs rated themselves somewhat higher in pro- requires that the SNL is able to in uence other
fessional practice decision-making than did the decision-makers, particularly at the executive and
CEOs. It is possible that the depth of SNL involve- governing board levels. These findings suggest
ment in professional practice decision-making that despite the significant restructuring within
was not readily apparent to the CEO. This could Canadian healthcare organizations, CEOs have
be related to two possible factors: considerable trust in the leadership competencies
professional practice decision-making occurs of the SNL and view these persons as having a
more frequently within committees, which very high degree of in uence. This augers sic
may not include the CEO, or well for members of the nursing workforce, with
538 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals

regard to knowing that frontline clinical issues (M = 4.23 vs. 3.83) where our overall mean was
and concerns will be heard when conveyed by higher.
SNLs to the executive team. Our findings suggest that concerns about the
Overall, SNLs reported having a large span of impact of restructuring (i.e. regionalization of
in uence the breadth of knowledge and skills the care services, programme management model
SNL brings to the executive table, allows them to and elimination of traditional distinct nursing
significantly in uence not only clinical care, but departments) on SNL decision in uence may not
also organizational policy and strategic direc- be warranted. In the present study, the predomi-
tions. In fact, SNLs are in an ideal position to nant SNL role configuration was operational/line
show the linkages between different types of authority for clinical programmes with a direct
decisions and ensure that there is alignment report to the CEO in 84 of organizations so
between the clinical and business decision spheres concern about direct line responsibility for
in organizations. There is likely an overlap among nursing was not an issue. Traditional distinct
the different decision types. Decisions in organi- nursing departments were rare (20 ) and were
zations are rarely totally independent of other found primarily in uebec and in community
decisions (Mintzberg et al 1976, Oetjen et al hospitals. Our sample was an experienced group
2008). For example, strategic planning decisions of SNLs suggesting their decision in uence skills
determine organizational priorities and ultimately were well developed. The need for leader succes-
affect most other decision types. Resource deci- sion planning, that is, identifying and developing
sions are also central to most other decision types the next generation of nurse leaders is critical
and set the boundaries for what is possible within given the average age of SNLs in Canada. Con-
the clinical, human resources and professional scious efforts to prepare future nurse leaders in
practice areas. mentored decision-making activities such as
When we compared our findings with those of learning about organizational decision processes,
Banaszak-Holl et al (1999), we found substantial role shadowing, guided project work and leader
similarity in overall decision-making across deci- development programmes are important to
sion types. Some caution is required with these augment their decision-making confidence and
comparisons given that a new decision-type expertise.
domain, professional practice, was added in our The demands of 21st century healthcare envi-
study. In the earlier study, means for clinical deci- ronments are increasingly characterized as
sions in all the decision variables were generally complex, dynamic, unpredictable and somewhat
higher where professional practice decisions were resistant to traditional management solutions to
often rated higher than clinical for all decision problems (Huston 2008). Strong nursing leader-
variables. Some aspects of the clinical domain ship is required to create cultures of safety and
may be incorporated in SNLs responses to the healthier work environments that promote patient
professional practice domain although we con- safety, excellence in care and recruit and retain
strued these two decision areas to be different. staff. To meet these challenges SNLs need expert
Both the current study and VA results were similar decision-making skills guided by sound empirical
overall for timing of involvement (M = 1.93 vs. evidence, innovative thinking and effective com-
1.92, respectively) and breadth of content exper- munication strategies to involve executive team
tise (M = 4.08 vs. 3.90, respectively). Areas of members and other stakeholders in creating new
greatest difference were for number of decision responses to these challenges. SNL decision
activities (M = 3.88 vs. 4.31, respectively) where in uence for change is now required on many
VA mean was higher, and decision in uence levels beyond the executive team: at the staff
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 539

level by creating alignment for organizational in uence. Moreover, CEOs validated these find-
decisions, at the board level by providing inter- ings, rating SNL involvement in decision-making
pretation of quality concerns, at the community even higher than SNLs. We found support for our
level by raising awareness of health service issues contention that involvement in decision-making
and, at the government level by advocating for predicted degree of perceived in uence over
policy change. decisions. In particular, breadth of content exper-
Study limitations include small sample size for tise and number of decision activities involving
testing the second hypothesis and potential mea- SNLs were significant predictors of decision
surement issues with regard to the instrument in uence, explaining 22 of the variance in
which may fail to adequately address overlap in uence.
among decision types and lack of specificity for
timing of decisions. Also, the unique aspects of ACKNOWLEDGEMENTS
the Canadian healthcare system may limit gener- Funding for this project: Canadian Health Ser-
alizability to nurse leaders in acute care settings vices Research Foundation (CHRSF)/Canadian
in other countries. While this study shows very Institutes of Health Research (CIHR), the Ontario
positive perceptions of SNLs as executive Health Services Research Co-sponsorship Fund
decision-makers, what is not known are the per- from Ontario Ministry of Health and Long-term
ceptions of the frontline clinicians and managers Care and Nursing Research Fund, Registered
about SNL effectiveness in presenting nursing Nurses Association of Ontario, Office of Nursing
issues. Future studies are needed to examine per- Policy (Health Canada), Centre FERASI,
ceptions at various levels within the organization Minist re de la Sant et des Services Sociaux du
as a means of validating the overall coherence u bec, CIHR Knowledge Translation Branch,
and confidence in SNL effectiveness. Nursing Leadership Network of Ontario, London
Health Sciences Centre, Mount Sinai Hospital
CONCLUSIONS and Vancouver Coastal Health Authority.
Senior nurse leaders play an in uential role in the
future of healthcare organizations through their REFERENCES
participation in decision-making at the senior Anderson R.A. McDaniel R.R. (1998) Intensity of
team level and their ability to in uence how registered nurse participation in nursing home
decision-making. The erontologist (1),
nursing is practiced and valued in the organiza-
90-100.
tion. Despite variations in how health system Anderson R.A. McDaniel R.R. (1999) RN participa-
restructuring has occurred across Canada, it is tion in organizational decision-making and improve-
clear that SNL involvement is critical and highly ments in resident outcomes. ealth Care
valued by CEOs. As a member of the senior lead- Management Revie (10), 7-16.
ership team, the SNL has the opportunity to in u- Anthony M.K., Standing T.S., Glick J. et al. (2005)
Leadership and nurse retention: the pivotal role of
ence team members by ensuring that patient care
nurse managers. ournal of Nursing Administration
and nursing practice perspectives are voiced (3), 146-155.
when decisions affect organizational directions, Arnold L., Drenkard K., Ela S. et al. (2006) Strategic
quality management and resource use. In general, positioning for nursing excellence in health
we found that SNLs perceived they had early systems—Insights from chief nursing executives.
Nursing Administration uarterly (1), 11-20.
involvement in decision-making processes, con-
Ashmos D.P. McDaniel R.R. (1991) Physician
tributed breadth of content expertise to most deci- participation in hospital strategic decision-making:
sion types, engaged in a variety of decision the effect of hospital strategy and decision content.
activities and had considerable decision-making ealth Services Research (3), 375-401.
540 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals

Ashmos D.P., Huonker J.W. McDaniel R.R. (1998) Havens D.S. (2001) Comparing nursing infrastructure
Participation as a complicating mechanism: the effect and out-comes: ANCC nonmagnet CNEs report.
of clinical professional and middle manager participa- Nursing Economics (6), 258-266.
tion on hospital performance. ealth Care Manage- Havens D.S. Laschinger H.K. (1997) Creating the
ment Revie (4), 7-20. environment to support shared governance: Kanter’s
Banaszak-Holl J., Alexander J., Valentine N.M., theory of power in organizations. ournal of Shared
Piotrowski M.M., Adams-Watson J.G. Davis J. overnance (1), 15-23.
(1999) Decision-making activity and in uence of Huston C. (2008) Preparing nurse leaders for 2010.
nurse executives in top management teams. ournal ournal of Nursing Management , 905-911.
of Nursing Administration (4), 18-24. Kirk H. (2008) Nurse executive director effectiveness: a
Baumann A., O’Brien-Pallas L. Armstrong-Stassen systematic review of the literature. ournal of
M. et al. (2001) Commitment and Care: The Bene ts Nursing Management , 374-381.
of a ealthy or place for Nurses, their Patients and Kleinman C.S. (2003) Leadership roles, competencies,
the System. Report submitted to the Canadian Health and education. ournal of Nursing Administration
Services Research Foundation, Ottawa ON. (9), 451-455.
Berggren I. Severinsson E. (2003) Nurse supervisors’ Krairiksh M. (2000) The Relationship among Staff
actions in relation to their decision-making style and Nurses Participation in Decision-Ma ing, Nurse
ethical approach to clinical supervision. ournal of Managers eadership Competencies, and Nurse-
Advanced Nursing (6), 615-622. Physician Collaboration. npublished doctoral
Canadian Nursing Advisory Committee (2002) ur dissertation, Case Western Reserve niversity,
ealth, ur uture: Creating uality or places for Cleveland, OH, SA.
Canadian Nurses, Advisory Committee on Health Laschinger H.K., Wong C.A., Ritchie J. et al. (2008)
Human Resources, Ottawa, ON. A profile of the structure and impact of nursing
Clancy T.R. (2003) The art of decision-making. ournal management in Canadian hospitals. ealthcare
of Nursing Administration (6), 343-349. uarterly (2), 85-94.
Clifford J.C. (1998) Restructuring: The Impact of Lauri S., Salantera S., Chalmers K. et al. (2001) An
ospital rgani ation on Nursing eadership, exploratory study of clinical decision-making in five
Jossey-Bass Publishing, San Francisco, CA. countries. ournal of Nursing Scholarship (1),
Duffield C., Moran P., Beutel J. et al. (2001) Profile of 83-90.
first-line nurse managers in New South Wales, Leatt P., Lemieux-Charles L. Aird C. (1994) Program
Australia in the 1990s. ournal of Advanced Nursing management: introduction and overview. In Program
(6), 785-793. Management and Beyond: Management Innovations
Dwore R.B., Murray B.P., Fosbinder D. et al. (2000) in ntario ospitals (P. Leatt, L. Lemieux-Charles
Integration of nurse executives into executive level C. Aird eds), pp. 1-10, Canadian College of Health
administration in tah hospitals. The ealth Care Services Executives, Ottawa, ON.
Manager (4), 22-36. Mass H., Brunke L., Thorne S. Parslow H. (2006)
Fedoruk M. (2000) The nurse executive: challenges for Preparing the next generation of senior nursing
the 21st century. ournal of Nursing Management leaders in Canada: perceptions of role competencies
(1), 13-20. and barriers from the perspectives of inhabitants and
Filkins J. (2003) Nurse directors’ jobs—A European aspirants. Canadian ournal of Nursing eadership
perspective. ournal of Nursing Management (1), (2), 75-91.
44-47. Mintzberg H., Raisinghani D. Th or t A. (1976)
Fonville A.G.M. (2002) o Nurse E ecutives Ac uire The structure of unstructured decision processes.
and se Ethics no ledge in Management Decision- Administrative Science uarterly (2), 246-
Ma ing. npublished doctoral dissertation, Columbia 275.
niversity Teachers College, New York, NY, SA. Murray B.P., Fosbinder D., Parson R.J. et al. (1998)
Freund C.M. (1988) Decision-making styles: managerial Nurse executives’ leadership roles: perceptions of
application of the MBTI and type theory. ournal of incumbents and in uential colleagues. ournal of
Nursing Administration (12), 5-11. Nursing Administration (6), 17-24.
Havens D.S. (1998) An update on nursing involvement Oetjen R.M., Oetjen D.M. Rotarius T. (2008)
in hospital governance: 1990-1996. Nursing Econom- Administrative decision making: a stepwise method.
ics (1), 6-11. The ealth Care Manager (1), 4-12.
 APPENDIX B Decisional Involvement of Senior Nurse Leaders in Canadian Acute Care Hospitals 541

Oroviogoicoechea C. (1996) The clinical nurse and Management in Canada, 3rd ed (J.M. Hibberd
manager: a literature review. ournal of Advanced D.L. Smith eds), pp. 199-237, Elsevier Canada,
Nursing , 1273-1280. Toronto, ON.
Robbins S.P. Langton N. (2003) rgani ational Thompson C. (1999) A conceptual treadmill: the need
Behaviour: Concepts, Controversies, Applications, for middle ground’ in clinical decision making
Pearson Education Canada, Toronto, ON. theory in nursing. ournal of Advanced Nursing
Shamian J. Lightstone E.Y. (1997) Hospital restruc- (5), 1222-1229.
turing initiatives in Canada. Medical Care (10), Thorman K.E. (2004) Nursing leadership in the
OS62-OS69. boardroom. ournal of bstetric, ynecologic, and
Shannon V. French S. (2005) The impact of the Neonatal Nursing , 381-387.
re-engineered world of health-care in Canada on penieks V.V. (2003) What constitutes effective
nursing and patient outcomes. Nursing In uiry (3), leadership Perceptions of magnet and nonmagnet
231-239. nurse leaders. ournal of Nursing Administration
Sharp N.D., Grelner G.T., Li Y. et al. (2006) Nurse (9), 456-467.
executive and staff nurse perceptions of the effects of Wangsness S.I. (1991) A Study of Decision-Ma ing
reorganization in Veterans Health Administration Activities of Nurse E ecutives in Acute Care
Hospitals. ournal of Nursing Administration (10), Pennsylvania ospitals. npublished doctoral
471-478. dissertation, Pennsylvania State niversity, Hershey,
Smith S.L. Friedland D.S. (1998) The in uence of PA, SA.
education and personality on risk propensity in nurse Wells R., Alexander J.A., Piotrowski M.M. et al. (1999)
managers. ournal of Nursing Administration (12), How other members of the top management team see
22-27. the nurse executive. Nursing Administration uar-
Smith D., Klopper H.E., Paras A. Au A. (2006) terly (3), 38-51.
Structure in health agencies. In Nursing eadership
AP PENDIX C

The Effect of Music on Parental Participation During

Pediatric Laceration Repair
Gregory Sobieraj, Maala Bhatt, Sylvie LeMay, Janet Rennick, Celeste Johnston

The purpose of this quasi-experimental study was to cipatory fear. A child’s fear, anxiety, and distress
test an intervention on the use of music during simple must be adequately addressed to ensure success-
laceration repair to promote parent-led distraction in
ful laceration repair and a positive hospital expe-
children aged 1 to 5. Children’s songs were broadcast via
speakers during laceration repair and parents were rience for parents and their children. The focus
encouraged to participate in distracting their child. The of this study was the testing of music as an inter-
proportion of parental participation was determined. vention to increase parent-led distraction in the
Laceration procedures were videotaped and objectively ED.
scored using the Procedure Behavior Check List. A total
of 57 children participated in the study. There was no LITERATURE REVIEW
difference in parental involvement between the control
and intervention groups. When age, sex, and condition There is a wealth of research supporting the
were controlled for, distress scores were significantly notion that children who are undertreated for pain
higher if the father was present in the procedure room suffer long-term deleterious effects. For example,
than if only the mother was present (43.68 vs. 23.39, Taddio, Katz, Ilersich, and Koren (1997) found
t (54) 4.296, p = < 0.001). It was concluded that distress
that neonates who were circumcised without
varies with the age of the child and the parent who is
present during the procedure. Providing music during analgesia experienced more distress in subse-
simple laceration repair did not increase the proportion quent routine immunizations than children who
of parents who were involved in distraction. were circumcised with topical analgesia. Simple
laceration repair, defined in this study as the
KEYWORDS: lacerations, music, intervention studies, application of tissue adhesives or sutures to repair
pediatrics, pain
torn or damaged tissue without the use of seda-
tives, is a relatively painless procedure. However,
children under 5 years of age are unable to dis-
INTRODUCTION tinguish between the experience and the sensation
In Canada more than 95,000 children visit emer- of fear (Carr, Lemanek, Armstrong, 1998
gency departments (EDs) annually because of Goodenough et al., 1999). Therefore, intense fear
injuries (Public Health Agency of Canada, 2002b) experienced by the child during laceration repair
and lacerations and other open wounds account may lead to long-term negative outcomes similar
for 25 of injuries among children aged 1 to to those described by Taddio and colleagues. If
4 (Public Health Agency of Canada, 2002a). we consider distress as the sum of anxiety and
Although laceration repair is a relatively painless pain (Walco, Conte, Labay, Engel, eltzer,
procedure due to the use of topical anesthesia, 2005), any medical procedure resulting in distress
children experience distress as a result of anti- in young children may lead to long-term deleteri-
ous effects. It should therefore be a priority for
From the Canadian ournal of Nursing Research, 2009 all pediatric health-care providers to implement
41:68–82. © McGill niversity School of Nursing. distress reduction in their practice.
542
 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair 543

Age has been identified as a significant vari- of being helpful (Piira, Sugiura, Champion,
able in studies assessing pediatric distress, espe- Donnelly, Cole, 2005). Finally, it has been
cially during acute, painful procedures. A study demonstrated that parental engagement in coping
by Goodenough and colleagues (1999) found that behaviours, such as use of humour and non-
ratings of pain and unpleasantness during a procedural talk directed at the child, serve to
painful medical procedure decreased with increas- decrease the amount of distress experienced by
ing age. An earlier study, similarly, found a nega- the child (Blount et al., 1989).
tive correlation between pain (both subjective and Interestingly, some coping strategies used by
objective) and age, indicating that the pain parents, such as verbal reassurance (e.g., it’s
response is attenuated by age (Fradet, McGrath, okay, don’t worry ), empathy, criticism, and
Kay, Adams, Luke, 1990). apologizing for the child’s behaviour, have been
Psychological interventions have been shown shown to heighten the child’s distress (Blount
to have a positive effect on procedural distress. et al., 1989 Manimala, Blount, Cohen, 2000
Techniques such as distraction have a clear benefit McMurtry, McGrath, Chambers, 2006). It is
in procedures such as venous cannulation or unclear how these parental behaviours directed
lumbar puncture (Cohen, 2002 man, Cham- towards the child serve to increase distress.
bers, McGrath, Kisely, 2006). There are, McMurtry and colleagues (2006) summarize the
however, very few studies exploring these bene- findings on reassurance and report that this coping
fits in painless procedures such as laceration strategy may increase distress via three mecha-
repair using topical anesthesia. Sinha, Christo- nisms. First, reassurance may cue the child to
pher, Fenn, and Reeves (2006) attempted to use prepare for an unpleasant event and incite fear
music as a distraction during laceration repair in and anxiety in the child. Second, it may reinforce
children aged 6 to 18. They found that music and encourage distress behaviour: The more the
effectively reduced anxiety in both the children child expresses feelings of distress, the more
and their parents during the procedure but that it attention he receives from the parent. Finally,
did not have an effect on the sensation of pain. reassurance may provide validation for the child’s
Conversely, a recent Cochrane review concluded feelings, effectively telling the child that it is
that music has a small but measurable effect on okay to be distressed (McMurtry et al., 2006).
the sensation of pain the authors recommend that Showing empathy and apologizing for the child’s
although music should not be used as a first-line behaviour likely work via similar mechanisms.
treatment for pain, it could serve as a useful Further, it has been demonstrated that parental
adjunct to analgesia (Cepeda, Carr, Lau, engagement in these distress-promoting behav-
Alvarez, 2006). Its positive effect on distress and iours can result in similar behaviours by those
its unobtrusive nature make music an ideal inter- treating the child, such as nurses and physicians
vention for testing in a busy environment such as (Frank, Blount, Smith, Manimala, Martin,
an ED. 1995).
Child caregivers are often closely attuned to It is therefore important that strategies be
the child and consequently can have a consider- developed whereby a parent can actively partici-
able effect on levels of distress experienced by pate in a procedure and thus be made to feel
the child. When a parent is present in the treat- helpful yet not engage in distress-promoting
ment room during a potentially painful event, behaviour. Distress-promoting behaviours may
positive effects include lower distress scores for be difficult to prevent, as a parent will intuitively
the parent and the child (Wolfram Turner, attempt to reassure a child who is experiencing
1996), increased parent satisfaction, and a sense distress. Music, as a recommended adjunct for the
544 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair

