MAKERERE UNIVERSITY SCHOOL OF PUBLIC HEALTH

MPH 7203: RESEARCH METHODS

Coordinating Department

Epidemiology and Biostatistics

Master of Public Health

Distance Education Programme

COURSE MATERIALS

2nd Edition
(December 2007)

1
 Contributors

Prof. Joseph Konde-Lule

Prof. Fred Wabwire-Mangen
Dr. Sheba Nakacubo Gitta
Dr. Patrick Okello
Ms Barbara Kirunda
Dr. Mayega Roy William
Dr. Danstan Bagenda

Instructional Design/Editing by

Dr. Mayega Roy William

Makerere University School of

Public Health
Edition 2
(December 2007)

2
TABLE OF CONTENTS
1.0 ABOUT THIS COURSE .......................................................................................................................... 6
1.1 COURSE TITTLE:............................................................................................................................... 6
1.2 GENERAL OUTLINE OF THE COURSE ............................................................................................ 6
1.3 INTRODUCTION ................................................................................................................................. 6
1.4 COURSE AIMS AND INSTRUCTIONAL GOALS .............................................................................. 7
1.5 TIME FRAME ...................................................................................................................................... 7
1.6 INSTRUCTIONAL MEDIA AND TECHNIQUES ................................................................................. 8
1.7 MODE OF EVALUATION ................................................................................................................... 8
1.8. EVALUATION OF THE INSTRUCTION PROCESS ........................................................................ 10
1.8.1 PRE-EVALUATION .......................................................................................................... 10
1.8.2 PRETEST ....................................................................................................................... 10
2.0 COURSE MATERIALS ......................................................................................................................... 11
2.1 Unit 1: CONCEPTS UNDERLYING RESEARCH I: DEVELOPING A RESEARCH PROBLEM ...... 11
2.1.1 Introduction to the Unit ..................................................................................................... 11
2.1.2 Unit Outline ..................................................................................................................... 11
2.1.3 Instructional goal.............................................................................................................. 12
2.1.4 Unit Objectives ................................................................................................................ 12
2.1.5 Time Frame..................................................................................................................... 12
2.1.6 Content ......................................................................................................................... 12
Session 1: Problem Identification and Analysis ...................................................................... 13
Session 2: Designing a Conceptual Framework in Research ................................................... 20
Session 3: Development of Research Objectives and Variables .............................................. 24
Session 4: Literature Review ................................................................................................ 26
2.1.7 Extension Activities .................................................................................................... 30
Extension Activity 1: Discussion Forum Question .................................................................... 30
Extension Activity 2: Self- Assessment Quiz ........................................................................... 30
2.2 Unit 2: CONCEPTS UNDERLYING RESEARCH II: METHODOLOGIES IN RESEARCH .............. 31
2.2.1 Introduction to the Unit ..................................................................................................... 31
2.2.2 Unit Outline ..................................................................................................................... 31
2.2.3 Instructional goal.............................................................................................................. 31
2.2.4 Unit Objectives ................................................................................................................ 31
2.2.5 Time Frame..................................................................................................................... 31
2.2.6 Content ......................................................................................................................... 31
Session 1: The Study Setting, Study Population and Study Designs ........................................ 32
Session 2: Sample Size Determination.................................................................................. 35
Session 3: Sampling Procedures .......................................................................................... 42
Session 4: Selection of Study participants ............................................................................. 47
Session 5: Quantitative and Qualitative Data Collection Methods ............................................ 50
Session 6: Quantitative and Qualitative Data Collection Tools ................................................. 56
Session 7: Study Variables and Measurements ..................................................................... 63
Session 8: Data Management and Quality Control in Research ............................................... 67
Session 9: Ethical Considerations in Development of a Research Problem ............................... 70
Session 10: Workplan and Budgeting .................................................................................... 74
Session 11: Alternative Study Designs ................................................................................... 76
2.3 Unit 3: QUALITATIVE RESEARCH METHODS .............................................................................. 80
2.3.1 Introduction to the Unit ..................................................................................................... 80