treatment of pain, may be a useful tool for dis- on parent-led distraction during simple lacera-
tracting the child and involving the parent in an tion repair in children aged to
activity that will prevent him or her from engag-
ing in distress-promoting behaviours. METHODS AND MATERIALS
PURPOSE Design
It has been demonstrated that distraction is an This quasi-experimental study was conducted in
effective means of decreasing distress. The a pediatric ED located in a large city. This study
purpose of this study was to test an intervention design was chosen over randomization because
using children’s songs to promote parent-led dis- no between-group differences were expected in
traction during simple laceration repair in chil- children presenting at the ED, based on a review
dren aged 1 to 5. Parents were encouraged to of the department’s patient-tracking software
participate in their child’s treatment by singing conducted by one of the investigators. As the
along to music being broadcast via speakers. A study site was experiencing a severe staff short-
parent who actively participates by singing will age at the time of the study, this method also
have less opportunity to engage in distress- served to minimize any burden associated with
promoting behaviour and, as reported by Sinha randomization (e.g., using randomization soft-
and colleagues (2006), may experience less ware) and to simplify study logistics for partici-
anxiety during the procedure. Maternal behav- pating ED staff.
iour could account for as much as 53 of the Recruitment took place in 2-week blocks.
variance in distress experienced by a child Those children presenting during the first 2 weeks
(Frank et al., 1995). This finding supports the of the study had laceration repair as per depart-
notion that an intervention targeting both parent ment protocol, without the music intervention.
and child could have a significant impact on the During the second 2 weeks, consenting patients
child’s distress. This simple and easily imple- received the intervention. Recruitment took place
mented intervention provides parents with a over an 8-week period, Monday to Friday from
medium through which to distract the child noon to 8 P.M. A review of the patient-tracking
while simultaneously avoiding distress-promoting software used at the study site, which tracks the
behaviours. Further, music is an inexpensive, presenting complaint, demographic data, and dis-
easily implemented, low-burden intervention, charge diagnosis, determined that these days and
requiring only the press of a button. This inter- times would allow for the greatest recruitment
vention could lead to a measurable reduction in potential, as they were when lacerations in the
distress during simple laceration repair by 1-to-5 age group were most likely to present at
increasing parent-led distraction, thereby improv- the ED.
ing the hospital experience for both young chil-
dren and their parents without placing an undue Sample
burden on professionals integrating the interven- Children aged 12 to 71 months, inclusive, pre-
tion into their practice. senting at the ED with a single, simple laceration
We hypothesized that parents in the interven- requiring repair with sutures or tissue adhesives
tion group would demonstrate a greater degree of and pretreated topically with lidocaine, epineph-
parent-led distraction than those in the control rine, and tetracaine (LET) were included in the
group. With this objective in mind, we formulated study. This age range was selected so the inter-
the following research question: Does music vention could be studied in a narrow develop-
broadcast via spea ers have a measurable effect mental range and to facilitate standardization of
 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair 545

the intervention and pre-procedural teaching. All procedure (bandage placed over laceration). The
children who required suturing received LET fol- parents were encouraged to sing along with
lowed by injectable lidocaine to ensure that the the music during the procedure. Participants in
procedure remained painless. Children were the non-intervention group (usual care) had no
recruited regardless of prior experience with lac- music played. All laceration-repair procedures
erations or laceration repair. Excluded were chil- were videotaped. The research assistant accom-
dren who had more than one laceration, required panied the physician, patient, and parents at all
sedation for their laceration repair, presented at the procedures and was responsible for proper
the ED without a family member, or were accom- positioning of video equipment and for starting
panied by a family member who did not speak the music at the beginning of the procedure.
English or French. Approval for the study was obtained from both
Families were identified as eligible for the the Nurse Manager and the Medical Director of
study by the triage nurse and agged for the the ED. Ethical approval was obtained from the
research assistant. The research assistant then ethical review board prior to implementation.
approached the family and requested consent for Informed consent was obtained by the research
participation prior to examination by the physi- assistant prior to videotaping the procedure. All
cian. Recruitment took place over the months of taped procedures were transferred to a dedicated
July and August 2008. Of the 69 families screened hard drive in a locked office at the end of each
for the study, 68 agreed to participate (98 ). study week. Videotapes were accessed and viewed
Eleven of the families were excluded from the only by the researchers and objective scorers.
final analysis because the child did not meet Patient confidentiality was ensured through the
inclusion criteria after being examined by the replacement of patient names with codes on all
physician (e.g., required sedation, required study materials. Consent forms were kept sepa-
complex laceration repair, had multiple injuries). rate from other study materials at all times.
One family did not provide a specific reason for
refusal to consent. In total, 57 families were Instruments
included in the final analysis, 27 of whom received
the music intervention. Parental Participation
The video scorers determined the amount of time
Intervention a parent spent distracting the child during the
All consenting families were met by a Child Life procedure. They were trained to recognize behav-
Specialist (CLS), who provided pre-procedural iours that distracted the child. Behaviours such as
teaching to the parent and child. The pre- singing to the child, diverting the child’s attention
procedural teaching was standardized between away from the laceration repair, or encouraging
the two groups. Children assigned to the interven- the child to sing were considered to be parental
tion group had audiorecorded children’s songs participation. The video scorers recorded the
played to them during the procedure. The song number of seconds spent on each distraction
choices included lullabies, educational songs, and event. For example, they timed exactly how long
songs performed by popular television characters a parent would sing along with the music being
in both English and French. Three songs were broadcast. A parental participation score was then
selected by the CLS and the parents prior to the derived by determining the proportion of time
procedure. These were played throughout the pro- spent on distracting the child (time distracting/
cedure on a repeating basis, from the start of the total procedure time). Interrater reliability
procedure (child placed on bed) to the end of the for proportion of parental participation was
546 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair

determined (0.767, p < 0.01, CI 95 0.632, study by comparing rater and investigator scores
0.923) and judged to be acceptable. on sample videotapes. Coding of the videotapes
Since the scores given by the two raters were was begun by the raters only when reliability was
similar, they were averaged to create an objective greater than 0.80 on sample videotapes. Follow-
distress score and a parental participation score ing data collection, interrater reliability was
(in seconds), which were used in the subsequent strong for the two video scorers on the objective
analysis. measure of distress (0.884, p < 0.01, CI 95 0.81,
0.93) and the time to complete the procedure
Procedure Behavior Check List (0.995, p < 0.01, CI 95 0.991, 0.997).
Videotapes of all laceration repairs were objec-
tively scored using the Procedure Behavior Check RESULTS
List (PBCL) (LeBaron eltzer, 1984). The The intervention and control groups were
PBCL is an observational measure of distress that similar for age, location of laceration, length
scores the presence and intensity of eight behav- of laceration, and family member present,
iours associated with child pain and anxiety (e.g., but dissimilar for gender. Children in the inter-
muscle tension, verbal stalling, crying). Each vention group more frequently required sutures to
behaviour is rated on a Likert-type scale ranging repair the laceration (26 vs. 7 ) (Table C-1)
from 0 to 5 (0 = no distress 1 = very mild dis- however, this difference was not statistically
tress 5 = extremely intense distress), for a score significant.
ranging from 0 to 40. This tool was originally Linear regression analysis was performed to
used to measure observable distress during lumbar determine whether parental involvement pre-
punctures in 67 pediatric oncology patients dicted distress scores and the degree to which age
between the ages of 6 and 18 years. Concurrent affected distress. In the control group (n = 30), 18
validity was found to be acceptable, with a cor- parents participated in distracting the child (60 )
relation of 0.80 (p < 0.001) to the children’s self- and the mean proportion of time spent participat-
reports of pain and anxiety (Lebaron eltzer, ing in the laceration repair was low (0.0647). Of
1984). Subsequent studies have shown the PBCL the 27 parents in the intervention group, 15 dis-
to be a reliable and valid measure of behavioural tracted their child (56 ), with a similar mean
distress in children (Cavender, Goff, Hollon, proportion of time spent distracting the child
Guzzetta, 2004 Luhmann, Schootman, Luhmann, (0.0669). There was no significant difference
Kennedy, 2006), with observed distress signifi- between the two groups in terms of parental
cantly correlated with patient ratings of pain and participation.
anxiety (Langer, Chen, Luhmann, 2005). There was no significant difference in distress
Finally, a recent review of observational mea- scores based on parental participation. The great-
sures of pain rated the PBCL one of the most est predictors of child distress were age (β =
accurate measures of pain-related distress cur- −0.434, t = −4.017, p < 0.01), with younger chil-
rently available, with a good balance of evidence, dren being more distressed, and the presence of
burden, and content validity (von Baeyer the father in the procedure room (β = −0.419, t =
Spagrud, 2007). −3.888, p < 0.01). Children had a significantly
Videotapes were scored by two reviewers higher mean distress score when the father was
naive to the study purpose using the PBCL. The present (43/100) than when only the mother was
reviewers were trained in the use of the PBCL by present (23/100) ( (1, 54) = 18.452, p < 0.01).
study investigators prior to the study start date. (See Table C-2 for descriptive and comparative
Interrater reliability was established prior to the data on distress scores.)
 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair 547

TABLE C-1
DEMOGRAPHIC AND PROCEDURAL CHARACTERISTICS OF INTERVENTION AND CONTROL GROUPS
CONTROL INTERVENTION CHI-SQUARED p

Age of child (months) 43.5 ± 14.4 39.9 ± 18.7 ns

Sex (n) 0.047*
Male 24 15
Female 6 12
Parent present (n) ns
Father 4 4
Mother 19 16
Both 6 7
Other 1 0
Location of laceration (n) ns
Scalp 9 3
Face 17 21
Other 3 3
Length of laceration (mm) 13.2 ± 7.6 14.5 ± 7.7 ns
Type of repair (n) ns
Tissue adhesive 28 20 ns
LETa + sutures 1 1
LET + lidocaine + sutures 1 6
N 30 27
*Significant at 0.05.
a
Topical anesthesia.

TABLE C-2
DISTRESS SCORE DATA: TREATMENT GROUP BY PARENTAL PRESENCE (DISTRESS SCORES 0–100)
CONTROL INTERVENTION MEAN F P

Mean distress (n) 33.1 (29) 28.6 (26) ns

Parent present
Mother (n) 27.1 (19) 19 (16) 23.05 (2, 52) 9.516 <0.01
Father (n) 40.2 (4) 44.7 (4) 42.5
Both (n) 48.1 (6) 43.4 (6) 45.8
Note: Two cases were excluded; one child was accompanied by an aunt and one child was missing objective data.

distress during laceration repair may decrease

DISCUSSION the need for sedatives such as midazolam, the
There is a paucity of research on distress during use of which increases observation time (post-
laceration repair. Although this is a relatively intervention) and the incidence of sequelae
simple, quick, and painless procedure, it is per- (Luhmann, Kennedy, Porter, Miller, Jaffe,
ceived by observers as extremely distressing 2001). Our results indicate that the strongest
(Babl, Mandrawa, O’Sullivan, Crellin, 2008). predictors of distress are age and the parent
As resolution of pain is an important predictor who accompanies the child in the treatment
of high patient satisfaction in children (Magaret, room. The finding that distress is strongly cor-
Clark, Warden, Magnusson, Hedges, 2002), it related with age is in concordance with the
may be inferred that effective resolution of dis- results of several other studies examining pedi-
tress during simple laceration repair may also atric distress (Carr et al., 1998 Goodenough
increase patient satisfaction. Further, reducing et al., 1999).
548 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair

Although the intervention did reduce distress the mean age for the group in which the father
in children (see Bhatt, Sobieraj, Johnston, was present was slightly lower (38.3 vs.
2009), in the present study parental participation 43.9 months), this is likely not a sufficiently large
was not higher in the intervention group. Although age difference to explain the stress differences.
parents were encouraged to distract their child There are no prior studies reporting a similar
during the procedure, the proportion of time spent finding. As different coping strategies are known
distracting the child, regardless of condition, was to provoke varying degrees of distress (Manimala
extremely small (6.6 of total procedure time). et al., 2000 McMurtry et al., 2006 Young, 2005),
The intervention may not provide sufficient stim- it may be that fathers in our study were using
ulus to overcome the unpleasantness of seeing coping strategies known to increase distress, such
one’s child in distress. In the future, more time as reassurance, criticism, or apologizing for the
with parents in pre-procedural teaching, to stress child’s behaviour, more frequently than mothers,
the importance of distraction, may serve to while mothers may have been using effective
increase the proportion of time spent participat- coping strategies, such as distraction, humour, or
ing. If the proportion of time spent participating non-procedural talk, with greater frequency.
is increased, we might observe a lowering of dis- Since families self-selected who would accom-
tress scores, as had been expected, since the pany the child in the procedure room, a second
parents will have less opportunity to engage in possibility for this difference in distress is that
distress-promoting behaviours. fathers chose to accompany difficult or expres-
Pre-procedural teaching has been demon- sive children more frequently than mothers alone
strated to reduce anxiety prior to a procedure did. Without collecting more data from parents
(Claar, Walker, Barnard, 2002 Spafford, von regarding their relationship with the child, or their
Baeyer, Hicks, 2002). Presumably the older preferred method of coping, it is hard to draw
children in our sample had learned more from the conclusions with respect to this difference in dis-
pre-procedural teaching and applied the informa- tress. A secondary analysis of the videotapes
tion more effectively. If distress is defined as the would allow us to determine the frequency and
sum of anxiety and pain (Walco et al., 2005), type of coping strategies used by family members,
then older children who are less anxious as a and to validate the hypothesis that different family
result of pre-procedural teaching will experience members use alternative coping strategies.
less distress. Therefore, we cannot rule out the One study has suggested that distraction loses
possibility that the difference in distress levels efficacy in reducing distress if the painful or
between age groups is a result of an association unpleasant stimulus is prolonged (McCaul
between increasing age and pre-procedural teach- Malott, 1984). Laceration repair in our study took
ing and is not in fact an accurate representation several minutes to complete (M = 328 seconds),
of distress scores. In the future it would be imper- in stark contrast to immunization, heel sticks, or
ative to add a third group to the study, one in blood sampling, which may take only seconds.
which no pre-procedural teaching has been pro- The degree to which a child is distracted may be
vided by a CLS, in order to control for this poten- further in uenced by their degree of stimulation.
tial confounding variable. As our intervention was fairly passive, it may not
A novel finding in our study was the difference have provided a sufficiently strong stimulus to
in distress scores depending upon which family overcome the unpleasantness of the laceration
member accompanied the child during the proce- repair.
dure. Children were significantly more distressed The present study had several limitations. A
if the father was in the treatment room. Although non-randomized design was chosen for the study,
 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair 549

because there were no differences expected in to families where there is greater need. Older
children presenting to the ED during the 2-week children may require less attention by auxiliary
study blocks. Despite this expectation, groups staff. This finding suggests that auxiliary staff can
differed on gender, family member present spend more time attending to the needs of other
during the procedure, and type of laceration patients on the unit. Further, the data suggest that
repair. A single-blind RCT may have prevented the younger population may require more atten-
the skewing of groups and increased the gener- tion from support staff than they are currently
alizability of our results. We cannot conclude receiving, to lower the increased level of distress
that the gender composition among groups experienced by these patients.
affected our results, as the literature on gender
differences and distress in children is inconclu- CONCLUSION
sive (e.g., Carr et al., 1998 Goodenough et al., Although the provision of music and pre-
1999). procedural teaching did not increase the propor-
Because audio was recorded and required for tion of parental participation, the study did find
proper scoring of the videotapes, the objective that children are more distressed in the presence
scorers were not blind to group assignment. of fathers—an important finding not described in
However, the scorers remained blind to study other studies. This finding will help inform future
purpose throughout the study, which reduced the studies where parent gender may be an important
risk of bias in video scoring. In any future research covariate.
it may be useful to apply a measure that does not
require audio cues, such as the Child Facial ACKNOWLEDGEMENTS
Coding System (CFCS) (Breau et al., 2001), to The authors would like to acknowledge the
reduce the risk of bias introduced by scorers who funding support of Groupe de recherche interuni-
are not blind to group assignment. versitaire en interventions en sciences infirmi res
The small sample size (N = 57) may be du u bec (GRIISI ) and the Montreal Chil-
a further limitation. A larger sample size dren’s Hospital Research Institute the Emer-
would have increased the power of the study and gency Department staff and the Child Life
allowed us to detect a smaller clinical effect. Services Department of the Montreal Children’s
Further, no qualitative data were collected from Hospital Fonds de la recherche en sant du
participating families and staff. Data such as sat- u bec (FRS ) and their research assistants,
isfaction with the intervention, likelihood of Sushmita Shivkumar and Kamy Apkarian.
adopting the intervention for future procedures,
and parents’ and staff members’ perceptions of Disclosure: There is no con ict of interest.
the effectiveness of the treatment might have
allowed us to infer the clinical usefulness of the
intervention. regory Sobieraj, RN, BN, is a PhD student at the School
of Nursing, Mc ill niversity, Montreal, uebec,
Practice Implications Canada Maala Bhatt, MD, MSc, is Attending Physician,
Our findings suggest that significant predictors of Division of Pediatric Emergency Medicine, Montreal
higher levels of distress during laceration repair Children s ospital Sylvie eMay, RN, PhD, is Associate
Professor, aculty of Nursing, niversit de Montr al
are younger age and paternal accompaniment in
anet Rennic , RN, PhD, is Nurse Scientist, Montreal
the procedure room. This information could in u- Children s ospital Celeste ohnston, RN, DEd, CA S,
ence unit managers/team leaders to more effec- is Professor and Associate Director of Research, School
tively allocate available resources, such as CLSs, of Nursing, Mc ill niversity
550 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair

Goodenough, B., Thomas, W., Champion, G. D.,

REFERENCES Perrott, D., Taplin, J. E., von Baeyer, C. L., et al.
Babl, F., Mandrawa, C., O’Sullivan, R., Crellin, D. (1999). nravelling age effects and sex differences in
(2008). Procedural pain and distress in young needle pain: Ratings of sensory intensity and
children as perceived by medical and nursing staff. unpleasantness of venipuncture pain by children and
Pediatric Anesthesia, (5), 412-419. their parents. Pain, , 179-190.
Bhatt, M., Sobieraj, G. S., Johnston, C. C. (2009, Langer, D. A., Chen, E., Luhmann, J. D. (2005).
May 5). Improving emergency department treatment Attributions and coping in children’s pain experi-
of pediatric lacerations ith the use of children s ences. ournal of Pediatric Psychology, (7),
songs as a distraction techni ue. Poster presentation 615-622.
at meeting of Pediatric Academic Societies, Balti- LeBaron, S., eltzer, L. (1984). Assessment of acute
more, Maryland. pain and anxiety in children and adolescents by
Blount, R. L., Corbin, S. M., Sturges, J. W., Wolfe, self-reports, observer reports, and a behavior
V. V., Prater, J. M., James, L. D. (1989). The checklist. ournal of Consulting and Clinical
relationship between adults’ behavior and child Psychology, (5), 729-738.
coping and distress during BMA/LP procedures: Luhmann, J. D., Kennedy, R. M., Porter, F. L.,
A sequential analysis. Behavior Therapy, (4), Miller, J. P., Jaffe, D. M. (2001). A randomized
585-601. clinical trial of continuous- ow nitrous oxide and
Breau, L., McGrath, P. J., Craig, K. D., Santor, D., midazolam for sedation of young children during
Cassidy, K. L., Reid, G. J. (2001). Facial expres- laceration repair. Annals of Emergency Medicine,
sion of children receiving immunizations: A principal (1), 20-27.
components analysis of the child facial coding Luhmann, J. D., Schootman, M., Luhmann, S. J.,
system. Clinical ournal of Pain, (2), 178-186. Kennedy, R. M. (2006). A randomized comparison of
Carr, T. D., Lemanek, K. L., Armstrong, F. D. (1998). nitrous oxide plus hematoma block versus ketamine
Pain and fear ratings: Clinical implications of age and plus midazolam for emergency department forearm
gender differences. ournal of Pain and Symptom fracture reduction in children. Pediatrics, (4),
Management, (5), 305-316. e1078-e1086.
Cavender, K., Goff, M. D., Hollon, E. C., Guzzetta, Magaret, N. D., Clark, T. A., Warden, C. R., Magnus-
C. E. (2004). Parents’ positioning and distracting son, A. R., Hedges, J. R. (2002). Patient satisfac-
children during venipuncture: Effects on children’s tion in the emergency department—a survey of
pain, fear, and distress. ournal of olistic Nursing, pediatric patients and their parents. Academic
(1), 32-56. Emergency Medicine, (12), 1379-1388.
Cepeda, M. S., Carr, D. B., Lau, J., Alvarez, H. Manimala, M. R., Blount, R. L., Cohen, L. L.
(2006). Music for pain relief. Cochrane Database of (2000). The effects of parental reassurance versus
Systematic Revie s, , CD004843. distraction on child distress and coping during
Claar, L. R., Walker, L. S., Barnard, J. A. (2002). immunizations. Children s ealth Care, (3),
Children’s knowledge, anticipatory anxiety, proce- 161-177.
dural distress, and recall of esophagogastroduodenos- McCaul, K. D., Malott, J. M. (1984). Distraction and
copy. ournal of Pediatric astroenterology and coping with pain. Psychological Bulletin, (3),
Nutrition, (1), 68-72. 516-533.
Cohen, L. L. (2002). Reducing infant immunization McMurtry, C. M., McGrath, P. J., Chambers, C. T.
distress through distraction. ealth Psychology, (2006). Reassurance can hurt: Parental behavior and
(2), 207-211. painful medical procedures. ournal of Pediatrics,
Fradet, C., McGrath, P. J., Kay, J., Adams, S., Luke, (4), 560-561.
B. (1990). A prospective survey of reactions to blood Piira, T., Sugiura, T., Champion, G. D., Donnelly, N.,
tests by children and adolescents. Pain, (1), 53-60. Cole, A. S. J. (2005). The role of parental presence in
Frank, N. C., Blount, R. L., Smith, A. J., Manimala, M. the context of children’s medical procedures:
R., Martin, J. K. (1995). Parent and staff behavior, A systematic review. Child: Care, ealth and
previous child medical experience, and maternal Development, (2), 233-243.
anxiety as they relate to child procedural distress and Public Health Agency of Canada. (2002a). Canadian
coping. ournal of Pediatric Psychology, (3), ospitals Injury Reporting and Prevention Program
277-289. C IRPP report: Age and se distribution of injured
 APPENDIX C The Effect of Music on Parental Participation During Pediatric Laceration Repair 551

persons. Retrieved September 4, 2009, from http:// during subsequent routine vaccination. ancet,
dsol-smed.phac-aspc.gc.ca/dsol-smed/is-sb/c chrp- (9052), 599-603.
eng.phtml. man, L. S., Chambers, C. T., McGrath, P. J., Kisely,
Public Health Agency of Canada. (2002b). Canadian S. (2006). Psychological interventions for needle-
ospitals Injury Reporting and Prevention Program related procedural pain and distress in children and
C IRPP report: Indication of the type of injury adolescents. Cochrane Database of Systematic
suffered by the injured person. Retrieved September 4, Revie s, , CD005179.
2009, from http://dsol-smed.phac-aspc.gc.ca/ von Baeyer, C. L., Spagrud, L. J. (2007). Systematic
dsol-smed/is-sb/c chrp-eng.phtml. review of observational (behavioral) measures of pain
Sinha, M., Christopher, N. C., Fenn, R., Reeves, L. for children and adolescents aged 3 to 18 years. Pain,
(2006). Evaluation of non-pharmacologic methods of (1-2), 140-150.
pain and anxiety management for laceration repair in Walco, G. A., Conte, P. M., Labay, L. E., Engel, R.,
the pediatric emergency department. Comment. eltzer, L. K. (2005). Procedural distress in children
Pediatrics, (4), 1162-1168. with cancer: Self-report, behavioral observations, and
Spafford, P. A., von Baeyer, C. L., Hicks, C. L. physiological parameters. Clinical ournal of Pain,
(2002). Expected and reported pain in children (6), 484-490.
undergoing ear piercing: A randomized trial of Wolfram, R. W., Turner, E. D. (1996). Effects of
preparation by parents. Behaviour Research and parental presence during children’s venipuncture.
Therapy, (3), 253-266. Academic Emergency Medicine, (1), 58-64.
Taddio, A., Katz, J., Ilersich, A. L., Koren, G. (1997). Young, K. (2005). Pediatric procedural pain. Annals of
Effect of neonatal circumcision on pain response Emergency Medicine, (2), 160-171.
AP PENDIX D

Patient Real-Time and 12-Month Retrospective

Perceptions of Difficult Communications in the
Cancer Diagnostic Period*
Sally Thorne,1 Elizabeth-Anne Armstrong,1 Susan R. Harris,1 T. Gregory Hislop,2 Charmaine
Kim-Sing,3 Valerie Oglov,1 John L. Oliffe,1 Kelli I. Stajduhar4

ABSTRACT KEYWORDS cancer, communication, health care, psycho-

social issues, quality of life, relationships
Communication is a notoriously complex challenge in the
cancer care context. Our program of research involves
exploration of patient–provider communications across the
A cancer diagnosis marks the experiential begin-
cancer trajectory from the patient perspective. Toward this ning of a process that transitions people to patients
end, we have been following a cohort of 60 cancer patients, and takes them on a journey through unfamiliar
representing a range of tumor sites, from immediately after and frightening territory toward an often uncer-
diagnosis through to recovery, chronic, or advanced disease. tain destination. Whether the diagnosis comes as
Drawing on interpretive description analytic techniques, we
an unanticipated and shattering moment in time,
documented patterns and themes related to various com-
ponents of the cancer journey. In this article, we report on or occurs in the context of a protracted season of
findings pertaining to poor communication during the anxiety and alarm that culminates in an eventual
initial diagnostic period, as described by patients at the time declaration, it is inherently characterized by
of diagnosis and 1 year later. These findings illuminate the meaningful communications with health care
dynamics of communication problems during that complex providers. Communication during the diagnostic
period, and depict the mechanisms by which patients
sought to confront these challenges to optimize their cancer
period sets the stage for the relationships cancer
care experience. On the basis of these findings, considered patients will have with those who manage and
in the context of the body of available evidence, sugges- support them on the cancer journey. The diagnos-
tions are proposed as to appropriate directions for system- tic period is therefore acknowledged as a time
level solutions to the complex communication challenges when communication between patients and clini-
within cancer care.
cians is particularly important and sensitive.
ntil malignant disease is confirmed or strongly
1
University of British Columbia, Vancouver, British Columbia, Canada suspected, patients with the possibility of a cancer
2
British Columbia Cancer Research Centre, Vancouver, British Colum- diagnosis tend not to have found their way into a
bia, Canada formalized cancer care system therefore, knowl-
3
British Columbia Cancer Agency, Vancouver, British Columbia, edge about the cancer diagnostic experience is
Canada pertinent to a wide range of care contexts, includ-
4
University of Victoria, Victoria, British Columbia, Canada ing primary care, ambulatory, and inpatient set-
tings. As recent studies confirm, a minority of
From ualitative ealth Research, 2009 19:1383–1394. patients are first told of their diagnosis by cancer
© 2009, The Authors and SAGE. specialists, with a majority officially learning that
552
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 553

they have cancer from a family physician or nature of patient needs and goals over the course
surgeon (Watson, Mooney, Peterson, 2007). of disease (Carlson, Feldman-Stewart, Tishelman,
However, some patients also report first knowing Brundage, SCRN Communication Team,
that they have cancer because of unintentional 2005), and the diversity of social and cultural
messages transmitted in the tone or urgency of the norms and expectations in relation to the commu-
clinic-scheduling clerk’s voice, nonverbal cues nication style preferred by patients (Schofield
from a radiation technologist, or the subtle shift et al., 2003 Siminoff, Graham, Gordon, 2006)
toward cautious language in a primary care pro- and by clinicians (Davis, Williams, Marin, Parker,
vider. Regardless of its form, the diagnosis experi- Glass, 2002 Feldman-Stewart et al., 2005
ence is retained in vivid memory by patients as a Ramirez, 2003). This science has also been
significant and disruptive milestone in their biog- complicated by the challenge of finding agree-
raphy and the entry point into a life-altering way ment on appropriate measures for cancer com-
of being (Evans, Tulsky, Back, Arnold, 2006 munication effectiveness (Carlson et al., 2005
Walsh Nelson, 2003). Thus, it is important that Feldman-Stewart, Brennenstuhl, Brundage,
we understand the dynamics of communications 2007 Schofield Butow, 2004).
surrounding diagnosis to ensure that clinicians Studies from the patient perspective have
who encounter patients in and around the diagnos- clearly documented that effective communication
tic phase are informed about and sensitive to its with health care providers is a high priority
impact, and possess the appropriate skills to concern at diagnosis and throughout the cancer
ensure that patients begin their cancer journey trajectory (Kreps, Arora, Nelson, 2003). During
with optimal support and guidance. the diagnostic phase in particular, obtaining and
applying information is understood to be a press-
BACKGROUND TO THE LITERATURE ing need for patients (Rutten, Arora, Bakos, Aziz,
Although it is well known that communication Rowland, 2005), although there is known to be
will have a strong impact on both the experience considerable diversity among patients relative to
and psychosocial outcomes of cancer (Baile the type of information that is sought and prefer-
Aaron, 2005 Butow, 2005 Fallowfield Jenkins, ence for the manner in which it is delivered (Cox,
2004 Fallowfield, 2008 Hack, Degner, Parker, Jenkins, Catt, Langridge, Fallowfield, 2006
SCRN Communication Team, 2005 Ong, Visser, Eggly et al., 2006).
Lammes, de Haes, 2000), poor communication These conceptual and methodological chal-
remains a significant problem facing cancer lenges make it exceedingly difficult to convinc-
patients across the spectrum of service (Thorne, ingly conclude what is best for all patients in
Bultz, Baile, SCRN Communication Team, relation to the communications surrounding a
2005). In part, this is because of the complexity of cancer diagnosis (Lockhart, Dosser, Cruickshank,
communication and the variability of the human Kennedy, 2007 Schofield Butow, 2004
experience of cancer, both of which create signifi- Schofield et al., 2003), and the absence of a solid
cant challenges for the development of a solid body of evidence-informed communications
empirical science within this field (Feldman- guidelines further compromises the capacity of
Stewart, Brundage, Tishelman, SCRN Com- concerned clinicians to take the appropriate steps
munication Team, 2005 Parker, Davison, toward skill development (Arora, 2003 Butler,
Tishelman, Brundage, 2005 Patrick, Intille, Degner, Baile, Landry, SCRN Communication
abinski, 2005). Recent reviews confirm that the Team, 2005). Where clinical guidelines exist,
science of cancer communication has been they tend to have been generated on the basis of
impeded by factors that include the changing expert professional opinion, with relatively little
554 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions

interpretation from the patient perspective care communication from the patient’s perspec-
(Edvardsson, Pahlson, Ahlstrom, 2006 Hoff, tive. In this study, we recruited a cohort of 60
Tidefelt, Thaning, Hermeren, 2007 Schofield newly diagnosed cancer patients and followed
et al., 2001). Although much of the advice avail- them over time (ideally, at least 2 years) to deter-
able to clinicians is quite general or self-evident, mine how their communication needs and prefer-
the literature contains considerable depth per- ences changed across their cancer trajectory.
taining to such communication practices as Our research uses interpretive description
responding to the distress cues of patients (Beach, methodology (Thorne, Reimer Kirkham,
Easter, Good, Pigeron, 2005), listening, engag- MacDonald-Emes, 1997 Thorne, 2008), an
ing in compassionate inquiry, and responding established qualitative inquiry approach that
empathetically (Back, Arnold, Baile, Tulsky, extracts elements from the social science tradition
Fryer-Edwards, 2005), and ensuring access to psy- (Glaser Strauss, 1967 Lincoln Guba, 1985
chosocial support services (Ambler et al., 2007 Miles Huberman, 1994), integrating their
Nelson, 2004 Taylor, Ismail, Hills, Ainsworth, application in a manner that makes them ideally
2004). Studies from the patient perspective have suited to the study of practice-based questions
generally confirmed that these elements during deriving from applied clinical fields (Sande-
the provision of medical advice are indeed appro- lowski, 2000).
priate (Ptacek Ptacek, 2001). However, because With ethics approval from our university-
of the significant challenge of tailoring com- based review board, we recruited over a 12-month
munication practices to the unique and particular period a cohort of 60 patients representing a range
needs and contextual circumstances of indivi- of disease and demographic variables. The study
dual patients, the standardized communication participants who voluntarily responded to our
approaches available within the literature have recruitment ads in various parts of the cancer
not been widely taken up (Davison Mills, system in a large western Canadian city included
2005). 43 women (72 ) and 17 men (28 ), a majority
The general problem that this article addresses residing in urban (92 ) rather than rural settings
is the continuing prevalence of poor communica- and of Euro-Canadian (95 ) ethnicity, and the
tion during the diagnostic window of time within remainder of Chinese or South Asian heritage.
the cancer experience. Specifically, we report Their ages ranged from mid-30s to late 80s, with
findings associated with those communication 22 (37 ) in the 50 to 59 bracket and 15 (25 ) in
encounters that patients experience as problem- each of the 40 to 49 and 60 to 69 groupings. A
atic during the diagnostic period within the range of tumor sites was re ected within the
context of a longitudinal cohort study of patient cohort, with the primary groupings constituting 23
perspectives of cancer communication across the (38 ) breast, 8 (13 ) hematological (lymphoma/
illness trajectory. By better understanding the leukemia), 7 (12 ) prostate, 6 (10 ) gastroin-
dynamics and nature of what patients consider to testinal, and 6 (10 ) gynecological (cervix/
be difficult communication, we increase our uterus/ovary), as well as 10 (17 ) across other
capacity to search for appropriate mechanisms sites.
with which to prevent and ameliorate these dif- Each member of the cohort was interviewed
ficulties for future patients. initially according to a semi-structured initial
interview guide by one of two interviewers, each
STUDY METHOD of whom held a graduate degree in a health pro-
The data upon which the current analysis draws fessional discipline and had specialized training
is part of an ongoing longitudinal study of cancer in qualitative research interview techniques.
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 555

Following the initial face-to-face individual inter- tion are not meant to re ect representativeness
view, interviewers conducted bimonthly follow-up of the population rather, when articulated in a
interviews either in person or by telephone, manner that is authentic and credible to the reader,
according to the preference and convenience of they can re ect valid descriptions of sufficient
the study participant. In most instances, the set of richness and depth that their products warrant a
follow-up interviews for each cohort member degree of generalizability in relation to a field of
re ected a mixture of the two formats. These understanding.
follow-up interviews were guided by themes and
questions derived from ongoing constant com- FINDINGS
parative analysis to develop emerging analytic Although patient descriptions of communication
themes and clarify individual variations across experiences both at the onset of the study and
accounts (Miles Huberman, 1994). In this ongoing were often complex, typically involving
manner, we followed the evolving trajectory of multiple health care professionals and circum-
each individual, as well as engaged them in revis- stances, our trigger questions were specifically
iting and re ecting on discrete events. designed to prompt elaboration on those elements
As might be predicted, the cohort members within communication they found particularly
varied in their trajectory, with some achieving a helpful and unhelpful. Within the extensive
level of confidence about survivorship, others accounts arising from these prompts, most of the
continuing active treatment, and still others study participants reported a range of helpfulness
approaching the terminal phase of the disease. At among and across encounters, with a majority
the time of writing, 6 members of the cohort have recalling at least one instance of what they con-
died of their disease. Thus, we have subsets sidered to be good and poor communication. In a
within the ongoing cohort representing a range of significant number of instances, the overall com-
events post-diagnosis. However, as the diagnosis munications these patients reported in this context
experience was an event common to all, each were depicted as sufficiently helpful or unhelpful
member was explicitly interviewed at length to permit legitimate characterization of their com-
about that aspect, both in the initial phases of munication experience as generally negative or
study participation and also in a targeted inter- positive during this time period. Within the cohort
view re ecting back on that period of time 12 of 60 patients in this study, 25 (42 ) reported
months later. It is these components of the overall decidedly negative communication experiences
continuing data set on which this article is based. pertaining to their diagnostic period, a proportion
In the interpretive description method, the col- perhaps not surprising given the evidence in the
lection and analysis of data are considered con- literature as to the prevalence of poor cancer care
current processes, with each informing the other communication.
iteratively over time. All interviews were tran- For the purposes of this analysis into dynamics
scribed verbatim into electronic format, and we of poor communication, we drew on the reported
used a qualitative computer software manage- instances patients characterized as unhelpful an
ment system to assist with data sorting and man- analysis of helpful communication is reported
agement. The analytic process we used involved elsewhere (Thorne et al., in press). Specifically,
rigorous engagement in interpretive cross- the current analysis derives from the multiple
comparison among and between cases in the transcripts from all 60 cohort members for illus-
context of our emerging thematic synthesis trative material relevant to understanding what
(Sandelowski, Trimble, Woodard, Barroso, constitutes difficult communication encounters
2006). Findings drawn from interpretive descrip- during the diagnostic time period from the patient
556 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions

perspective, with a particular emphasis on those explain what it was about that experience that
25 for whom the experience was predominantly affected them so strongly. Although each patient’s
negative. In doing so, we use the language of dif- story revealed unique features and nuances, we
cult to signal our recognition that this kind of noted clear patterns within the dimensions that
study can pass no judgment on what actually constituted what they considered to be poor com-
occurred, but rather re ects what the patient per- munication. These patterns pertained to problem-
ceived to have occurred. Thus, the term dif cult atic indirect messages, a disrespectful manner, a
captures what patients perceived as negative, failure to individualize, or mismatches between
unhelpful, or poor, without conveying the impli- the nature and quantity of the information that
cation that this situation was associated with poor was received in contrast to what they felt was
practice by a clinician. For additional depth, we required.
drew purposively from the 7 of those 25 (repre-
senting 28 of the original negative group, and Indirect Messages
11 of the cohort) whose characterization of A significant subset of the study participants sus-
their ongoing cancer communication experiences pected or knew they had cancer because of
remained negative 1 year later. Although a major- signals that were inadvertently communicated to
ity of the participants in our study ran into occa- them prior to the actual delivery of the formal
sional problems in communication or encountered diagnosis. For example, during her screening
individual professionals with whom communica- colonoscopy, one woman overheard a nurse
tion seemed unusually difficult, all members of saying, It will be okay, because there is enough
the cohort considered good communication with to resect. Another woman overheard a techni-
their providers as a fundamental requirement of cian in the diagnostic clinic muttering, Oh no
effective cancer care. We found no instances in when reviewing her scans and then walking away.
which an individual whose initial diagnostic A third patient reported being phoned by a recep-
cancer communication experience was positive tionist to book a radiotherapy work-up prior to
arriving at the 1-year mark with a generally nega- being told of his diagnosis or the recommenda-
tive view of the current communication context. tion for radiation treatment. Yet another patient
Thus, careful attention to what we might charac- reported that a relative had been advised to take
terize as communication casualties seems of par- very good care of her because it isn’t good news
ticular significance to understanding the potential while she was in recovery following surgery.
negative impact of communication that misses For some patients, the memory of the diagnos-
the mark in the early stages of the cancer journey. tic consultation was strongly shaped by the
The findings presented here therefore summarize message implied in the visible discomfort, awk-
patterns and themes within descriptions of the wardness, or anxiety of the health care profes-
diagnostic encounters patients found difficult, sional delivering the news. For example, one
and interpret what explains the resolution or con- patient recalled his doctor pacing nervously
tinuation of difficult communications 1 year into throughout the session. He didn’t know what to
the cancer journey. say, you know, he really didn’t. He just said,
Well,’ he says, that’s it.’ I didn’t know what to
The Dynamics of Difficult do and he didn’t really guide me. The sense that
Communications a health care professional charged with delivering
When study participants characterized their diag- news as important as a cancer diagnosis ought to
nostic communication experience as poor, they be in possession of a particular communication
provided detailed illustrations and examples to skill set was shared by most of these patients.
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 557