2.3.2 Unit Outline ..................................................................................................................... 80
2.3.3 Instructional goal.............................................................................................................. 80
2.3.4 Unit Objectives ................................................................................................................ 80
3
 2.3.5 Time Frame..................................................................................................................... 80
2.2.6 Content ......................................................................................................................... 80
Session 1: Concepts in Social Health Research ...................................................................... 81
Session 2: Qualitative Research and its role in Epidemiological Inquiry ..................................... 83
Session 3: Social Science Methods Appropriate In Public Health.............................................. 87
Session 4: Rapid Assessment Procedures ............................................................................. 92
Session 5: Design of Qualitative Data Collection Tools ............................................................ 98
Session 6: Qualitative Data Analysis .................................................................................... 101
Session 7: An In-depth view of Triangulation ........................................................................ 106
2.2.7 Extension Activities ........................................................................................................ 109
Extension Activity 1: Discussion Forum Question .................................................................. 109
Extension Activity 2: Self- Assessment Quiz ......................................................................... 109
2.4 Unit 4: RESEARCH PROPOSAL DEVELOPMENT, IMPLEMENTATION AND WRITING SKILLS
.............................................................................................................................................................. 111
2.4.1 Introduction to the Unit ................................................................................................... 111
2.4.2 Unit Outline ................................................................................................................... 111
2.4.3 Instructional goal............................................................................................................ 111
2.4.4 Unit Objectives .............................................................................................................. 111
2.4.5 Time Frame................................................................................................................... 111
2.4.6 Content ....................................................................................................................... 111
Session 1: Research Proposal Format and Writing Skills ...................................................... 112
Session 2: Tips in Implementing the Research Proposal ....................................................... 119
Extension Activities ................................................................................................................ 122
Extension Activity 1: Discussion Forum Question .................................................................. 122
Extension Activity 2: Self- Assessment Quiz ......................................................................... 122
2.5 Unit 5: DATA ANALYSIS AND REPORT WRITING ...................................................................... 123
2.5.1 Introduction to the Unit ................................................................................................... 123
2.5.2 Unit Outline ................................................................................................................... 123
2.5.3 Instructional goal............................................................................................................ 123
2.5.4 Unit Objectives .............................................................................................................. 123
2.5.5 Time Frame................................................................................................................... 123
2.5.6 Content ....................................................................................................................... 123
Session 1: Data Analysis ................................................................................................... 124
Session 2: Research Report Writing ................................................................................... 128
Session 3: Dissertation, Abstract and Manuscript Formats .................................................... 130
2.4.7 Extension Activities .................................................................................................. 132
Extension Activity 1: Discussion Forum Question .................................................................. 132
Extension Activity 2: Self- Assessment Quiz ......................................................................... 132
3.0 ADDITIONAL RESOURCES ............................................................................................................... 133
3.1 TEXT DOCUMENTS FOR ADDITIONAL READING ............................................................ 133
3.2 GLOSSARY OF TERMS ................................................................................................... 133
3.3 REFERENCES ................................................................................................................. 133
3.4 ANSWERS TO QUIZ QUESTIONS .................................................................................... 134
3.5 INDEX OF URLs FOR INTERNET RESOURCES ................................................................ 135
3.6 INDEX OF DISCUSSION FORUM QUESTIONS ................................................................. 135
3.7 INDEX OF ADDITIONAL RESOURCES FOLDER ............................................................... 135
3.8 INDEX OF SELECTED LECTURE NOTES ......................................................................... 135
3.9 SUMMATIVE EVALUATION OF THE INSTRUCTION PROCESS ......................................... 136
Progressive Assessment – Hand-in Assignments.................................................................. 136