However, as one recalled, I felt like he was in context within which individual diagnostic mes-
training, that I was a guinea pig to this guy. In sages were received, digested, and interpreted,
another example, a patient remembered the deliv- sometimes with compromising results.
ery of the diagnostic news as a speech or mono-
logue that was delivered by a specialist who Disrespectful or Unprofessional Manner
refused to take questions until the end, by which The most prevalent context in which diagnostic
point the patient had forgotten all of the questions communication was perceived as negative by par-
he might have incrementally raised. One patient ticipants in this study was associated with the
recalled a physician who delivered the news feeling that they were being treated as a case
standing rather than sitting in the manner of his or a number, rather than a unique individual
usual consultation. A third recalled an encounter worthy of respect. As one patient angrily
in which the clinician demanded that she sit, expressed it, I don’t think the system gives a
despite her instinct, in the context of a growing damn about the patients. I really think I am
sense of alarm, that an insistence on standing a number— Blah, blah, blah’—in the cancer
might actually be the only bit of control she charts. I’m totally irrelevant, and not an individ-
might legitimately hold onto in that particular ual at all. Another described the experience of
moment. feeling treated as a piece of meat rather than a
In several instances, patients recalled the indi- person.
rect use of language as a complicating factor in One particularly negative example of a com-
their understanding of their diagnosis. One munication encounter that characterized such
described the physician as pussyfooting around behavior was an instance in which a patient
the diagnosis, whereas another remembered described the physician as being late for the diag-
being informed that she had a tumor without nostic consultation, and then proceeding to accept
being told whether or not it was malignant. A a personal phone call during which the content of
third patient described coming out of the diagnos- the conversation was sufficiently humorous to
tic encounter uncertain as to whether he had lym- occasion audible laughter. The patient perceived
phoma or stomach cancer because the lesions this as a violation of sufficient concern that he
were referred to simply as lumps in your subsequently reported it to the physician’s regula-
stomach. tory body. Other instances in which the affect or
Another kind of indirect message that compro- language of clinicians was less overtly unprofes-
mised the diagnostic communication pertained to sional or dismissive could be perceived as equally
system dynamics and processes. For example, hurtful to patients. One patient described being
one patient was able to interpret from the speed informed of her diagnosis by a surgeon who stood
with which a referral to a specialist was enacted at hospital room door and said, It’s cancer, and
that he most likely had cancer. Another, whose walked away. Feeling rushed during a diagnostic
cancer diagnosis occurred over the summer when consultation was also associated with a percep-
a number of regular members of the health care tion that one was being dismissed or not fully
team were on vacation, experienced the informa- dealt with. One patient recalled a diagnostic
tion as fragmented clues to her impending diag- encounter in which the clinician said, I just want
nosis from pieces shared by different health care you to know it’s not that I’m choosing not to
professionals because no one person was pre- spend time with you but you need to realize we
pared or equipped to tell her the whole story. have very sick patients here. Another reported a
Thus, for these study participants, the overall diagnostic exchange in which he felt the clini-
climate of communication served as a powerful cian’s manner was highly abrupt. From his per-
558 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions

spective, I was just somebody that was expen- the typical patient. For example, a patient who
dable. That’s the way I took it anyway. was a professional statistician resented the
Patients described this feeling of being treated assumption that he would be satisfied with general
as a number or case as an absence of what they statistical information without specific reference
considered normal human connection between to his own demographic or disease features.
them and the health care provider with whom they Others reported what they perceived as standard-
were interacting. For example, one patient ized approaches to how cancer was discussed as
described a particularly negative experience when dismissive of their personal communication style
she was unexpectedly given a diagnosis and and preferences. As one patient cleverly expressed
various details about her cancer although she was this disjuncture, I want the abrupt information
both mentally unprepared and physically exposed but I don’t want it in an abrupt tone
during a diagnostic procedure. As one participant A particular instance of the failure to individu-
theorized, health care professionals who deal with alize approaches during the diagnostic phase was
cancer might believe they need to keep a seal on evident in the various reports we heard from
their emotions because, if they do not, their study participants relative to the use of the tele-
doctor–patient relationships might get too phone to deliver the news that they had cancer.
messy. For these patients, however, failure to For some patients, the willingness of a physician
show compassion or attempting to block the inher- to call them at home with the news was a power-
ent emotionality associated with receiving a ful signal that they should not have to prolong the
cancer diagnosis denied them a human connection anxiety of waiting and that they deserved the
with their health care provider and added signifi- information as soon as it was available. In such
cantly to the distress they were experiencing. instances, patients understood that the alternative
might have been an urgent request that they make
an office appointment, which would have signifi-
Failure to Individualize cantly heightened their anxiety. However, for
Even when communications were considered other participants in this study, a telephone diag-
generally respectful and professional, a number nosis exemplified a distancing mechanism and a
of the negative experiences reported by the study failure to individualize communications. One
participants related to what they perceived as a patient recalled being horrified by a phone
failure to recognize them as unique individuals, message that was left late on a Friday afternoon
different from any other patient, and to tailor while she was at work, breaking the news that she
approaches accordingly. For example, for some had cancer and suggesting she make an appoint-
patients, not being called by their preferred name ment for the following week when she had had
was a symbolic way of communicating that they time to digest the information. Another, who
were in no way memorable or special despite was given her cancer diagnosis directly over the
their situation. For others, lack of recognition of phone during a work day, while she was in the
their full-time employment status and the chal- midst of running a home-operated business, felt
lenge it posed for scheduling consultations the physician had demonstrated no consideration
re ected disinterest in their personal circum- for her particular professional circumstances.
stances. Several patients felt that standardized Thus, the issue for these patients seemed not the
approaches disregarded their particular back- telephone diagnosis in and of itself, but what that
ground and expertise, and believed their special communication approach revealed about the cli-
circumstances warranted communications that nician’s awareness of and sensitivity to their dis-
differed from those that might be appropriate for tinctive needs and circumstances.
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 559

was given far too much information when con-

Inappropriate Nature or Quantity siderable detail about possible treatments was
of Information offered before all her diagnostic test results were
A final major problematic communication theme fully analyzed and interpreted. In her case,
within these accounts of the diagnostic period information-induced anxiety emerged as a
pertained to what patients perceived as mis- byproduct of trying to understand and then
matches between the quality and quantity of reinterpret what would happen over time.
information they needed and what they received. In some instances, patients could point to spe-
Although most patients begin the cancer journey cific information they had been provided that
with relatively little technical information, many turned out to have been factually or interpretively
quickly develop considerable sophistication with incorrect. One man reported being initially told
regard to the complexity associated with a cancer that he most likely had bladder wall cancer, only
diagnosis and related issues including treatment to be later informed that it was actually a stone
decision making and prognostication. In a number and that he was home free. However, he was
of instances, study participants ultimately con- subsequently notified that both guesses were
cluded that they had not been given the complete wrong and that the correct diagnosis was actually
story about their cancer based on subsequent prostate cancer. While he was capable of sympa-
understandings of important information that was thizing with the difficulties posed by differential
available to their health professionals at the time diagnosis, he found the degree of certainty with
of diagnosis. From the perspective of these which the earlier incorrect conclusions had been
patients, in some instances this occurred because put forward made it difficult for him to trust the
the communication competencies of the clinician subsequent recommendations of health care
were inadequate to deliver such news, and in professionals.
others it was because the clinician delivering the A final element associated with the informa-
diagnosis had knowledge deficits. For example, tion provided during the diagnostic encounter
one patient believed that the physician had with- was associated with the degree of hopefulness
held information about the cancer stage, even that was communicated by the clinician to the
though it was available at the time of diagnosis. patient. Although patients acknowledged that a
Another recalled that his family physician cancer diagnosis in which there is a bleak outlook
explained that he did not actually know much is exceedingly difficult to discuss, they were par-
about this type of cancer and therefore simply ticularly affronted by communications that
provided him the name of it on a piece of paper. seemed designed to deprive them of any hope.
Though incomplete information was a distinct One woman, diagnosed with a particularly aggres-
worry for some patients, others experienced prob- sive cancer, vividly recalled asking her physician,
lematic communication related to too much Can you give me any hope As she remem-
information or information that was badly timed bered the encounter, his hasty reply was, That’s
or inappropriate. One patient recalled that, imme- not my area. Even as patients recognized that
diately on being told that the diagnosis was they needed accurate information, many found
cancer, her physician went on to explain that she that they also relied on their clinicians to find the
should not feel guilty and that there were medica- positive spin they needed to face it. Thus, they
tions to treat depression. Another reported being were particularly distressed by health care profes-
told that he ought to be pleased about his particu- sionals whose perpetual approach seemed to be
lar kind of cancer, as it was easier to treat than knocking my spirits down rather than helping
some of the other cancers. A third felt that she them find a way forward.
560 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions

patients were able to move on from the specific

Reflections a Year Later clinician with whom the initial problematic com-
At the 12-month post-diagnosis mark, the patients munication occurred and others were not. Some
in this study were asked to re ect on their experi- patients encountered numerous professionals, in
ence with communications during the diagnostic which case the effective communication of one
period, retelling it from the perspective of time might have buffered the effects of problematic
and experience. We were particularly interested communication with another. Other patients
in comparing these retrospective accounts with might have had options with regard to switching
those acquired in the more immediate diagnostic services to obtain a better communication match.
timeframe to determine whether they surfaced Thus, a progression with regard to the cancer
any substantive reinterpretation of what effective communication experience 1 year later might not
and problematic communication entailed, or re ect a change in the dynamic of any one
whether elements viewed with a particular valence patient–professional dyad, but rather changing
in the initial experience might have been reframed circumstances and opportunities within a com-
based on knowledge gained with experience over prehensive system of care. Furthermore, regard-
time. In these accounts, we detected various shifts less of the generally positive or negative
in the precision of details and the intensity of communication experience of any individual
emotion associated with recall. As one patient study participant, a majority reported both posi-
described it, tive and negative individual encounters along the
There’s a sort of haziness of memory about the actual
way. Nevertheless, it is informative to consider
sequence of events but the feelings that I felt at the how patients whose communication experiences
time are still pretty clear, you know. I felt concerned did and did not improve accounted for their
and worried. And you’re so thirsty for as much evolving circumstances.
information as I possibly could get, but also cognizant For 7 study participants, the early diagnostic
of the fact that I had limited ability to take it in.
experiences characterized as predominantly dif-
In most instances, however, the initial and ficult were recalled as similarly problematic 12
12-month accounts were remarkably similar in months later. This group included 2 men and 5
their content, tone, and quality, even to the point women, aged 49 to 66 years (mean = 55.7), with
of identical wording in key passages, as if they a range of tumor sites (two lung, two colorectal,
had been the source of considerable rumination, and one each of breast, prostate, and leukemia).
rendered repeatedly in accounts to others, and Among them, 3 had undergone surgery, 3 chemo-
committed to memory. This strong emotional therapy, and 2 radiation therapy, with 1 having
tone, even 1 year later, reveals the importance of had no treatment. The initial diagnosis was deliv-
these early communication experiences in the tra- ered to them by their own family physician (in 4
jectory of the patient’s cancer journey. cases), a locum covering for the usual general
Following diagnosis, cancer patients enter a practitioner (1), a specialist (2), or a radiologist
trajectory that might involve various health care (1). Over the course of the initial year, 4 had
professionals, treatments, and care contexts, discontinued the relationship with the initial clini-
depending on such factors as their disease type cian they described as problematic. All of these
and stage, and the geographical location in which patients had encountered multiple health care
they lived. In some instances, their cancer was professionals over this period, and 5 described
managed by clinicians other than those who pro- themselves as very happy with the communica-
vided the diagnosis, and in others, the primary tion of at least one of their specialist practitioners.
management remained fairly constant. Thus some However, although only 1 patient characterized
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 561

communication with all of her health care profes- considerably more attention paid to the value of
sionals as problematic, all 7 were continuing to effective communication as well as skill with
receive at least some of their care within a com- enacting it among cancer care specialists com-
munication context that they characterized as pared to the generalist health care professionals
poor. When asked to explain why they thought who might have been involved during the diag-
this to be the case, most cited the continuing sense nostic period. Thus, the switch to clinicians with
that they were not being treated as individuals by whom communication was more effective
the clinicians overseeing their care. Though this occurred as a matter of course within the care
group was reasonably certain that their clinical trajectory for some patients. In other cases,
care was technically competent, some described patients felt that it simply took time to develop
feeling highly frustrated that they had never rapport, and found that they were able to respond
received the compassion they felt they deserved, differently to particular clinicians once they better
and reported a litany of continuing communica- understood how to engage with them and inter-
tion difficulties, including misinformation, intim- pret their particular communication approaches.
idation, and distrust that tainted aspects of their Among this group of patients whose commu-
life quality with cancer. Others seemed more nications had progressed from difficult to gener-
resigned to their situation, indicating that they felt ally positive, a majority attributed this change to
they had learned to tolerate problematic commu- their own capacity to take active control of the
nication and accept that the particular clinician communication circumstances. For example, in
overseeing their care was incapable of anything some cases, they sought alternative opinions or
better. changed health care professionals. Some study
The remaining 18 of the original 25 whose participants reported actively avoiding certain cli-
communications were initially characterized as nicians, through such mechanisms as refusing to
predominantly difficult had resolved that situa- be seen by trainees or by clinicians of a certain
tion and reported their communications to be gen- gender. Others described themselves as actively
erally positive at the 1-year point. This group petitioning to be referred to particular clinicians
included 14 women and 4 men, aged 35 to 81 who came well recommended by fellow patients.
years (mean = 54.4), whose tumor sites repre- A number of these participants also took
sented breast (8), lymphoma (4), prostate (3), control in advocating for themselves in terms of
gynecological (2), and colorectal (1) cancer. All their own communication needs. For example,
but 1 had undergone cancer treatment, including one patient, who had found her oncologist’s
13 who had surgery, 9 chemotherapy, and 4 radia- serious demeanor highly intimidating, took the
tion therapy. The initial diagnosis had been deliv- initiative of some innocuous joking with him as
ered to them by an oncologist (5), a radiologist a strategic way to set the course for the kind of
(4), a surgeon (3), another specialist (3), or a ongoing relationship she felt she needed. From
general practitioner (2). When asked to account her perspective, her success with this maneuver
for this change over the course of the year, they made their subsequent communications more
described factors relating to their own learning lighthearted and hopeful. Other patients claimed
and agency as well as taking advantage of the they had developed an approach of advocating on
opportunity to engage with clinicians who were their own behalf by repeatedly reminding their
more skilled communicators. For a number of health care professionals of their particular com-
these patients, communication improved imme- munication needs. A significant number of these
diately on shifting the locus of care to the study participants referenced having taught their
oncology clinical specialty service. Many noted clinicians how to engage with them in a manner
562 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions

that was supportive of their need for hope. As one to be seen by the cancer specialist who will deter-
patient explained, I mean let’s be truthful, it’s mine the course of action. The first consultation
pretty grim. But I’ll say to them, I need hope with that specialist becomes a time in which treat-
because I’m such a believer that hope will increase ment plans must be formulated on the basis of the
my chances.’ As another reiterated, I read a lot cancer, any comorbidities, and distinctive cir-
of body language. And they can tell me some- cumstances. At the same time, the specialist is
thing that might be taken as bad news, but if they charged with attempting to establish a rapport
have a bit of a smile or a twinkle in their eye when with the patient, learning about their concerns
they’re talking to me, that gives me hope. and needs, understanding their expectations, and
Thus, for the participants in this study, com- ascertaining how they would prefer to receive and
munication was a powerfully in uential element titrate information. From the patient perspective,
of their initial cancer care experience and remained these encounters become part of the momentum
important 1 year later. For those whose initial that is starting to build, sweeping them up in a
communication experiences were difficult, the biological, psychosocial, emotional, and experi-
impact of those experiences was lasting. Fortu- ential turmoil. In this context, complex concepts
nately, many were able, by circumstance or design, are to be understood and decisions made that
to resolve the negative communication for the might well have a profound in uence on the
most part and to find at least one point of effective remainder of their lives.
communication within the system. Their insights This situational context helps us appreciate
into the necessity of ensuring that their communi- why the diagnostic period plays a particularly
cation was as effective as possible reinforce the meaningful role in creating the foundation for a
power of the communication encounter to shape cancer journey that is characterized by either sup-
the tone and structure of the cancer experience. portive or compromised communication (Epstein
Street, 2007). As cancer care evolves, the
DISCUSSION agenda for many subsequent clinical consulta-
Although many of the communication difficulties tions often becomes more problem- or treatment-
articulated by these patients as compromising focused. By the time there might be subsequent
their experience during the cancer diagnosis bad news, such as regarding metastastic sic
period are not dissimilar to the kinds of difficult disease, many specialists will have formed good
communication patterns that have been reported working relationships with their patients. Thus,
elsewhere in relation to cancer care in general although these subsequent bad news occasions
(Arora, 2003 Hack et al., 2005), these findings remain times of particular vulnerability to poor
orient us to the particularities of these early, initial communication, many patients experience them
encounters and reveal the need to thoughtfully within a relational context that is more solidly
consider their potential impact on the journey on established than was the case during their initial
which the patient is embarking. The initial diag- diagnostic experience.
nosis period stands out for patients as the moment, Vivid recall and emotionality associated with
frozen in time, during which their worst fears are poor communication in the initial diagnostic
realized (Walsh Nelson, 2003). In many period remained remarkably fresh 12 months
instances, the clinician delivering that initial news later for these patients. This was especially the
is not a cancer specialist, and is incapable of case for the subset whose communication expe-
answering the urgent questions that patients inev- riences could be characterized overall as nega-
itably have. Following the moment of first being tive. For those unfortunate individuals unable
informed one has cancer, there is typically a wait to negotiate a more positive communication
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 563

environment, whether through circumstance or certain cancer care clinicians whose ways of
capacity for self-efficacy, the burden of continu- engaging helped them envision what was possi-
ing communication problems added to their dis- ble. Thus, though preventing communication
tress and deprived them of an important source problems remains a high priority, it would seem
of comfort. However, except for a small minority that intervening after they have occurred could be
of the participants in our study who were unable another powerful tool for building a base toward
to fully resolve their communication challenges, future sensitive communication occasions that
most did find their way forward into a cancer might unfortunately become necessary in the
care environment within which they had at least course of the cancer illness. Furthermore, the
some access to helpful and effective communica- accounts confirmed that although some initially
tion. This suggests that, although the initial prob- problematic patient–provider relationships can
lematic communication might have set the stage indeed be rescued, there might be others for
for future difficulties, there were also multiple which it is feasible and perhaps preferable to
ways in which a communication rescue could change clinicians and start over.
occur. For many patients, this came about natu- Among the many features of difficult commu-
rally in the course of encountering multiple pro- nication that the participants in this study articu-
fessionals, especially oncology specialists and lated and that have been reported in the literature,
allied health professionals with a specialty in negotiations around hope perpetually stand out as
cancer care. For other patients, it required active the single most challenging and troublesome bat-
intervention on their part to strategize and tleground (Evans et al., 2006 Hack et al., 2005).
develop ways of motivating the specific profes- Although information, decisional guidance, com-
sionals involved in their care toward more con- passion, and support feature strongly in the sub-
structive communications. And in a significant stance of cancer care communication across
number of cases, communication became in- the spectrum, the spectre of hope re ects a per-
creasingly effective because of a combination vasive complicating dimension (Thorne, Oglov,
of both professional and patient strategic Armstrong, Hislop, 2007). Patients typically
engagement. report the denial of hope as the most serious vio-
Although the value of communication in lation inherent in poor cancer communication
cancer care might not have been self-evident (Clayton, Butow, Arnold, Tattersall, 2005). Cli-
prior to their diagnosis, it was strongly recog- nicians note the profound complexity inherent in
nized thereafter by all of the patients in this study. their obligation to provide patients with the level
Their accounts revealed accessing an impressive and detail of information to which they have a
array of suggestions and techniques for improv- legal and ethical right, while at the same time
ing their communication experiences from such supporting their emotional well-being (Salander,
resources as patient support services, Internet Bergenheim, Bergstrom, Henriksson, 1998
chat sites, mass media consumer advice, and their Sardell Trierweiler, 1993). Fostering unrealis-
own social networks. Clearly, the presence of a tic hope is commonly understood as incompetent
powerful network of community-based advice practice. The clinician is inevitably caught
and support was instrumental to the development between the dual requirements of providing accu-
of ideas and options, especially toward prevent- rate information about the seriousness of the
ing future problems in their communications with bodily condition and concurrent hopefulness for
cancer care providers. Additionally, many attrib- the patient’s personal well-being, without inad-
uted important experiential learning associated vertently misleading the patient or causing addi-
with effective communication to exposure to tional harm (Hagerty et al., 2005 Leydon, 2008).
564 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions

In the context of newly formed patient– a significant iatrogenic problem within cancer
provider relationships, when the emotional shock care (Parker et al., 2005).
is profound and the information to be processed
complex, the diagnostic period becomes a phase CONCLUSIONS
within the cancer trajectory when supportive Although considerable knowledge transfer effort
approaches to ensure optimal communication has been directed toward preventing poor com-
seem most acutely needed. In this context, diag- munication in cancer care, the problematic prac-
nostic and prognostic information are most easily tice patterns among a subset of clinicians have
con ated in the consideration of what to expect proven astonishingly difficult to eradicate (Epstein
and what can be hoped for. The initial diagnostic Street, 2007 Thorne et al., 2005) analyses of
phase therefore seems a situation for which ade- both basic and continuing education communica-
quate time and support ought to be paramount tions training outcomes suggest modest results at
planning concerns. It is well recognized that most best (Butler et al., 2005 Fallowfield Jenkins,
cancer care systems are under severe stress 2006 Fellowes, Wilkinson, Moore, 2004).
(Canadian Strategy for Cancer Control, 2006), Thus, studies such as this from the cancer patient
and equitable apportioning of scarce expert perspective provide a valid basis for considering
resources precludes the kind of time and access alternative approaches to both strengthening
to medical specialists that most patients would general awareness of the inherent value of com-
optimally prefer (Epstein Street, 2007 Thorne, munication across the cancer care system and
Hislop, Stajduhar, Oglov, 2008). Although addressing the sequelae of problematic commu-
many efforts are under way to minimize undue nication when it occurs. In the current context,
waiting times in the initial consultation context, effective and engaged interprofessional cancer
delays are likely inevitable. It is therefore unreal- care teams, capitalizing on the distinct com-
istic to expect that medical specialists within petencies of a range of professionals, working
existing cancer system structures can accommo- collaboratively with patient networks and con-
date the full range of communication and sup- sumer advocacy systems, seems a highly promis-
portive needs of newly diagnosed patients. Thus, ing direction for advancement. Aligning the
it seems time to consider better ways to capitalize communication challenge with the larger policy
on the communication resources that reside mandate of a fully coordinated cancer care
within the allied health professions, especially system, we can advance services at a population
advanced practice nurses (Ambler et al., 2007 level while continuing to be mindful of the very
Gentry Sein, 2007 Jarrett Payne, 2000 real needs of the individuals we are privileged to
Skrutkowski et al., 2008), as well as the expand- serve.
ing array of available patient/consumer advocacy
networks (Decter Grosso, 2006) and Internet DECLARATION OF
information/support platforms (Decter Grosso, CONFLICTING INTERESTS
2006 Intille abinski, 2005 Weiss Lorenzi, The authors declared no con icts of interest with
2005). Ensuring that patients have a range of respect to the authorship and/or publication of
options to communicate their needs and concerns, this article.
with the proviso that there are effective mecha-
nisms for aligning the insights and issues that FUNDING
arise from these communications into the care The authors disclosed receipt of the following
management process, seems a logical priority for financial support for the research and/or author-
preventing and ameliorating what we know to be ship of this article: Funding was provided by
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 565

research grant 74545 from the Canadian Insti- Cox, A., Jenkins, V., Catt, S., Langridge, C.,
tutes of Health. Fallowfield, L. (2006). Information needs and
experiences: An audit of K cancer patients.
European ournal of ncology Nursing, (4),
263-272.
REFERENCES Davis, T. C., Williams, M. V., Marin, E., Parker, R. M.,
Ambler, N., Rumsey, N., Harcourt, D., Khan, F., Glass, J. (2002). Health literacy and cancer
Cawthorn, S., Barker, J. (2007). Specialist nurse communication. CA: A Cancer ournal for Clini-
counsellor interventions at the time of diagnosis of cians, (3), 134-149.
breast cancer: Comparing advocacy with a Davison, R., Mills, M. E. (2005). Cancer patients’
conventional approach. ournal of Advanced Nursing, satisfaction with communication, information and
(3), 445-453. quality of care in the K region. European ournal
Arora, N. K. (2003). Interacting with cancer patients: of Cancer Care, , 83-90.
The significance of physicians’ communication Decter, M., Grosso, F. (2006). Navigating Canada s
behavior. Social Science and Medicine, , health care: A user guide to getting the care you
791-806. need. Toronto, ON, Canada: Penguin.
Back, A. L., Arnold, R. M., Baile, W. F., Tulsky, J. A., Edvardsson, T. M., Pahlson, A., Ahlstrom, G. P.
Fryer-Edwards, K. (2005). Approaching difficult (2006). Experiences of onset and diagnosis of
communication tasks in oncology. CA: A Cancer low-grade glioma from the patient’s perspective.
ournal for Clinicians, (3), 164-177. Cancer Nursing, (5), 415-422.
Baile, W. F., Aaron, J. (2005). Patient–physician Eggly, S., Penner, L. A., Greene, M., Harper, F. W. K.,
communication in oncology: Past, present, and future. Ruckdeschel, J. C., Albrecht, T. L. (2006).
Current pinion in ncology, (4), 331-335. Information seeking during bad news oncology
Beach, W. A., Easter, D. W., Good, J. S., Pigeron, E. interactions: uestions asked by patients and their
(2005). Disclosing and responding to cancer fears companions. Social Science and Medicine, (11),
during oncology interviews. Social Science and 2974-2985.
Medicine, , 893-910. Epstein, R. M., Street, R. L. (2007). Patient-centered
Butler, L., Degner, L., Baile, W., Landry, M., SCRN communication in cancer care: Promoting healing
Communication Team. (2005). Developing communi- and reducing suffering. NIH publication No. 07-6225.
cation competency in the context of cancer: A critical Bethesda, MD: National Institutes of Health.
interpretive analysis of provider training programs. Evans, W. G., Tulsky, J. A., Back, A. L., Arnold,
Psycho- ncology, (10), 861-872. R. M. (2006). Communication at times of transitions:
Butow, P. (2005). The communication goals and needs How to help patients cope with loss and redefine
of cancer patients: A review. Psycho- ncology, , hope. Cancer ournal, (5), 417-424.
846-847. Fallowfield, L., Jenkins, V. (2004). Communicating
Canadian Strategy for Cancer Control. (2006). The sad, bad, and difficult news in medicine. ancet, ,
Canadian strategy for cancer control: A cancer plan 312-319.
for Canada. Retrieved January 2, 2008, from http:// Fallowfield, L., Jenkins, V. (2006). Current concepts
www.cancer.ca/vgn/images/portal/ of communication skills training in oncology. Recent
cit 86751114/10/2/1404842209cw CSCC Results in Cancer Research, , 105-112.
Discussion Paper July 2006 v2.pdf Fallowfield, L. J. (2008). Treatment decision-making in
Carlson, L. E., Feldman-Stewart, D., Tishelman, C., breast cancer: The patient–doctor relationship. Breast
Brundage, M. D., SCRN Communication Team. Cancer Research Treatment, (Suppl 1), 5-13.
(2005). Patient–professional communication research Feldman-Stewart, D., Brennenstuhl, S., Brundage,
in cancer: An integrative review of research methods M. D. (2007). A purpose-based evaluation of information
in the context of a conceptual framework. Psycho- for patients: An approach to measuring effectiveness.
ncology, , 812-828. Patient Education and Counseling, (3), 311-319.
Clayton, J. M., Butow, P. N., Arnold, R. M., Feldman-Stewart, D., Brundage, M. D., Tishelman, C.,
Tattersall, M. H. (2005). Fostering coping and SCRN Communication Team. (2005). A conceptual
nurturing hope when discussing the future with framework for patient–professional communication:
terminally ill cancer patients and their caregivers. An application to the cancer context. Psycho-
Cancer, (9), 1965-1975. ncology, (10), 801-809.
566 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions

Fellowes, D., Wilkinson, S., Moore, P. (2004). Nelson, C. M. (2004). The needs of recently diagnosed
Communication skills training for health care cancer patients. Nursing Standard, (13), 42-44.
professionals working with cancer patients, their Ong, L. M., Visser, M. R., Lammes, F., de Haes, J. C.
families and/or carers. Cochrane Database of (2000). Doctor–patient communication and cancer
Systematic Revie s, Issue , Art No: CD patients’ quality of life and satisfaction. Patient
D I: CD pub . Education and Counseling, (2), 145-156.
Gentry, S., Sein, E. (2007). Taking the wheel. Parker, P. A., Davison, B. J., Tishelman, C.,
Oncology nurses help patients navigate the cancer Brundage, M. D. (2005). What do we know about
journey. NS Connect, (3), 8-11. facilitating patient communication in the cancer care
Glaser, B. G., Strauss, A. L. (1967). The discovery of setting Psycho- ncology, , 848-858.
grounded theory: Strategies for ualitative research. Patrick, K., Intille, S. S., abinski, M. F. (2005). An
Chicago: Aldine. ecological framework for cancer communication:
Hack, T. F., Degner, L. F., Parker, P. S., SCRN Implications for research. ournal of Medical Internet
Communication Team. (2005). The communication Research, (3), e23.
goals and needs of cancer patients: A review. Ptacek, J. T., Ptacek, J. J. (2001). Patients’ percep-
Psycho- ncology, (10), 831-845. tions of receiving bad news about cancer. ournal of
Hagerty, R. G., Butow, P. N., Ellis, P. M., Lobb, E. A., Clinical ncology, (21), 4160-4164.
Pendlebury, S. C., Leighl, N., et al. (2005). Commu- Ramirez, A. G. (2003). Consumer–provider communica-
nicating with realism and hope: Incurable cancer tion research with special populations. Patient
patients’ views on the disclosure of prognosis. Education and Counseling, (1), 51-54.
ournal of Clinical ncology, (6), 1278-1288. Rutten, L. J. F., Arora, N. K., Bakos, A. D., Aziz, N.,
Hoff, L., Tidefelt, ., Thaning, L., Hermeren, G. Rowland, J. (2005). Information needs and sources of
(2007). In the shadow of bad news—Views of information among cancer patients: A systematic
patients with acute leukaemia, myeloma or lung review of research (1980-2003). Patient Education
cancer about information, from diagnosis to cure or and Counseling, (3), 250-261.
death. BMC Palliative Care, (1), doi:10.1186/1472- Salander, P., Bergenheim, A. T., Bergstrom, P.,
684 -6-1. Henriksson, R. (1998). How to tell cancer patients:
Intille, P., abinski, M. F. (2005). An ecological A contribution to a theory of communicating the
framework for cancer communication: Implications diagnosis. ournal of Psychosocial ncology, (2),
for research. ournal of Medical Internet Research, 79-93.
(3), e23. Retrieved January 9, 2009, from http:// Sandelowski, M. (2000). Whatever happened to
www.jmir.org/2005/3/e23/ qualitative description Research in Nursing
Jarrett, N., Payne, S. (2000). Creating and maintain- ealth, , 334-340.
ing optimism in cancer care communication. Sandelowski, M., Trimble, F., Woodard, E. K.,
International ournal of Nursing Studies, , 81-90. Barroso, J. (2006). From synthesis to script: Trans-
Kreps, G. L., Arora, N. K., Nelson, D. E. (2003). forming qualitative research findings for use in
Consumer/ provider communication research: practice. ualitative ealth Research, ,
Directions for development. Patient Education and 1350-1370.
Counseling, (1), 3-4. Sardell, A. N., Trierweiler, S. J. (1993). Disclosing
Leydon, G. (2008). Yours is potentially serious but the cancer diagnosis: Procedures that in uence patient
most of these are cured : Optimistic communication hopefulness. Cancer, (11), 3355-3365.
in K outpatient oncology consultations. Psycho- Schofield, P. E., Beeney, L. J., Thompson, J. F., Butow,
ncology, , 1081-1088. P. N., Tattersall, M. H. N., Dunn, S. M. (2001).
Lincoln, Y. S., Guba, E. G. (1985). Naturalistic Hearing the bad news of a cancer diagnosis: The
in uiry. Beverly Hills, CA: Sage. Australian melanoma patient’s perspective. Annals of
Lockhart, K., Dosser, I., Cruickshank, S., Kennedy, ncology, (3), 365-371.
C. (2007). Methods of communicating a primary Schofield, P. E., Butow, P. N. (2004). Towards better
diagnosis of breast cancer to patients systematic communication in cancer care: A framework for
review . Cochrane Database of Systematic Revie s developing evidence-based interventions. Patient
, CD . Education and Counseling, (1), 32-39.
Miles, M. B., Huberman, A. M. (1994). ualitative Schofield, P. E., Butow, P. N., Thompson, J. F.,
data analysis (2nd ed.). Thousand Oaks, CA: Sage. Tattersall, M. H. N., Beeney, L. J., Dunn, S. M.
 APPENDIX D Patient Real-Time and 12-Month Retrospective Perceptions 567

(2003). Psychological responses of patients receiving interpretive description of patient preferences.

a diagnosis of cancer. Annals of ncology, (1), European ournal of Cancer Care.
48-56. Thorne, S., Reimer Kirkham, S., MacDonald-Emes,
Siminoff, L. A., Graham, G. C., Gordon, N. H. J. (1997). Interpretive description: A non-categorical
(2006). Cancer communication patterns and the qualitative alternative for developing nursing
in uence of patient characteristics: Disparities in knowledge. Research in Nursing ealth, (2),
information-giving and affective behaviors. Patient 169-177.
Education and Counseling, (3), 355-360. Thorne, S. E. (2008). Interpretive description. Walnut
Skrutkowski, M., Saucier, A., Eades, M., Swidzinski, Creek, CA: Left Coast Press.
M., Ritchie, J., Marchionni, C., et al. (2008). Impact Thorne, S. E., Bultz, B. D., Baile, W. F., SCRN
of a pivot nurse in oncology on patients with lung or Communication Team. (2005). Is there a cost to poor
breast cancer: Symptom distress, fatigue, quality of communication in cancer care : A critical review of
life, and use of health-care resources. ncology the literature. Psycho- ncology, , 875-884.
Nursing orum, (6), 948-954. Walsh, D., Nelson, K. A. (2003). Communication
Taylor, E., Ismail, S., Hills, H., Ainsworth, S. (2004). of a cancer diagnosis: Patients’ perceptions of
Multi-component Psychosocial support for newly when they were first told they had cancer.
diagnosed cancer patients: participants’ views. American ournal of ospice Palliative Care,
International ournal of Palliative Nursing, (6), (1), 52-56.
287-295. Watson, D. E., Mooney, D., Peterson, S. (2007).
Thorne, S., Hislop, T., Stajduhar, K., Oglov, V. Patient e periences ith ambulatory cancer care in
(2008). Time-related communication skills from the British Columbia, . Retrieved July 20, 2008,
cancer patient perspective. Psycho- ncology, from http://www.chspr.ubc.ca/files/publications/2007/
Pre-press: D I: pon . chspr 07 4R.pdf
Thorne, S., Oglov, V., Armstrong, E.-A., Hislop, Weiss, J., Lorenzi, N. (2005). nline communication
T. G. (2007). Prognosticating futures and the human and support for cancer patients: A relationship-centric
experience of hope. Palliative and Supportive Care, design frame or . Paper presented at the Proceedings
(3), 227-239. of American Medical Informatics Association Annual
Thorne, S., Oliffe, J., Kim-Sing, C., Hislop, T. G., Symposium, pp. 799-803. Retrieved January 9, 2009,
Stajduhar, K., Harris, S. R., et al. (in press). Helpful from: http://www.pubmedcentral.nih.gov/
communications during the diagnostic period: An articlerender.fcgi artid 1560480
Glossary
A ani al rights Guidelines used to protect the rights of
a riori From Latin, meaning the former that is, animals in the conduct of research.
before the study or analysis. anony ity A research participant’s protection in a
abstract A brief, comprehensive summary of a study at study so that no one, not even the researcher, can link
the beginning of an article. the subject with the information given.
accessible o ulation A population that meets the antecedent variable A variable that affects the
population criteria and is available. dependent variable but occurs before the introduction
accuracy The characteristic of all aspects of a study of the independent variable.
systematically and logically following from the assent An aspect of informed consent that pertains to
research problem. protecting the rights of children as research
a ter only design An experimental design with two participants.
randomly assigned groups: a treatment group and a assu tions Accepted truths, key concepts and ideas,
control group. This design differs from the true reasons and justifications, supporting examples,
experiment in that both groups are measured only parallel experiences, implications and consequences,
after the experimental treatment. Also known as and any other structural features of the written text
posttest-only control group design used to interpret and assess it accurately and fairly.
a ter only none uivalent control grou design A attrition The loss of a subject from a study between
quasiexperimental design similar to the after-only time 1 data collection and time 2 data collection. Also
experimental design except that participants are not called mortality
randomly assigned to the treatment group or the auditability The characteristic of a qualitative study,
control group. developed by the investigator’s research process, that
ai o in uiry The goals or specific objectives of the allows another researcher or a reader to follow the
research, which vary with the paradigm. thinking or conclusions of the investigator.
al ha Considered an a priori probability because it is
set before the data are collected. Also considered a
conditional probability because the null hypothesis is behavioural aterialist ers ective In ethnographical
assumed to be true. studies, the observation of culture through a group’s
al ha coe cient See reliability coef cient patterns of behaviour and customs, its way of life,
alternate or reliability A reliability measure in and what it produces.
which two or more alternate forms of a measure are bene cence An obligation to do no harm and to
administered to the same participants at different maximize possible benefits.
times. The scores of the two tests determine the bene ts Potential positive outcomes of participation in
degree of relationship between the measures. Also a research study.
called parallel-form reliability bias A distortion in the interpretation of the results of
analysis The division of the content into parts to the data analysis.
understand each aspect of the study. biological easure ent se of specialized equipment
analysis o covariance ANCO A A statistic that to determine the biological status of participants in a
measures differences among group means and uses a study.
statistical technique to equate the groups under study oolean o erator In a literature search, the word that
in relation to an important variable. defines the relationships between words or groups of
analysis o variance ANO A A statistic that tests words for example, AND, OR, NOT, or
whether group means differ from each other instead NEAR.
of testing each pair of means separately, ANOVA brac eting A process by which the researcher identifies
considers the variation among all groups. personal biases about the phenomenon of interest to
568
 Glossary 569

clarify how personal experience and beliefs may conce t An image or symbolic representation of an
colour what is heard and reported. The term comes abstract idea.
from the mathematical metaphor of putting brackets conce tual de nition The general meaning of a
around our beliefs so they can be put aside. concept.
conce tual ra e or A structure of concepts,
C theories, or both that is used to construct a map
case study The study of a selected phenomenon that for the study.
provides an in-depth description of its dimensions concurrent validity The degree of correlation between
and processes. two measures of the same concept that are
case study ethod The study of a selected administered at the same time.
contemporary phenomenon over time to provide an conduct o research The analysis of data collected
in-depth description of the essential dimensions and from a homogeneous group of participants who meet
processes of the phenomenon. study inclusion and exclusion criteria for the purpose
chance error An error attributable to uctuations in of answering specific research questions or testing
subject characteristics that occur at a specific point in specified hypotheses.
time and are often beyond the awareness and control con dence interval An estimated range of values,
of the examiner an error that is difficult to control, which are likely to include an unknown population
unsystematic, and unpredictable and thus cannot be parameter calculated from a given set of sample data.
corrected. Also called random error Abbreviated CI
chi s uare χ A nonparametric statistic used to con dentiality Assurance that a research participant’s
determine whether the frequency found in each identity cannot be linked to the information that was
category is different from the frequency that would provided to the researcher.
be expected by chance. consent Agreement to participate in a study. See
citation anage ent so t are A software program informed consent
that formats and stores the researcher’s citations so consistency An aspect of the data collection process
that they are available for electronic retrieval. that requires that data be collected from each subject
clinical uestion An inquiry that is the basis of in the study in exactly the same way or as close to
evidence-informed practice. A clinical question the same way as possible.
concerns five components: population, intervention, constancy An aspect of control in data collection that
comparison, outcome, and time (PICOT). ensures that methods and procedures of data
closed ended ite A question that the respondent may collection are the same for all participants that is,
answer with only one of a fixed number of choices. each participant is exposed to the same
cluster sa ling A probability sampling strategy that environmental conditions, timing of data collection,
involves successive random sampling of units. The data collection instruments, and data collection
units sampled progress from large to small. Also procedures.
known as multistage sampling constant co arative ethod In the grounded theory
codes Tags or labels that are assigned to themes in a method, a process of continuously comparing data as
qualitative study. they are acquired during research.
coding The progressive marking, sorting, resorting, and constant error See systematic error
defining and redefining of the collected data. construct validity The extent to which a test measures
cognitive ers ective In ethnographical studies, the a theoretical construct or trait.
view that culture consists of beliefs, knowledge, and constructivis The basis for naturalistic (qualitative)
ideas people use as they live. research, which developed from writers such as
cohort The participants of a specific group that are Immanuel Kant, who sought alternative ways of
being studied. thinking about the world a belief that reality is not
co unity based artici atory research A method fixed but rather is a construction of the people
by which the voice of a community is systematically perceiving it.
accessed in order to plan context-appropriate constructivist aradig The basis of most qualitative
action. research, which is concerned with the ways in which
conceal ent An observational method that refers to people construct their worlds.
whether or not the participants know that they are consu er A person whose activity uses and applies
being observed. research.
570 Glossary

content analysis A technique for the objective, critical social thought A philosophical orientation that
systematic, and quantitative description of suggests that reality and a person’s understanding of
communications and documentary evidence. reality are constructed by people with the most power
content validity The degree to which the content of the at a particular point in history.
measure represents the universe of content or the critical thin ing The rational examination of ideas,
domain of a given behaviour. inferences, assumptions, principles, arguments
conte t The personal, social, and political environment conclusions, issues, statements, beliefs, and actions.
in which a phenomenon of interest (time, place, criti ue The process of objectively and critically
cultural beliefs, values, and practices) occurs. evaluating the content of a research report for
conte t de endent Condition in which the meaning of scientific merit and application to practice, theory, or
an observation is defined by its circumstance or the education.
environment. criti uing criteria The standards, appraisal guides, or
contrasted grou s a roach A method used to assess questions used for objectively and critically
construct validity. A researcher identifies two groups evaluating a research article.
of individuals who are suspected of having either an Cronbach s al ha A test of internal consistency in
extremely high or an extremely low score on a which each item in a scale is simultaneously
characteristic scores from the groups are obtained compared with all others.
and examined for sensitivity to the differences. Also cross sectional study Nonexperimental research in
called no n-groups approach which data at one point in time—that is, in the
control The measures used to hold uniform or constant immediate present—are examined.
the conditions in a research study. culture The system of knowledge and linguistic
control grou The group in an experimental expressions used by social groups that allows the
investigation that does not receive the experimental researcher to interpret or make sense of the world
intervention or treatment the comparison group. the structures of meaning through which people shape
controlled vocabulary A selected list of words and experiences.
phrases that are applied to similar pieces of m la i e nde o r ing and llied eal h
information units (e.g., life skills). i era re CINA L A print and computerized
convenience sa ling A nonprobability sampling database computerized CINAHL is available on
strategy in which the most readily accessible persons CD-ROM and online.
or objects serve as participants or participants of a
study.
convergent validity A type of construct validity in data Information systematically collected in the course
which two or more tools that theoretically measure of a study the plural of datum.
the same construct are positively correlated. data dis lay Compression and organization of data that
correlation The degree of association between two promotes understanding and visualization and enables
variables. conclusions to be drawn.
correlational study A type of nonexperimental research data reduction The process of selecting and
that examines the relationship between two or more transforming the data from field notes or
variables. transcriptions.
credibility A characteristic of qualitative research that data saturation A point when the information
refers to the accuracy, validity, and soundness of data. collected by the researcher becomes repetitive
criterion related validity The degree of relationship ideas conveyed by the participant have been
between performance on the measure and the actual shared previously by other participants, and
behaviour, either in the present (concurrent) or in the inclusion of additional participants does not result
future (predictive). in new ideas.
critical reading An active interpretation and objective debrie ng The opportunity for researchers to discuss
assessment of an article, during which the reader is the study with the participants and for participants to
looking for key concepts, ideas, and justifications. refuse to have their data included in the study.
critical social theory The use of both qualitative and deductive Concluded from data.
quantitative research to highlight historical and deductive reasoning A logical thought process in
current experiences of suffering, con ict, and which hypotheses are derived from theory reasoning
collective struggles. moves from the general to the particular.
 Glossary 571