4
Post-test............................................................................................................................ 136
Post Evaluation .................................................................................................................. 137

5
 Makerere University School of Public Health
Master of Public Health – Distance Education Programme

1.0 ABOUT THIS COURSE

1.1 COURSE TITTLE: RESEARCH METHODS

1.2 GENERAL OUTLINE OF THE COURSE

This course will cover the following areas:
UNIT 1: CONCEPTS UNDERLYING RESEARCH I: DEVELOPING A RESEARCH PROBLEM
UNIT 2: CONCEPTS UNDERLYING RESEARCH II: METHODOLOGIES IN RESEARCH
UNIT 3: QUALITATIVE RESEARCH METHODS
UNIT 4: RESEARCH PROPOSAL DEVELOPMENT, IMPLEMENTATION AND WRITING SKILLS
UNIT 5: DATA ANALYSIS AND REPORT WRITING

1.3 INTRODUCTION
Welcome message from the Course Coordinator: I wish to welcome you to this exciting ‘science and art’
of research methodology. As you will learn, Research Methods is a cross-cutting discipline allowing both
the elucidation and discovery of new knowledge in both arts and science subjects. The ability to identify
problems and to conduct research to seek solutions for the identified problems is a core competence in the
health services. This course takes a practical approach, translating theory and assumptions about the
everyday problems and challenges faced in public health practice into new knowledge and the application
of that knowledge to the control of such problems.
About this course: This course introduces you to the principles, concepts and terminologies used in
research methods and their application in public health. Research is a core discipline of public health. This
course will cover the following areas:
- Introduction to research methods and its uses
- Identifying a research question
- Developing a research proposal
- Designing and conducting a study
- Analysis of data
- Writing a research report (field report, Dissertation)

Contributors and resource persons:

I wish to extend sincere thanks to Ms. Barbara Kirunda and Dr. Patrick Okello for their invaluable help in
putting together the course materials. May I take this opportunity to introduce to you all the other facilitators
for this course: Prof Fred Wabwire-Mangen, Dr. David Guwatudde, Dr David Ndungutse, Mr. David
Mukanga Odaka, Mr. Lynn Atuyambe, Dr. Robinah Najjemba, Mr. Simon Kasasa and Dr. Sheba Nakacubo
Gitta. If you have any specific problems related to this course, feel free to contact the course coordinator
for guidance and support [Email:].

6
The Course Coordinator

Dr. Danstan Bagenda holds a PhD in Epidemiology from the Case

Western Reserve University in Cleveland Ohio. He is also a
Statistician and has worked in several research programmes and key
clinical and field trials. He worked for the Rakai Health Sciences
Programme, the Makerere University-Johns Hopkins Collaboration
and the Department of Obstetrics and Gynaecology and Makerere.

He is currently a Lecturer in the Department of Epidemiology and

Dr. Danstan Bagenda
Biostatistics and is involved in current research especially on
HIV/AIDS, including vaccine trials and sero-behavioural surveys.

The Instructional Designer:

This course guide has been designed, edited and enhanced by Dr. Mayega Roy William. Dr. Mayega
holds a Bachelor of Medicine and Bachelor of Surgery Degree and is a graduate of the Master of Public
Health Programme. He is charged with the pedagogical aspects of the MPH (DE) Programme, including
materials development, quality development and moderating the methods of delivery. His current interests
are in the areas of Epidemiology, Instructional Design and e-Learning. Specific issues related to the content
and design of the materials may be routed to him on the address de_materials@musph.ac.ug.

1.4 COURSE AIMS AND INSTRUCTIONAL GOALS

1.4.1 Aim
The overall aim of this course is to equip the MPH Officer with the skills to design, conduct and evaluate
research projects on selected health problems of Public Health importance using internationally acceptable
standards and methodologies.
1.4.2 Instructional Goals
By the end of this course, the MPHO should be able to:
1. Select and analyse research problems so as to formulate research objectives and questions

2. Select, develop and use appropriate methodologies in conducting research in a given research
problem

3. Write and implement a research proposal, incorporating appropriate methodologies to address

specified research questions

4. Analyse data and prepare a scientific report for a given research design

1.5 TIME FRAME

You will cover this course during the second semester of the first academic year of the MPH Distance
Education Program. The semester runs for a total of 17 weeks including time for face to face lectures and
examinations.

7
1.6 INSTRUCTIONAL MEDIA AND TECHNIQUES
1. Orientation Session: Brief instructor led face-to-face orientation sessions will be delivered at the
School of Public Health at the beginning of the course. The purpose of the sessions is to give you
an overview of what you are expected to read about, and guide you on the resources available.
2. Print material: You will be provided with handouts of print materials (MUIPH Distance Education
Resource Kit) and other materials. You will also be provided with printed case studies where
necessary. There is an additional resources kit containing a collection of handouts and readers that
have been sorted for you. You will be expected to acquire this at your own cost. We highly
recommend that you get yourself a copy of this kit. Copies are available at the Business Centre at
the going rate for photocopying services in the school.
3. E-mail: E-mail shall be the main means of communication for submission of assignments,
announcements and consultation with faculty. Please ensure that you obtain a reliable e-mail
address and register it with the Programme Administrative Secretary. In case you change your e-
mail, please promptly notify the same officer.
4. E-Learning Platforms and Tools: Learning Management tools shall be used to conduct online
discussion forums and chatting. You will be informed in due course about the site to be used for
hosting these interactive activities.
5. Textbooks, internet and independent study: You are required to search for the relevant
references and acquire the core recommended readings for each course. There are also many
internet sources from which you can obtain information.
6. Activities, examples and exercises: Please note that the biostatistics course involves a lot of
calculations. It is important, especially, that you do all the exercises and examples given, so that
you internalize the application of the statistical methods they convey.