degree o reedo The number of quantities that are e ide iological study Examination of factors affecting
unknown minus the number of independent equations the health and illness of populations in relation to the
linking these unknowns a function of the number in environment.
the sample. Abbreviated df e iste ology The theory of knowledge the branch of
deli itations Characteristics that restrict the population philosophy concerned with how people know what
to a homogeneous group of participants. they know, or what is known to be truth.
de endent variable In experimental studies, the e uivalence Consistency or agreement among
presumed effect of the independent or experimental observers using the same measurement tool, or
variable on the outcome. Variation in the independent agreement among alternative forms of a tool.
variable changes this effect. The dependent variable error variance The extent to which the variance in test
is observed but not manipulated. scores is attributable to error rather than to a true
descri tive e loratory survey A type of measure of behaviours.
nonexperimental research in which descriptions of ethics The theory or discipline dealing with principles
existing phenomena are collected for the purpose of of moral values and moral conduct.
using the data to justify or assess current conditions ethnogra hic ethod A method that scientifically
or to make plans for improvement of conditions. describes cultural groups. The goal of the
descri tive statistics Statistical details used to describe ethnographer is to understand the natives’ view of
and summarize sample data. their world.
design The plan or blueprint for conduct of a study. ethnogra hy A qualitative research approach designed
develo ental study A type of nonexperimental to produce cultural theory. Also called ethnographic
research that is concerned not only with the existing research
status and interrelationships of phenomena but also etic ers ective An outsider’s view of another’s world.
with changes that occur as a function of time. evaluation The process of determining the value of
directional hy othesis A hypothesis that specifies the data.
expected direction of the relationship between evaluation research The use of scientific research
independent and dependent variables. methods and procedures to evaluate a program,
disse ination The communication of research treatment, practice, or policy outcomes the analytical
findings. means used to document the worth of an activity.
divergent validity A type of construct validity in evidence based ractice The conscious, explicit, and
which two or more tools that theoretically measure judicious use of the current best evidence in the care
the opposite of the construct are negatively of patients and the delivery of health care services.
correlated. evidence in or ed ractice Acknowledging and
do ains In an ethnographic study, symbolic categories considering the myriad factors beyond such evidence
that include smaller categories. as local indigenous knowledge, cultural and religious
norms, and clinical judgement.
E evidence in or ed ractice guidelines Principles that
e ect si e Measurement of the magnitude of a help the researcher better understand the evidence
treatment effect how large of a difference is observed base of certain practices.
between the groups. e ost acto study A type of nonexperimental research
ele ent The most basic unit about which information that examines the relationships among variables after
is collected. variations have occurred. Also known as a causal-
eligibility criteria Characteristics of a population that comparative study, a comparative study, and (by
meet requirements for inclusion in a study. epidemiologists) a retrospective study
e ic ers ective The native’s or insider’s view of the e clusion criteria Criteria used to exclude individuals
world. from participating in a study.
e irical analytical A general label for quantitative e eri ent A scientific investigation in which
research approaches that test hypotheses. observations are made and data are collected by
e irical actors Those things that can be observed means of the characteristics of control,
through the senses the obtaining of evidence or randomization, and manipulation.
objective data. e eri ental design A research design that has the
e irical literature A synonym for data-based following properties: randomization, control, and
literature. manipulation.
572 Glossary

e eri ental grou The group in an experimental of the world as it is lived by a selected group of
investigation that receives the experimental people.
intervention or treatment. grounded theory ethod An inductive approach in
e ternal criticis A process used to judge the which a systematic set of procedures is used to
authenticity of historical data. develop theory about basic social processes.
e ternal validity The degree to which the findings of a
study can be generalized to other populations or
environments. a thorne e ect See reactivity
e traneous variable A variable that interferes with the her eneutics A theoretical framework in which to
operations of the phenomena being studied. Also understand or interpret human phenomena from the
called mediating variable study of those phenomena.
heterogeneity Dissimilarities of a sample group,
which inhibit the researchers’ ability to interpret
ace validity A type of content validity in which an the findings meaningfully and make
expert’s opinion is used to judge the accuracy of an generalizations.
instrument. hierarchical linear odelling L A type of
actor analysis A strategy for assessing construct regression analysis that allows for analysis of
validity in which a statistical procedure is used to hierarchically structured data simultaneously at all
determine the underlying dimensions or components levels.
of a variable and to assess the degree to which the historical research ethod The systematic approach
individual items on a scale truly cluster around one or for understanding the past through collection,
more dimensions. organization, and critical appraisal of facts.
easibility The capability of the study to be history threat The threat to internal validity that events
successfully carried out. outside of the experimental setting may affect the
ndings The statistical results of a study and the dependent variable.
conclusions, interpretations, recommendations, ho ogeneity A similarity of conditions. Also called
generalizations, and implications for future research internal consistency
and nursing practice. ho ogeneous Having limited variation in attributes or
isher s e act robability test An analysis used to characteristics.
compare frequencies when samples are small and hy othesis A best guess or prediction about what a
expected frequencies are less than six in each cell. researcher expects to find with regard to the
ttingness The degree to which study findings are relationship between two or more variables.
applicable outside the study situation and how hy othesis testing a roach A strategy for assessing
meaningful the results are to individuals not involved construct validity in which the theory or concept
in the research. underlying a measurement instrument’s design is used
or ative evaluation Assessment of a program as it is to develop hypotheses that are tested. Inferences are
being implemented, usually focusing on evaluation of made based on the findings about whether the
the process of a program rather than the outcomes. rationale underlying the instrument’s construction is
re uency distribution A descriptive statistical method adequate to explain the findings.
for summarizing the occurrences of events under
study. I
incidence The number of cases occurring in a particular
period.
generali ability generali e The extent to which data inclusion criteria Criteria that an individual must
can be inferred to be representative of similar satisfy to participate in a study.
phenomena in a population beyond the studied inde endent variable The antecedent or variable that
sample. has the presumed effect on the dependent variable.
grand tour uestion A question in a qualitative The independent variable is manipulated in
study that re ects a broad overview of the issue to be experimental research studies.
studied. inductive Generalizing from specific data.
grounded theory A research approach that is inductive reasoning A logical thought process in which
constructed inductively from a base of observations generalizations are developed from specific
 Glossary 573

observations reasoning moves from the particular to ite to total correlation The relationship between
the general. each item on a scale and the total scale.
in erential statistics Statistical details that combine
mathematical processes and logic to test hypotheses
about a population with the help of sample data. ustice The principle that human participants should be
in or ed consent An ethical principle that requires a treated fairly.
researcher to inform individuals about the potential
benefits and risks of a study before the individuals
can participate voluntarily. a a The level of agreement observed beyond the
instru ental case study Research undertaken to level that would be expected by chance alone.
pursue insight into an issue or to challenge a ey in or ants Individuals who have special
generalization. knowledge, status, or communication skills and who
instru entation threats Changes in the measurement are willing to share their expertise with the
of the variables that may account for changes in the ethnographer.
obtained measurement. no ledge ocused triggers Research ideas that are
internal consistency The extent to which items within generated when staff read research, listen to scientific
a scale re ect or measure the same concept. Also papers at research conferences, or encounter
called homogeneity evidence-based practice guidelines published by
internal criticis The process of judging the reliability federal agencies or specialty organizations.
or consistency of information within a historical no n grou s a roach See contrasted-groups
document. approach
internal validity The degree to which the experimental uder ichardson coe cient The estimate
treatment, not an uncontrolled condition, resulted in of homogeneity used for instruments in which a
the observed effects. dichotomous response pattern is used.
interrater reliability The consistency of observations urtosis The relative peakness or atness of a
between two or more observers often expressed as a distribution.
percentage of agreement between raters or observers
or a coefficient of agreement that takes into account L
the element of chance generally used with the direct level o signi cance al ha level The risk of making a
observation method. type I error, set by the researcher before the study
intersub ectivity A person’s belief that other people begins.
share a common world with him or her an important levels o easure ent Categorization of the precision
tenet in phenomenology. with which an event can be measured (nominal,
interval easure ent A type of measurement in ordinal, interval, and ratio).
which events or objects are ranked on a scale, with Li ert ty e scale A list of statements for which
equal intervals between numbers but with a ranking responses are varying degrees of agreement or
set arbitrarily at zero (e.g., Celsius temperature). opinion for example, strongly agree, agree, no
intervening variable A condition that occurs during an opinion, disagree, or strongly disagree.
experimental or quasiexperimental study that affects li itations The weaknesses of a study.
the dependent variable. Linear Structural elationshi s LIS EL A
intervention An observational method that deals with computer program developed to analyze covariance
whether or not the observer provokes actions from and the testing of complex causal models.
those who are being observed. literature revie An extensive, systematic, and
intervention delity Consistency in data collection. critical review of the most important published
intervie A method of data collection in which a data scholarly literature on a particular topic. In most
collector questions a subject verbally. Such an cases, the literature review is not considered
interview may occur face to face, over the telephone, exhaustive.
or by Skype or other electronic media, and may lived e erience In phenomenological research, the
consist of open-ended or close-ended questions. focus on undergoing events and circumstances
intrinsic case study Research undertaken to gain a (prelingual), as opposed to thinking about these
better understanding of the essential nature of the events and circumstances (conceptualized
case. experience).
574 Glossary

logistic regression logit analysis The analysis of a type of systematic review applied to qualitative
relationships between multiple independent variables research.
and a dependent variable that is binary, ordinal, or ethodological research The controlled investigation
polynomial. and measurement of the means of gathering and
longitudinal study A nonexperimental research analyzing data the development and evaluation of
design in which a researcher collects data from data collection instruments, scales, and techniques.
the same group at different points in time. Also ethodology Discipline-specific principles, rules, and
called prospective study and repeated-measures procedures that guide the process through which
study knowledge is acquired.
idrange theory A focused conceptual structure that
synthesizes the link between practice and research
ani ulation The provision of some experimental into ideas central to the discipline of nursing.
treatment, in varying degrees, to some of the i ed ethods research Research in which the
participants in the study. investigator collects and analyzes data, integrates the
atching A special sampling strategy used to construct findings, and draws inferences using both qualitative
an equivalent comparison sample group by filling it and quantitative approaches or methods in a single
with participants who are similar to each subject in study or a program of inquiry one form of
another sample group in terms of pre-established triangulation.
variables, such as age and gender. odal ercentage A measure of variability percentage
aturation Developmental, biological, or of cases in the mode.
psychological processes that operate within an odality The number of modes, or peaks, in a
individual as a function of time and are external to frequency distribution.
the events of the investigation. ode A measure of central tendency the most frequent
ean A measure of central tendency the score or result.
arithmetic average of all scores. odel A symbolic representation of a set of concepts
easure ent The assignment of numbers to objects or that is created to depict relationships.
events according to rules. ortality The loss of a subject from time 1 data
easure ent e ects Changes in the generalizability of collection to time 2 data collection. Also called
study findings to other populations, as a result of attrition
administration of a pretest. ulti le analysis o variance ANO A A test used
easures o central tendency Descriptive statistical to determine differences in group means when a
techniques that describe the average member of a study has more than one dependent variable.
sample (e.g., mean, median, and mode). ulti le regression The measure of the relationship
easures o variability Descriptive statistical between one interval-level dependent variable and
procedures that describe the level of dispersion in several independent variables. Canonical correlation
sample data. is used when a study has more than one dependent
edian A measure of central tendency in a range of variable.
scores, the middle score (50 of the scores are above ultistage sa ling A sampling method that involves
it and 50 of the scores are below it). successive random sampling of units (clusters) that
E LINE The print or computerized database of progresses from large to small and meets sample
standard medical literature analysis and retrieval eligibility criteria. Also known as cluster sampling
system online it is also available on CD-ROM. ultitrait ulti ethod a roach A type of validation
e ber chec ing In participatory action research, in which more than one method is used to assess the
sharing the findings with the participants to know accuracy of an instrument (e.g., observation and
whether the interpretation of their responses is interview of anxiety).
accurate.
eta analysis A research method in which the results N
of multiple studies in a specific area are examined narrative in uiry A field of hermeneutics that focuses
and the findings are synthesized to make conclusions on the lived experience and perceptions of
regarding the area of focus. experience, in which materials such as in-depth
etasynthesis A technique for drawing inferences or interview transcripts, memoirs, stories, and creative
synthesizing findings from similar or related studies nonfiction are used as sources of data.
 Glossary 575

naturalistic research A general label for qualitative O

studies that involve the researcher going to a natural ob ective An adjective describing data that are not
setting where the phenomenon being studied is taking in uenced by anyone who collects the information.
place. ob ectivity The use of facts without distortion by
naturalistic setting The environment in which people personal feelings or bias.
live in every day, such as homes, schools, and observed test score The actual score obtained in a
communities. measure the true score plus error.
net or sa ling A strategy used for finding odds ratio The probability of an event, which is
samples that are difficult to locate. It entails the calculated by dividing the odds in the treated or
use of social networks and the fact that friends exposed group by the odds in the control group.
tend to have characteristics in common participants one grou retest osttest design A study approach
who meet the eligibility criteria are asked for used by researchers when only one group is available
assistance in getting in touch with others who meet for study participants act as their own controls, and
the same criteria. Also known as sno ball effect no randomization occurs, thus enhancing the internal
sampling validity of the study.
no inal easure ent The level used to classify online database An Internet collection of journal
objects or events into categories without any relative sources (periodicals) of research and conceptual
ranking (e.g., gender, hair colour). articles on a variety of topics (e.g., doctoral
nondirectional hy othesis A hypothesis that indicates dissertations), as well as the publications of
the existence of a relationship between the variables professional organizations and various governmental
but does not specify the anticipated direction of the agencies.
relationship. ontology The science or study of being or existence
none uivalent control grou design A quasi- and its relationship to nonexistence.
experimental design that is similar to the true o en ended ite A question that respondents may
experiment, but participants are not randomly answer in their own words.
assigned to the treatment or the control group. o erational de nition The description of how a
none eri ental research design A research design in concept is measured and what instruments are used to
which an investigator observes a phenomenon capture the essence of the variable.
without manipulating the independent variable or o erationali ation The process of translating concepts
variables. into observable, measurable phenomena.
non ara etric statistics Statistics that are usually o inion leaders Individuals from the local peer group
used when variables are measured at the nominal or who are viewed as respected sources of in uence,
ordinal level because they do not estimate population considered by associates to be technically competent,
parameters and involve less restrictive assumptions and trusted to judge the fit between the evidence-
about the underlying distribution. Also called based practice and the local situation.
distribution-free tests ordinal easure ent A calculation to show rankings
non ara etric tests o signi cance Inferential of events or objects numbers are not equidistant, and
statistics that make no assumptions about the zero is arbitrary (class ranking).
population distribution. orientational ualitative in uiry A qualitative
non robability sa ling A selection technique in approach in which researchers begin with an ideology
which elements are chosen by nonrandom or orientation (e.g., feminism, Marxism, critical
methods. theory) to direct the investigation, including the
nor al curve A statistical curve that is unimodal and research question, methodology, fieldwork, and
symmetrical about the mean. analysis of the findings.
null hy othesis A statement that no relationship exists
between the variables and that any relationship P
observed is a result of chance or uctuations in value The conditional probability of obtaining, from
sampling. Also known as a statistical hypothesis or the study data, the value of the test statistic that is at
least as extreme as that calculated from the data,
null value When an experimental value indicates no given that the null hypothesis is true.
difference between the treatment and control groups aradig From the Greek word meaning pattern : a
known as the value of no effect. set of beliefs and practices, shared by communities of
576 Glossary

researchers, that guide the knowledge development however, this view is tempered by the belief that
process. It is a synonym of orldvie See also science offers an imperfect understanding of the
philosophical beliefs world.
arallel or reliability See alternate-form reliability ost ositivist aradig The basis of most
ara eter A characteristic of a population. quantitative research and, to a smaller extent,
ara etric statistics Inferential statistics that involve qualitative research.
the estimation of at least one parameter, require osttest only control grou design See after-only
measurement at the interval level or higher, and design
involve assumptions about the variables being o er The conditional prior probability that the
studied. These assumptions usually include the fact researcher will make a correct decision to reject the
that the variable is normally distributed. null hypothesis when it is actually false, denoted as
artici atory action research A form of orientation 1 − b.
research that seeks to change society the researcher rediction study A type of nonexperimental research
studies a particular setting to identify problem areas design in which the investigator attempts to make a
to improve practice, identify possible solutions, and forecast or prediction on the basis of particular
take action to implement changes. phenomena.
ath analysis A statistical technique in which the redictive validity The degree of correlation between
researcher hypothesizes how variables are related and the measure of a concept and some future measure of
in what order and then tests the strength of those the same concept.
relationships or paths. revalence The number of people affected by a disease
Pearson correlation coe cient Pearson r A statistic or health problem.
that is calculated to re ect the degree of relationship ri ary sources Scholarly literature that is written by
between two interval level variables. Also called a person or persons who developed the theory or
Pearson product-moment correlation coef cient conducted the research articles and books by the
ercentile A measure of rank the percentage of scores original author or authors. Primary sources include
that a given score exceeds. eyewitness accounts of historical events provided by
heno ena Occurrences, circumstances, or facts that original documents, films, letters, diaries, records,
are perceptible by the senses. artefacts, periodicals, and tapes.
heno enological ethod A process of learning and rint inde es Paper based listings of published
constructing the meaning of human experience material, generally used to find journal sources
through intensive dialogue with persons who are (periodicals) of data-based and conceptual articles on
living the experience. a variety of topics, as well as publications of
heno enology A qualitative research approach that professional organizations and various governmental
aims to describe experience as it is lived through, agencies. Most information is now entered onto
before it is conceptualized. electronics (online) databases.
hiloso hical belie s The system of motivating values, robability The long-run relative frequency of an event
concepts, principles, and the nature of human in repeated trials under similar conditions.
knowledge of an individual, group, or culture see robability sa ling A selection technique in which
also paradigm and orldvie some form of random selection is used when the
hysiological easure ent The use of specialized sample units are chosen.
equipment to determine the physical status of roble ocused triggers Research ideas that are
participants in a study. identified by staff through quality improvement, risk
ilot study A small, simple study conducted as a surveillance, benchmarking data, financial data, or
prelude to a larger-scale study (which is often called recurrent clinical problems.
the parent study ). roble state ent A statement in a research article in
o ulation A well-defined set that has certain specified which the research question is articulated.
properties. rocess consent A request for the respondent’s
ost hoc analysis Comparison of all possible pairs of continued participation in a study.
means after an omnibus ANOVA to determine where roduct testing The testing of medical devices.
the difference lies. ro osition A linkage of concepts that lays a
ost ositivis The view that a reality exists foundation for the development of methods that test
that can be observed, measured, and understood relationships.
 Glossary 577

ros ective study A nonexperimental study that begins equal chance of being assigned to either the
with an exploration of assumed causes and then experimental group or the control group.
moves forward in time to the presumed effect. Also range A measure of variability the difference between
called longitudinal study and repeated-measures the highest and the lowest scores in a set of sample
study data.
sycho etrics The theory and development of ratio easure ent The ranking of the order of events
measurement instruments. or objects that has equal intervals and an absolute
ur ose The aims or objectives the investigator hopes zero (e.g., height, weight).
to achieve with the research. reactivity The distortion created when those who are
ur osive sa le A group consisting of particular being observed change their behaviour because they
people who can illuminate the phenomenon they want know that they are being observed. Also known as
to study. the a thorne effect
ur osive sa ling A sampling strategy in which the reco endations An investigator’s suggestions for the
researcher’s knowledge of the population and its application of a study’s results to practice, theory, and
elements is used to select the participants. future research.
records or available data Information that is
collected from existing materials, such as hospital
ualitative research The systematic, interactive, and records, historical documents, and audio or video
subjective research method used to describe and give recordings.
meaning to human experiences. ualitative research re ereed eer revie ed ournal A scholarly journal
is often conducted in natural settings and uses data that has a panel of external and internal reviewers or
that are words or text, as opposed to numerical data, editors the panel reviews manuscripts submitted for
to describe the experiences being studied. possible publication. The review panel uses the same
uantitative research The process of testing set of scholarly criteria to judge whether the
relationships, differences, and cause-and-effect manuscripts are worthy of publication.
interactions among and between variables. These re e ivity The situation wherein researchers must
processes are tested with hypotheses and research monitor whether their own perspectives are affecting
questions through the use of objective, precise, and their research methods, analyses, or interpretations.
highly controlled measurement techniques to gather relationshi di erence study A study that traces the
information that can be analyzed and summarized relationships or differences between variables that can
statistically. provide a deeper insight into a phenomenon.
uasie eri ent Research in which the researcher reliability The consistency or constancy of a measuring
initiates an experimental treatment, but some instrument the extent to which the instrument yields
characteristic of a true experiment is lacking. the same results on repeated measures.
uasie eri ental design A research approach in reliability coe cient A number between 0 and 1 that
which random assignment is not used, but the expresses the relationship between the error variance,
independent variable is manipulated and certain true variance, and the observed score. A correlation of
mechanisms of control are used. 0 indicates no relationship the closer to 1 the
uestionnaire An instrument designed to gather data coefficient is, the more reliable is the tool. Also
from individuals. called the alpha coef cient
uota sa ling A nonprobability sampling strategy re resentative sa le A sample whose key
that identifies a specific strata of the population and characteristics closely approximate those of the
represents the strata proportionately in the sample. population.
research The systematic, rigorous, logical investigation
that aims to answer questions about nursing
rando error See chance error phenomena.
rando selection A selection process in which research based rotocols Practice standards that are
each element of the population has an equal and formulated from the findings of several studies.
independent chance of being included in the research ethics board E A board established in
sample. agencies to review biomedical and behavioural
rando i ation A sampling selection procedure in research involving human participants within the
which each person or element in a population has an agency or in programs sponsored by the agency to
578 Glossary

assess whether ethical standards are met in relation to scale A self-report measurement tool in which items of
the protection of the rights of human participants. indirect interest are combined to obtain an overall
research hy othesis A statement about the expected score, A set of symbols is used to respond to each
relationship between variables. Also known as a item. A rating or score is assigned to each response.
scienti c hypothesis scatter lot A visual representation of the strength
research uestion A presentation of an idea that forms and magnitude of the relationship between two
the foundation for a study it is developed from the variables.
research problem and results in the research scienti c hy othesis The researcher’s expectation
hypothesis. about the outcome of a study. Also known as the
research utili ation A systematic method of research hypothesis or
implementing sound research-based innovations in scienti c literature A synonym for data-based
clinical practice, evaluating the outcome, and sharing literature.
the knowledge through the process of research scienti c erit The degree of validity of a study or
dissemination. group of studies.
res ect or ersons The idea that people have the right scienti c observation The collecting of data about the
to self-determination and to being treated as environment and participants. The observations
autonomous agents that is, they have the freedom to undertaken are consistent with the specific objectives
participate or not participate in research. of the study the collection of data is systematically
retros ective data Data that have already been planned and recorded all observations are checked
recorded, such as scores on a standard examination. and controlled and the observations are related to
retros ective study A nonexperimental research design scientific concepts and theories.
that begins with the phenomenon of interest (the secondary analysis A form of research in which the
dependent variable) in the present and examines its researcher takes previously collected and analyzed
relationship to another variable (the independent data from one study and reanalyzes the data for a
variable) in the past. Also known as a causal- secondary purpose.
comparative study, a comparative study, and (by secondary source Scholarly material written by a
social scientists) an e post facto study person or persons other than the individual who
rigour The strictness with which a study is conducted developed the theory or conducted the research.
to enhance the quality, believability, or Secondary sources are usually published. Often a
trustworthiness of study findings. secondary source represents a response to or a
ris bene t ratio The extent to which the benefits of summary and critique of a theorist’s or researcher’s
the study are maximized and the risks are minimized work. Examples are documents, films, letters, diaries,
in such a way that the participants are protected from records, artefacts, periodicals, and tapes that provide
harm during the study. a view of the phenomenon from another’s
ris s The potential negative outcomes of participation perspective.
in a research study. selection The generalizability of the results to other
populations.
S selection bias The threat to internal validity that arises
sa le A subset of sampling units, or elements, from a when pretreatment differences exist between the
population. experimental group and the control group.
sa ling A process in which representative units of a selection e ects The threat to external validity that
population are selected for study in a research occurs when the ideal sample population participants
investigation. are either too few or unavailable to the researcher.
sa ling error The tendency for statistics to uctuate se i uartile range se i inter uartile range A
from one sample to another. measure of variability the range of the middle 50
sa ling ra e A list of all units of the population. of the scores.
sa ling interval The standard distance between the si le rando sa ling A probability sampling
elements chosen for the sample. strategy in which the population is defined, a
sa ling unit The element or set of elements used for sampling frame is listed, and a subset from which the
selecting the sample. sample will be chosen is selected members are
saturation The repetition of information until no randomly selected.
further useful data are forthcoming. s e The measure of the asymmetry of a set of scores.
 Glossary 579

sno ball e ect sa ling A strategy used for finding sy etry When the two halves of a distribution are
samples that are difficult to locate. This strategy mirror images of one another (i.e., when folded over,
entails the use of social networks and the fact that they can be superimposed on each other).
friends tend to have characteristics in common syste atic A term used when data collection is carried
participants who meet the eligibility criteria are asked out in the same manner with all participants and by
for assistance in getting in touch with others who all persons collecting the data.
meet the same criteria. Also known as net or syste atic error An error attributable to the lasting
sampling characteristics of the subject that do not tend to
social desirability The tendency of a subject to uctuate from one time to another. Also called
respond in a manner that he or she believes will constant error
please the researcher rather than in an honest syste atic revie A summary of research evidence
manner. from several studies.
Solo on our grou design An experimental design syste atic sa ling A probability sampling strategy
with four randomly assigned groups: the pretest- that involves the selection of participants randomly
posttest intervention group, the pretest-posttest drawn from a population list at fixed intervals.
control group, a treatment or intervention group with
only posttest measurement, and a control group with T
only posttest measurement. statistic The test of whether two groups’ means are
s lit hal reliability An index of the comparison more different than would be expected by chance.
between the scores on one half of a test with those on The groups may be related or independent.
the other half to determine the consistency in target o ulation A population or group of individuals
response to items that re ect specific content. who meet the sampling criteria and about whom the
stability An instrument’s ability to produce the same researcher hopes to make generalizations.
results with repeated testing. test retest reliability The stability of the scores of an
standard deviation A measure of variability measure instrument when it is administered twice to the same
of average deviation of scores from the mean. In participants under the same conditions within a
equations, abbreviated SD prescribed time interval. The scores from the different
standard error o the ean The standard deviation of times are paired and then compared to determine the
a theoretical distribution of sample means. It indicates stability of the measure.
the average error in the estimation of the population testability The ability of the variables in a proposed
mean. study to be observed, measured, and analyzed by
statistic A characteristic of a sample, described in quantitative methods.
mathematical terms (e.g., percentage). testable Measurable by quantitative methods.
statistical hy othesis A statement that no relationship testing e ect The effect on the scores of a posttest as
exists between the independent and dependent the result of having taken a pretest.
variables. Also known as a null hypothesis te t Data in a contextual form that is, narrative or
strati ed rando sa ling A probability sampling words that were written from recorded interviews and
strategy in which the population is divided into strata then transcribed.
or subgroups members of each strata are the atic analysis The process of recognizing and
homogeneous with regard to certain characteristics. recovering the emergent themes in data.
An appropriate number of elements from each the es Clusters of data with structured meaning that
subgroup are randomly selected based on their occur frequently.
proportion in the population. theoretical ra e or A structure for concepts,
su ative evaluation Assessment of the outcomes theories or both used to construct a map for the study
of a program, conducted after the program’s based on a philosophical or theorized belief or
completion. understanding or why the phenomenon under study
survey study A descriptive, exploratory, or comparative exists.
study in which researchers collect detailed theoretical sa ling In the grounded theory
descriptions of existing variables and use the data to method, the sampling method used to select
justify and assess current conditions and practices or experiences that will help the researcher test ideas
to make more plans for improving health care and gather complete information about developing
practices. concepts.
580 Glossary

theory A set of interrelated concepts, definitions, and ty e I error The researchers’ incorrect decision to
propositions that present a systematic view of reject the null hypothesis.
phenomena for the purpose of explaining and making ty e II error As a result of the sample data, the failure
predictions about those phenomena. to reject the null hypothesis when it is actually false.
ti e series design A quasiexperimental design used to Also known as beta (β).
determine trends before and after an experimental
treatment. Measurements are taken several times
before the introduction of the experimental treatment, validation sa le The sample that provides the initial
the treatment is introduced, and measurements are data for determining the reliability and validity of a
taken again at specified times afterward. measurement tool.
trans erability The extent to which findings from one validity The determination of whether a measurement
qualitative research study have meaning to other instrument actually measures what it is purported to
studies in similar situations. measure.
translation science The investigation of methods, values Personal beliefs of the researcher.
interventions, and variables that in uence the variable A defined concept a property that takes on
adoption of evidence-based practices. different values and is studied by quantitative
triangulation The expansion of research methods in a researchers.
single study or multiple studies to enhance diversity,
enrich understanding, and accomplish specific goals.
true e eri ent A study design in which participants eb bro ser A software program used to connect to
are randomly assigned to an experimental group or a or search the World Wide Web (e.g., Internet
control group, pretest measurements are performed, Explorer).
an intervention or treatment occurs in the orldvie The way people in society think about the
experimental group, and posttest measurements are world a synonym for paradigm See also
performed. Also known as the pretest-posttest control philosophical beliefs
group design or classic e periment
trust orthiness An accurate portrayal of the
experience of the study’s participants a measure of score A rating used to compare measurements in
rigour in qualitative research that includes the standard units an examination of the relative distance
concepts of credibility, audibility, and fittingness. of the scores from the mean.
Index
A Attentively embracing story, 38–39, 38f
A priori, 228, 364–365 Attrition, 208t, 209, 222
Aboriginal population, 121, 133–134 Audio recording interviews, 332
Abstraction, 335–336, 336t Auditability, 324t, 325–326, 417
Abstracts, 57–58, 101 Authenticity of records, 300
Accessible populations, 261, 279, 279f Autonomous agents, 117t–119t
Accuracy, 201 Available data, 299–300
Actigraphy, 318 Average, arithmetical. See Mean (M)
Advanced statistics, 376–377
Aesthetic knowledge, 25f
After-only design, 221f, 222–223 Background, 62
After-only nonequivalent control group design, 225f, Backward solution, 372
226–227 Barnason Efficacy Expectation Scale (BEES), 317–318
Aim of inquiry, 26–29, 27t–28t Basic research, 199
Aims, research, 74–75 Behavioural/materialistic perspective, 179
Alpha coefficient, 308–309 Beneficence, 114, 114b
Alpha level, 364–366 Benefits, 136
Alternate-form reliability, 309–310, 315 Beta (β), 363–364
Analysis of covariance (ANCOVA), 369–370 Bias, 172–173, 199–200, 263t
Analysis of variance (ANOVA), 368–369 Bibliographic databases, online, 101
Analysis understanding, 51, 51b Biographical history, 185–186
Analytic coding, 335 Biological measurements, 291–292
Anecdotes, 294 Blinding, 205–206
Animal rights, 135–136 Blobbograms, 375–376, 376f
Anonymity, 117t–119t, 123–128 Boolean operator, 105–106
Antecedent variables, 220 Bracketing, 160, 172
Appraisal, 62 Brainstorming, 67, 69b
Appraising evidence British Medical ournal, 102b
data-collection methods, 302b
descriptive/inferential statistics, 378b–380b C
experimental/quasiexperimental designs, 230b–231b Canadian Association for Nursing Research (CANR), 14b
legal/ethical issues, 137b Canadian Association of Schools of Nursing (CASN), 9
literature reviews, 108b–109b Canadian Consortium for Nursing Research and Innovation
nonexperimental designs, 249b (CNRI), 14b
qualitative data analysis, 344b Canadian Council on Animal Care (CCAC), 135–136
qualitative research, 191b Canadian Health Services Research Foundation (CHSRF), 6–7,
quantitative research, 212b–213b 13–14, 14b, 17
reliability/validity, 326b–328b Canadian Institutes of Health Research (CIHR), 14b, 18,
Results / Discussion sections, 392b 135–136
sampling strategies, 280b–283b Canadian ournal of Nursing Research, 13
Assent, 131–132 Canadian Nurses Association (CNA), 6, 113, 121, 130b
Association, tests of, 359t Canadian Nursing Foundation (CNF), 14b
Associative relationship statement, 81 Canadian Registered Nurse Examination (CRNE), 9
Assumptions, 51 Canonical correlation, 372
Caregiver Reciprocity Scale, 318–320, 319t
Case study method, 182–185, 343t
Page numbers followed by f indicate figures t, tables Categorical variable, 349
b, boxes. Causal modelling, 318–319
581
582 INDEX

Causal-comparative studies, 241 Constructivism, 26–30, 27t–28t

Causality, 217, 243–244 Constructivist paradigm, 30
Cause-and-effect relationships, 218, 221–222, 224, Consumer of research
228–229 intelligence as, 8–11
Chance error, 322–324 literature reviews by, 90
Change champions, 501–502 perspective of, 81–83, 95
Chi-square ( 2), 370 Content analysis, 296
Citation management software, 104–106 Content validity, 315–316
Clinical guidelines, 62–63 Context, 26–31, 27t–28t, 179–180
Clinical questions, 81–83, 83b–84b, 83t Context dependent, 149, 149f
Close-ended items, 296, 297f Contrasted-groups approach, 320–321, 321t
Cluster sampling, 270–271 Control, 199–200, 203–206, 219
Cochrane Library, 101, 102b Control group, 205–206
Code of ethics, 113 Controlled vocabulary, 105
Code of Ethics for Registered Nurses, 113, 130b Convenience sampling, 263–264
Codes, 335, 336t Convergent validity, 318–320
Coding, 335, 337–338, 337t–338t Core group, 502
Coercion, 115t–119t Correlation, 370
Cognitive application, 482 Correlation coefficients, 370–373, 371f
Cognitive perspective, 179 Correlational studies, 206, 238–239
Cohen’s kappa, 314–315 Cost-effectiveness, 15–16
Cohort, 240 Covert data collection, 117t–119t
Committee for the Study of Nursing Education, 11–13 Creating ease, 38–39, 38f
Communication, 499–502 Credibility, 324t, 325, 400
Community-based participatory research (CBPR), 187. See Criterion sampling, 275
also Participatory action research (PAR) Criterion-related validity, 316–317
Comparative studies, 241 Critical appraisal, 52, 53b, 68t
Comparative surveys, 236 Critical ethnography, 180
Comparison group. See Control group Critical reading, 49–52, 51b
Composite reliability, 318–319 Critical theory, 26–31, 27t–28t, 169f, 175
Comprehensive understanding, 51b Critical thinking, 49–52
Computer database search, 99 Critical thinking decision path
Computer-assisted qualitative data analysis software assessing sampling strategy/generalizability relationship,
(CA DAS), 332–333 274b
Computerized data collection, 300 assessing study results, 385b
Concealed observation with intervention, 293f, 294 assessing types of validity/reliability, 323b
Concealment, 293, 293f choosing experimental/quasiexperimental design, 218b
Concepts, 42, 92 choosing nonexperimental design, 236b
Conceptual definition, 41, 92, 288–289 choosing theoretical path, 43b
Conceptual framework, 36f, 43–44, 92–93 descriptive statistics choice, 349b
Conceptualization, 200–201 determining type of hypothesis, 82b
Concern for welfare, 119–120 evaluating risk-benefit ratio, 132b
Conclusion drawing, 341–342 for inferential statistics, difference questions, 360b
Conclusions, in systematic reviews, 62 for inferential statistics, relationship questions, 361b
Concurrent validity, 316 locating evidence, 100b
Conduct of research, 114, 114b, 480–481, 481f noting threats to validity, 210b
Confidence intervals, 373–374, 373t, 389 selecting data-collection method, 289b
Confidentiality, 117t–119t, 123–128 selecting qualitative research method, 170b
Confirming/disconfirming cases, 275 selecting research process, 32b
Connecting with self-in-relation, 38–39, 38f Critical values method, 365
Consent, 123, 131–132 Critiquing criteria
Consistency, 56, 290 data-collection methods, 303b
Constancy, 204–205 defined, 52
Constant comparative method, 178, 341 descriptive/inferential statistics, 380b
Constant error, 324 experimental/quasiexperimental designs, 231b
Construct, 301 frameworks, 44–45, 45b
Construct validity, 317–324 hypotheses, 85–86, 87b
 INDEX 583

Critiquing criteria Continued Deductive reasoning, 39–40

legal/ethical issues, 137b Degree of freedom, 367
literature reviews, 109b Delimitations, 259–261
nonexperimental designs, 250b Delivery of health care, 15–16
qualitative data analysis, 344b Dependent variables, 71–72, 203–204
qualitative research, 191b Descriptive coding, 335
quantitative research, 213b Descriptive correlational studies, 239, 239t
reliability/validity, 328b Descriptive phenomenology, 169
research question, 84–85, 87b Descriptive statistics, 347–359, 349b, 378b–380b
Results / Discussion sections, 392b–393b Descriptive/exploratory surveys, 236
sampling strategies, 283b Design
Critiquing process, 44–45, 45b critique of, 212b–213b
Critiquing strategies, 52–54 emergent nature of, 159
Cronbach’s alpha, 312, 312b, 312t interrelationships of, 199, 199f
Cross-sectional studies, 239–240, 239t literature reviews for, 93–94
Culture, 179 purpose of, 199–200
Cumulative Index to Nursing and Allied Health Literature in qualitative research, 151
(CINAHL), 69, 99–107, 102b, 105f, 106b Determinant of health framework, 36–37
Cyclical uctuations, 272 Developmental studies, 239–243
Deviant cases, in purposive sampling, 275
Dichotomous response format, 311
Data, 8 Dichotomous variable, 349
Data analysis Differences
in case studies, 184–185 between groups, 360b
in ethnographic method, 181–182 between means, 359t
in grounded theory method, 177–178 tests of, 361–362, 368–370
in participatory action research, 189 Differential loss of participants, 208t, 209
in phenomenological method, 173–174 Dignity, 121, 130b
qualitative, 153–154, 333–342, 334f, 343t, 344b Directional hypothesis, 78–80, 79t
quantitative, 348–378 Discomfort and harm, 117t–119t, 121
in research articles, 60–61 Discussion section
Data-collection methods critique of, 392b–393b
critique of, 302b–303b features of, 61, 388–392, 389t
instrument development, 300–301 Dissemination, 480, 481f
online/computerized, 300 Distribution-free tests, 367
in research articles, 60 Divergent (discriminant) validity, 318–320
selection of, 288–291, 289b Doctoral degree, 10–11, 13
types of, 291–300 Domains, 181–182
Data collection/gathering Double blinding, 205–206
in case studies, 184
constancy in, 204–205 E
in ethnographic method, 181 Education, 7–8
in grounded theory method, 177 Education Resource Information Center (ERIC), 102b
in historical research, 186 Edwards, Dr. Nancy, 16–17
in participatory action research, 189 Effect size, 277
in phenomenological method, 173 Effectiveness, 229
for qualitative research, 152–153 Element, 261
Data display, 338–340, 339f Eligibility criteria, 259–261
Data management, 332–333 Embryonic research, 136–138
Data reduction, 334–338, 334f Emergent nature of research design, 159
Data saturation, 152–153, 173, 275, 430–431 Emic perspective, 180, 183–184
Data triangulation, 161 Empirical analytical approach. See uantitative research
Databases, 69, 99, 102b Empirical knowledge (empirics), 24–26, 25f
Debriefing, 293 Epidemiological studies, 248–250
Deception, 117t–119t Epistemology, 26–31, 27t–28t
Declarative statement, 76 Equivalence, 308, 314–315
Deductive paradigm, 161, 162t Error variance, 322
584 INDEX

Ethical issues Experiment, 218

critique of, 137b Experimental designs
current/future dilemmas, 120–121 advantages/disadvantages, 223–224
evolution of, 121 critique of, 230b–231b
history of, 113–121, 115t–116t for field/laboratory, 223
in qualitative research, 158–160, 158t overview, 217, 217b–218b
Ethical knowledge, 25f properties of, 218–220
Ethical principles, 11, 114b types of, 220–223, 221f
Ethical Research uidelines for Registered Nurses, 121 Experimental group, 205–206
Ethics, 120 Explanatory models, 180
Ethnogeriatrics, 180 Exploitation, 159
Ethnographic research method External criticism, 186
abstraction in, 336t External validity, 207b, 208t, 210–212, 210b
data analysis in, 343t Extraneous variables, 203–204, 220
features of, 178–182
Ethnography, 179–180
Etic perspective, 180, 183–184 test, 368–369
Evaluation, of evidence-informed practice, 502–504, 503b, Face validity, 316
504t Face-to-face interviews, 298
Evaluation research, 229–232 Fact, in historical research, 186, 187b
Evidence Factor analysis, 321–324, 322t, 377
appraisal of. See Appraising evidence Fair treatment, 117t–119t
grading of, 491–494, 492t–493t Feasibility, 70, 201–203, 202t
levels of, 55–56, 55f, 98f Feminist ethnography, 180
retrieval of, 489–491 Feminist theory, 175
searching for, 96–99, 97t, 100b Field experiments, 223
use of, 482–483 Findings
Evidence-based practice, 6, 481–482 in case studies, 184–185
Evidence-informed practice in ethnographic method, 182
building culture of, 504–507, 505f in grounded theory method, 178, 179f
changes in, 497 in participatory action research, 189–190
communication in, 499–502 in phenomenological method, 174–175
critiquing practice guidelines, 494 presentation of, 384–392
critiquing studies, 494–495 in qualitative research, 154
defined, 481–482 Fisher’s exact probability test, 370
development of, 497–498, 498b Fittingness, 324t, 325–326, 417
evaluation of, 502–504, 503b, 504t Fixed-response items, 296
future directions, 508 Flexibility, 206
guidelines for, 483–485, 494 Focus group(s), 299, 498, 498b
implementation of, 498–504, 499f–500f Format, 57–61
literature for, 489–494 Formative evaluation, 229–230
models of, 483–486 Forward solution, 372
overview, 480–486 Frameworks, 36–45, 45b
in qualitative research, 155–157, 156t–157t Fraudulent research, 134
recommendations for, 495–497, 497b Frequency distribution, 351–352, 351t, 352f
research and, 54–57 Frequency polygons, 352, 353f
significance of, 5–7 Funding, 13–14, 14b, 18
steps of, 57f Future directions, 15–20
synthesis of research, 495, 496t
team formation, 487–489, 488f
topic selection, 486, 486b, 487t Generalizability, 16, 210, 274b, 390
translation science and, 507–508 Genetics, 18
Ex post facto studies, 241–243, 241t Global research community, 17
Excerpta Medica, 102b Goldmark report, 11–13
Exclusion criteria, 151, 259–261 Goodness, 326
Expedited review, 129 Graduate degree, 10
Experiential knowledge, 25f Grand theory, 43f, 44, 46
 INDEX 585