Note to the Reader!

The MPH Distance Education Programme is committed to providing the best possible distance learning
environment for you. As such, the Programme has invested in the development of high quality user-friendly
instructional materials that will better facilitate students’ learning. A full time Instructional designer/editor
has been appointed for this purpose. The materials shall mainly be available in print format. CD-ROMs,
interactive media, e-learning platforms and any other viable resources may from time to time be used by
the Instructors to reach out to you. Assignments will be sent to your e-mail box when you have registered at
the beginning of the semester. There is also an additional resources kit that contains Additional Readers,
worked examples and exercises. You should endeavour to acquire a set of these readers during the face-
to-face session.

1.7 MODE OF EVALUATION

1.7.1 Self Assessment
Self-evaluation: You are encouraged to attempt a set of exercises and questions at the end of each
presentation for your own assessment. Assignments and activities in the course are for you to test yourself
and evaluate your performance. They will enable you gauge your understanding of the content. Answers to
these questions may be availed in the Additional Resources section of this document. You will also be
required from time to time to contribute to specific discussion points that have been set up for analysis on
the discussion forums.
Participation in learning activities: We may also request you to contribute meaningfully to the topics put up
on the discussion boards and where indicated, your contribution may be graded.
Practice: Research methodology is a practical discipline. You are encouraged to identify research
questions to which you can apply the knowledge gained.

8
 1.7.2 Progressive Assessment – Hand-in Assignments
In line with the University regulations, you will be evaluated in two segments; progressive assessment,
which accounts for 30% of the total mark and the end of semester university examination which accounts
for the remaining 70%. Apart from the self-assessment exercises, the institute shall require you to hand in
one or more assignments for marking. These assignments may be in form of quizzes, structured
questionnaires, long or short answer questions, term papers or project reports, to be forwarded on-line. An
assignment for assessment will be indicated and will either be given to you at the time of the face-to-face,
or forwarded to you by mail or internet. Please pay close attention to the deadlines for handing in the
assignments. The progressive test will account for 30% of the overall course score.

TAKE NOTE THAT THE PROGRESSIVE ASSESSMENT IS A PRE-REQUISITE FOR THE END
OF SEMESTER EXAMINATION.

1.7.3 University Exam

You will sit for a final course examination at the end of the semester to be held at the School of Public
Health. You must make arrangements to travel to the School for this examination once the date has been
communicated to you. This will contribute 70% of the final mark

9
1.8. EVALUATION OF THE INSTRUCTION PROCESS

1.8.1 PRE-EVALUATION
NOTE: Before using these materials, you are kindly requested to fill the Evaluation Questionnaire for this
semester and send it to the Instructional Designer. This questionnaire is not a test, but it will enable us to
measure your expectations from the instruction process and to gauge how much you will benefit from the
instruction materials for this semester, as well as inform us the extent to which the materials given in the
previous semester assisted you. The questionnaire has two parts: A Post Evaluation of the previous
materials and a Pre-evaluation of the materials that you expect in the new semester. In the post evaluation
section, you are requested to evaluate the materials that you received in the last semester. In the pre-
evaluation section, you are expected to inform us of the key competencies and qualities you expect from
the new materials. Your responses will be compared with responses in the Post evaluation section at the
beginning of the next semester. The information generated will be used in an iterative process designed to
improve the materials. This evaluation may be administered during the face-to-face session, before the
materials are dispatched. In the event that it is not delivered then, you can access it in your Additional
Resources Folder. In case you do not fill it during the Face-to-face sessions but fill it later, you are
requested to send the completed questionnaire to the Instructional Designer at the e-mail address:
de_materials@musph.ac.ug.

1.8.2 PRETEST
NOTE: It is important that before you read these materials, you complete a Pre-test. The purpose of this
test is to make a baseline assessment of what current knowledge you have. It is also an important guide to
what areas you need to emphasise in your reading. For some courses, where the Course Coordinator
considers it a requirement, the test may be administered at the time of face-to-face, during the introduction
to this course. Otherwise for other courses, it is strictly optional, but you are encouraged to take it prior to
your reading. The test is also contained in the Additional Resources Folder.