Grand tour questions, 153 Informed consent

Grounded theory research method elements of, 122b
abstraction in, 336t examples of, 124f–127f
critique of, 326 human rights and, 122–128
critiquing guidelines, 403b observational studies and, 293
data analysis in, 343t Instrument development, 300–301
features of, 175–178 Instrumental case study, 182–183
premises of, 176b Instrumentation threats, 208–209, 208t
sample critique of, 430–432 Integrative reviews, 61–62
sample report as, 419–427 Intellectual history, 185
theoretical sampling and, 275–276 Intentional dialogue, 38–39, 38f
Interdisciplinary triangulation, 161
Internal consistency, 308, 310–312
Harm studies, 374–375, 374t Internal criticism, 186
Hawthorne effect, 211, 293–294 Internal validity, 207–210, 207b, 208t, 210b
Health care disparities, 18 International Council of Nurses, 19
Health care shifts, 15–16 International research, 17, 19
Heideggerian phenomenology, 171 Internet data collection, 300
Hermeneutic philosophy, 171 Internet resources, 308b
Hermeneutics, 37, 171 Interrater reliability, 290–291, 314–315
Heterogeneity, 260–261 Intersubjectivity, 171–172
Hierarchical linear modelling (HLM), 243 Interval measurement, 350
Hierarchy of evidence, 155, 266. See also Levels of evidence Intervening variables, 220
Histograms, 352, 352f–353f Intervention, 293, 293f
Historical research method, 185–187, 187b Intervention fidelity, 290
History, of nursing research, 11–15, 12b Interviews, 295–299, 332
History threat, 207, 208t Intrinsic case study, 182–183
Homogeneity, 308, 309b, 310–312 Introduction, in research articles, 49t, 58
Homogeneous sampling, 204, 260–261 Investigator triangulation, 161
Human genome, 120, 136–138 Iowa Model of Evidence-Based Practice to Promote uality
Human rights, 117t–119t, 121–133 Care, 483–485, 484f
Hypothesis Item reliability, 318–319
characteristics of, 76–77 Item-to-item correlation, 310–311, 310t
critique of, 85–86, 87b
development of, 69b, 75–80, 76f
directional vs. nondirectional, 78–80 Jewish Chronic Disease Study, 115t–116t
importance of, 66–67 Johanson, Sue, 20
literature reviews for, 93 Justice, 114, 114b
of research articles, 59
research question, design and, 80–81
statistical vs. research, 77–78 Kearney’s categories, 155–157, 156t–157t
theory base for, 77 Kendall’s tau, 359t, 361f, 371–372
variables and, 42 Key informants, 177
wording of, 77, 79t Knowledge, 24–26, 24f, 39–40
Hypothesis-testing approach, 317–318, 359t, 361–362 Knowledge creep, 482
Knowledge-focused triggers, 486
I Known-groups approach, 320–321
Implications, 61 Kuder-Richardson (KR-20) coefficient, 311–312
Incidence studies, 248–250 Kurtosis, 351
Inclusion criteria, 151, 259–261
Independent variables L
features of, 71–72 Laboratory experiments, 223
manipulation of, 205–206, 219–220 Ladder of abstraction, 40–41, 41f
in quantitative research, 203–204 Legal issues, 113–121, 137b
Inductive paradigm, 161, 162t Level of significance, 364–366
Inductive reasoning, 39–40 Levels of evidence, 55–56, 55f. See also Hierarchy of evidence
Inferential statistics, 348, 359–377, 360b–361b, 378b–380b Levels of measurement, 348–359, 348t
586 INDEX

Likert-type scales, 296, 297f, 313f Mortality, 208t, 209

Limitations, 386 Multimethod research, 160–164, 162t–163t, 226
Linear Structural Relationships (LISREL), 318–319, 319t Multiple regression, 372
Literature reviews Multistage sampling, 270–271
characteristics of, 110b Multitrait-multimethod approach, 320
critique of, 108b–109b Multivariate analysis of variance (MANOVA), 369
data-collection methods and, 288–289
format, 107–108 N
overview, 91–92, 91f Narrative inquiry, 175
purposes of, 92–96, 93b National Commission for the Protection of Human Subjects of
in qualitative research, 150–151 Biomedical and Behavioral Research, 114
research ideas and, 67–70, 69b National Research Act, 114
searching, 96–99, 97t, 98f, 103t Naturalistic research. See ualitative research
theoretical framework and, 58–59 Naturalistic setting, 148, 158–159
types of resources, 99–107 Network sampling, 273
Literature search, 96–107, 97t, 98f, 103t Nightingale, Florence, 11, 12b
Lived experience, 172 Nineteenth century, nursing research in, 11
Logistic regression (logit analysis), 367–368, 376–377 Nominal measurement, 348–349
Longitudinal studies, 240–241 Nondirectional hypothesis, 78–80, 79t
Nonequivalent control group design, 225–226, 225f
Nonexperimental research designs
Manipulation of independent variables, 205–206, 219–220 additional quantitative studies, 243–244
Master’s degree, 10, 13 choice of, 236b
Matching, 272–273 critique of, 249b–250b
Maturation effect, 207, 208t epidemiological studies, 248–250
Mean (M), 348t, 353–354 hypotheses in, 79t, 81, 86
Measurement, 288, 348 overview, 235, 235b, 235f
Measurement effects, 212 prediction/causality in, 243–244
Measurement instruments survey studies, 236–238
development of, 246t Nonparametric statistics, 367
equivalence of, 308, 314–315 Nonparametric tests, 359–360, 359t, 370
homogeneity of, 308, 309b, 310–312 Nonparametric tests of significance, 367
reliability of, 307–315, 309b, 323b Nonprobability sampling, 262–266, 263t, 273–276
rigour determined by, 307 Normal curve, 354
stability of, 308–310 Normal distribution, 354–355, 354f
validity of, 307, 315–324 Null hypothesis, 79, 80t, 361–362
Measures of central tendency, 347, 348t, 352–354, 354t Nuremberg Code, 113, 114b
Measures of variability, 355–359 Nursing Ethics: or ospital and Private se (Robb), 121
Median, 348t, 353 Nursing knowledge, 25f
Medical Outcomes Study Short-Form 36, 317 Nursing research. See also Research
MEDLINE, 69, 101, 102b Nursing Research, 13
Member checking, 342 Nursing research
Mental health illnesses, 18 dissemination of, 20
Meta-analysis, 164, 245–248, 375–376, 376f history of, 11–15, 12b
Metasynthesis, 164 methods, 190–191, 190t
Method, 62, 93–94 promoting depth in, 16–17
Methodological research, 244–245, 246t Nursing Research Fund, 13–14, 16
Methodological triangulation, 161 Nursing science, 157–158, 158f
Methodology, 26–29, 27t–28t
Microrange theory, 43f, 44, 46 O
Midrange theory, 43f, 44, 46 Objective data, 288
Milestones, historical, 12b Objectives, of research study, 74–75
Misconduct, 134–135 Objectivity, 200–201, 387b
Mixed-methods studies. See Multimethod research Observational data-collection methods, 292–295, 293f
Modality, 352 Observed test score, 322
Mode, 348t, 352 Odds ratio, 374–375, 374t
Models, 36f Office of Nursing Policy, 14b
 INDEX 587

Older adults’ human rights, 132–133 PICOT format, 81, 83b, 95, 96b, 489t
One-group (pretest-post-test) design, 225f, 227 Pilot study, 201–203, 276
Online data collection, 300 Point-biserial correlation, 361b, 371–372
Online databases, searching, 99, 101, 102b Populations
Ontario Training Centre in Health Services and Policy features of, 259–261
Research, 14 of interest, 18–19
Ontology, 26–30, 27t–28t for research questions, 73t
Open coding, 178 statistical significance in, 359
Open-ended items, 296, 297f stratified, 264–266, 268–270
Operational definition, 41, 92, 288–289 Possibility, in historical research, 186, 187b
Operationalization, 288 Post hoc analysis, 369–370
Opinion leaders, 501, 502b Post-positivism, 26–30, 27t–28t
Ordinal measurement, 349–350 Post-positivist paradigm, 30
Orientational qualitative inquiry, 175 Post test-only control group design, 221f, 222–223
Ottawa Model of Research se (OMR ), 485–486, 485f Post tests, 220–221, 221f
Outcomes Power, 363–364
evaluating, 229–232 Power analysis, 277–278
literature reviews for, 94–95 Practical experience, 68t
statistical decision making, 364f Practical significance, 366
Practice, link to theory and education, 7–8
P Prediction, 243–244
P value, 365 Prediction studies, 243–244
Paired comparisons, 369–370 Predictive correlational studies, 239, 239t
Paradigms, 26–29, 27t–28t, 148 Predictive validity, 316–317
Parallel-form reliability, 309–310, 315 Preliminary understanding, 51b
Parameters, 359 Pretests, 220–221, 221f
Parametric statistics, 367 Prevalence studies, 248–250
Parametric tests, 359–360, 359t, 368–369 Primary sources, 93, 94t, 186, 187b
Parametric tests of significance, 366 Print databases, 101–104
Participants, 31, 32b Print indexes, 99
Participatory action research (PAR), 159–160, 187–190, 342 Prisoners, 133
Patterns of knowing, 5 Privacy, 117t–119t, 119–121, 130b
Pearson correlation coefficient (Pearson r), 370–371 Probability, 186, 187b, 362–363
Pediatric research, 131–133 Probability sampling, 262, 263t, 266–272
Percentile, 356 Problem statements, 66, 200–201
Perfect negative correlation, 371, 371f Problem-focused triggers, 486
Perfect positive correlation, 371, 371f Procedures, in research articles, 60
Personal bias, 172–173 Process consent, 123
Personal digital assistants (PDAs), 300 Process evaluation, 229–232
Personal judgement, 437 Product testing, 135
Personal knowledge, 25f Proportional stratified random sampling, 268–270, 269f
Perspective. See Researcher’s perspective Propositions, 176
Phenomena, 5–6, 401–404 Pro uest, 102b
Phenomenological reduction, 160, 172 Prospective studies, 239t, 240–241
Phenomenological research method Psychometrics, 244–245
abstraction in, 336t Psychosocial variables, 288, 312
data analysis in, 343t PsycINFO, 69, 102b
overview, 169–175 PubMed, 69
Phenomenon identification Purpose, of research study, 58, 74–75
in case studies, 183 Purpose statements, 75, 75b
in ethnographic method, 180 Purposive sampling, 151, 273–275
in grounded theory method, 176–177
in participatory action research, 188
in phenomenological method, 172 ualitative outcome analysis ( OA), 404
Philosophical beliefs, 26 ualitative research
Photovoice, 189 application of, 401–405
Physiological data-collection methods, 291–292 appropriate use of, 150
588 INDEX

ualitative research Continued

beliefs of, 149 Random error, 322–324
components of, 150–154, 150b Random numbers, 266, 267f
continuum of methods, 167, 169f. See also speci c methods Random selection, 266
critique of, 191b Randomization (random assignment), 206, 218–219
critiquing guidelines, 402b Randomized clinical trial (RCT), 218
data analysis. See Data analysis, qualitative Range, 356, 357t–358t
design for, 59–60 Ranking, 348t, 349–350
ethical issues in, 158–160, 158t Ratio measurement, 350
evidence in, 155–157, 156t–157t Reactivity (reactive effects), 210–212, 210f, 293–294
features of, 148–149 Reciprocity. See Caregiver Reciprocity Scale
format for, 49t, 73t Recommendations, 61, 391b, 392
foundation of, 164–165 Records, for data collection, 299–300
informed consent for, 126f–127f Recruitment, in qualitative research, 152
levels of evidence in, 56 Refereed (peer-reviewed) journals, 99–101
metasynthesis in, 164 References, 61
nonprobability sampling in, 273–276 Re exivity, 160
nursing science and, 157–158, 158f Rejection of null hypothesis, 362
process of, 33–34, 33f Relationship
research questions in, 71 tests of, 73–74, 370–373
rigour in, 307 between themes, 341
sample critique of, 416–418 Relationship questions, 361b
sample report as, 402, 407–414 Relationship statements, 76–77
sample size in, 278 Relationship/difference studies, 235b, 238–243
selecting, 32b, 170f Reliability
stylistic considerations, 400–401 critique of, 326b–328b
triangulation and, 160–164, 162t–163t Internet resources and, 308b
ualitative Research Evaluation Methods (Patton), 171 overview, 60
uality, 56 of research instruments, 307–315, 309b, 323b
uality of life, 289–290 Reliability coefficient, 308–309
uantitative control, 206 Repeated measures studies, 239t, 240–241
uantitative research Replication studies, 16
critique guidelines for, 435b–436b Representative samples, 262
data analysis. See Data analysis, quantitative Research. See also Nursing research
design for, 59 Aboriginal people and, 133–134
format for, 50t, 73t challenges, 15
informed consent for, 124f–125f consumers of, 8–11
literature reviews and, 92–95, 95f content sections of, 435b–436b
process of, 34–36, 35f critique of, 11
research questions in, 71 dissemination of, 10, 19–20
rigour in, 307 ethical/legal issues, 11, 113–121
sample critique of, 453–455, 470–472 frameworks for, 40–44
sample report as, 438–450, 456–467 funding of, 13–14, 14b, 16
sample size in, 276–278 future of, 15–20
selecting, 32b global, 17
stylistic considerations, 437 linking education/theory/practice, 7–8
uantity, 56 nurse’s role in, 9–11
uasiexperimental designs philosophies of, 26–31
advantages/disadvantages, 228–229 significance of, 5–7
critique of, 230b–231b Research articles, 57–61
features of, 224–225 Research design, 59–60, 80–81. See also Design
overview, 217, 217b–218b Research ethics boards (REBs), 128–131, 131b, 279, 279f
types of, 225–228, 225f Research hypothesis, 77–78, 79t, 361–362
ueer theory, 175 Research paradigms, 26–31, 27t–28t
uestionnaires, 295–299 Research priorities, 17–20, 18b
uota sampling, 264–266, 265t Research questions
uotations, 154, 340, 340t about health issues, 38t
 INDEX 589

Research questions Continued Sampling Continued

in case studies, 183 matching, 272–273
clinical questions as, 81–83 procedures, 60, 278–280, 279f
critique of, 84–85, 87b purpose of, 259, 261–262
developing, 67–70, 68t in qualitative research, 273–276
in ethnographic method, 180–181 strategies, 262–276, 263t, 274b
fully developed, 71–74 Sampling distribution of the means, 354
in grounded theory method, 177 Sampling distributions, 363
hypothesis, design and, 80–81 Sampling error, 362–363
importance of, 66 Sampling frame, 266
interrelationships of, 75–76, 76f Sampling interval, 271–272
literature reviews for, 93 Sampling plan, 279, 279f
in participatory action research, 188 Sampling unit, 270
in phenomenological method, 172 San Antonio Contraceptive Study, 115t–116t
in research process, 59, 66–67 Scales, 295–296, 297f
study purpose and, 74–75 Scatter plots, 347
Research themes. See Themes Scientific hypothesis, 77–78, 361–362
Research utilization, 480–482, 481f, 485–486, 485f Scientific merit, 437
Researcher, as instrument, 160 Scientific observation, 292, 294–295
Researcher-participant interaction, 159–160 Search, literature, 96–107, 97t, 98f, 103t
Researcher’s perspective Search history, 104–105, 105f
in case studies, 183–184 Secondary analysis, 248
in ethnographic method, 181 Secondary sources, 93, 94t
in grounded theory method, 177 Selection, sample, 173
of literature reviews, 92–95 Selection bias, 208t, 209–210
in participatory action research, 188 Selection effects, 211
in phenomenological method, 172–173 Self-determination, 117t–119t, 121
Respect for persons, 114, 114b Semiquartile/semi-interquartile range, 356
Results section Setting section, in qualitative research, 152
assessing, 384–388 Significance
communicating, 61 examining, 70
critique of, 392b–393b level of, 364–366
features of, 384–388, 385t, 387b, 387t–388t tests of, 366–374
Retrospective data, 242 Simple random sampling, 266–268, 267f
Retrospective studies, 239t, 241–243, 241t Skew, 354–355, 355f
Rigour Snowball effect sampling, 177, 273
credibility/auditability/fittingness, 324–326, 324t Social desirability bias, 296–298
critique of, 326b–328b Social history, 185–186
Internet resources for, 308b Social processes, 175–176
qualitative/quantitative research and, 307 Sociopolitical knowledge, 25f
reliability, 307–315 Solomon four-group design, 221f, 222
validity, 315–324 Spearman-Brown formula, 311
Risk-benefit ratio, 132f, 136, 138 Spearman’s rho, 359t, 371–372
Risks, 136 Split-half reliability, 311
Stability, 308–310
S Stakeholders, 487–489, 488f
Sample selection Standard deviation (SD), 354f, 356–359
in case studies, 184 Standard error of the mean, 363
in ethnographic method, 181 Standard error (SE), 318–319, 319t
in grounded theory method, 177 Statistical hypothesis, 79, 80t, 361–362
in participatory action research, 188–189 Statistical significance, 366–374
in phenomenological method, 173 Statistics
Sample size, 276–278, 277t advanced, 376–377
Samples, for qualitative research, 151–152, 261–262 critique of, 377–378
Sampling for data analysis, 347–348
concepts, 259–262 function of, 359, 377
critique of, 280b–283b Statistics Canada, 103t
590 INDEX

Step activity monitor (SAM), 314 Topic coding, 335

Stepwise solution, 372 To ard : Visions for Nursing (Villeneuve MacDonald),
Stratified random sampling, 268–270, 269f 15
Structural equation modelling, 377 Transferability, 391
Structure Translation science, 507–508
of case studies, 183–184 Treatment effect, 364
of ethnographic method, 180–181 Trends, 15–16, 227
of grounded theory method, 177 Triangulation, 160–164, 162t–163t, 325
of participatory action research, 188–189 Tri-Council Policy Statement-2, 115t–116t, 120–121
of phenomenological method, 172–173 True experiment, 218–220. See also Experimental
Structured observations, 294 designs
Study design. See Design Trustworthiness, 342, 401
Subjectivity, 29, 160 Tuskegee Syphilis Study, 115t–116t
Subsets, 268–270 Twentieth century, nursing research in, 11–15, 12b
Subthemes, 338–339, 338t Type I errors, 363–366, 364f
Summative evaluation, 229–230 Type II errors, 363–366, 364f
Survey studies, 235b, 236–238 Typical purposive sampling cases, 275
Symmetry, 351
Synthesis understanding, 51b
Systematic data, 288 nauthorized research, 135
Systematic error, 324 nethical research studies, 115t–116t
Systematic reviews, 61–62, 245–248, 375–376, 376f nidimensional scale, 310
Systematic sampling, 271–272 nited Nations research goals, 19
niverse of content, 315
T nstructured observations, 294
t statistic, 368–369
t tests, 359t, 367–370
Table of random numbers, 267f Validation sample, 315
Target populations, 261, 279, 279f Validity
Telephone interviews, 298 critique of, 326b–328b
Testability Internet resources for, 308b
of hypotheses, 77, 86 of research instruments, 60, 307, 315–324, 323b
of research questions, 73–74, 74t threats to, 206–212, 207b, 208t, 210b
Testing effects, 207–208, 208t, 222 Values, 26–29, 27t–28t
Test-retest reliability, 309 Variables
Text, 148 data collection and, 288–291
Thematic analysis, 334, 337–338 defined, 41
Themes, 14–15, 15t, 334–335, 338–339, 338t in developing research ideas, 69, 69b, 74t
Theoretical framework hypotheses and, 42
choosing, 43b types of, 71–72, 73t
in historical research, 185–186 Verification, 341–342
in literature reviews, 92–93 Vignettes, 339
for research, 36–44, 38t, 58–59 Vulnerable groups, protection of, 131–133
Theoretical sampling, 177, 275–276, 430–431
Theoretical/empirical knowledge, 24–26, 25f
Theories. See also speci c theories Web browsers, 101–104
application of, 8b Web sites, 103t
defined, 42, 92 Weir report, 11–13
linking education/practice, 7–8 Willowbrook Hospital Study, 115t–116t
linking research/practice, 39–40 World Health Organization (WHO), 17, 19
Theory base, for hypotheses, 77 Worldview, 26, 40–41, 41f
Theory triangulation, 161
Theory-practice-research connection, 39f
Time series design, 225f, 227 scores, 356–359