10
 2.0 COURSE MATERIALS

2.1 Unit 1: CONCEPTS UNDERLYING RESEARCH I:

DEVELOPING A RESEARCH PROBLEM
2.1.1 Introduction to the Unit

Let us start by posing the question: What is research?

In day-to-day language, Research has variously been defined as either a quest for new knowledge or the
act of searching for knowledge so as to elucidate some concept or derive new meaning and understanding
of some phenomenon.

For our purposes, we will define research as the systematic (organized) collection, analysis and
interpretation of data to answer a certain question or solve a problem.

In relation to the mission and objectives of the MPH training, being able to design, conduct research and
disseminate the findings of the research are considered as core competencies expected of an MPH
graduate.

The MPH trainee should be able to:

 Conduct basic research – necessary to generate new knowledge and technologies to deal with
major unsolved health problems
 Conduct applied research – necessary to identify priority problems and to design and evaluate
policies and programs that will deliver the greatest health benefits and make optimal use of the
available resources
Thus, research is major function of Makerere University Institute of Public Health.

This document will guide the learner on the basic steps in designing and conducting research.

During the course of MPH training, the learner should be able to:
 Recognize health problems, prioritize the problem, define and specify a research question and
state the problem in a researchable way,
 Develop a research proposal, conduct the study, and report the findings

Research often starts by identification of a researchable problem. In this unit, we shall consider the first part
of the formative aspects of a research process by looking at aspects of problem identification, problem
analysis and development of research objectives.

2.1.2 Unit Outline

The following topics will be covered:
Session 1: Problem Identification and Analysis
Session 2: Designing a Conceptual Framework in Research
Session 3: Development of Research Objectives and Variables
Session 4: Literature Review

11
2.1.3 Instructional goal

The MPHO should be able to select and analyse research problems so as to formulate research objectives
and questions

2.1.4 Unit Objectives

By the end of this unit, the student should be able to:

1. Identify researchable problems
2. Use conceptual frameworks to analyse research problems and elucidate their underlying
factors
3. Develop SMART research objectives, given a research problem
4. Conduct meaningful literature review and literature citation for a given research problem

2.1.5 Time Frame

2.1.6 Content

12
Session 1: Problem Identification and Analysis
Introduction: The research process often starts with identification of a researchable problem. In this
session, we shall discuss the principles of problem identification and problem analysis in order to formulate
a research problem.

Session Topics: The following topics will be covered:

a. Identifying a health problem
b. Problem Prioritization
c. Problem Analysis
d. Statement of the problem

Session Objectives:
By the end of this lesson, the MPHO should be able to:
1. Identify researchable problems in health service delivery
2. Compare a given set of identified research problems so as select priority problems for
research
3. Conduct problem analyses of given research problems to identify their probable causes
4. Organise facts from problem analyses to write a meaningful problem statement

a. Identifying a health problem

The challenge of the health system in Uganda is to promote, improve and maintain the health of Ugandan
people equitably and to the optimal level possible with the resources available. In order to meet this
challenge, the MPH training at MUIPH attaches significant importance to acquisition of knowledge and
skills in research methodology.

A proposal writing process always begins with a statement of the problem. Finding a problem is not difficult
but identifying one for the purpose of research is not always easy. The first and most important task of
research process is to identify and define clearly the problem one wishes to study.

The following may help you to identify a study problem:

 Some unknown or missing information concerning a public health issue/ problem
 A perceived difference or discrepancy between what exists and the ideal or planned situation
(what is and what should be)
 The reason(s) for this difference may be unclear so that it makes sense to develop a research
question
 There may be more than one possible answer to the question or solution to the problem

Example of a research problem:

Problem situation:
A recent study revealed great differences among villages in the prevalence of HIV positive persons.
Despite the fact that all villages receive the same level of health education and services from the Ministry of
Health, some villages have an HIV prevalence rate as high as 32% among adults from 15 – 49 years old
while other villages have a rate as low as 6%.

13
Discrepancy
In a relatively small geographical area, you would expect that all villages would have approximately the
same sero-prevalence rate but in fact there is great variation among villages.

Problem question
Which factors are responsible for the geographic variation of HIV prevalence between villages?

Possible answers
Villages differ in their socio-economic environments, and these differences influence the context within
which HIV is transmitted. Some are stable agricultural villages while some are mobile fishing communities.
Some villages are located on major roads and have easy access to market towns and others are more
remote with very difficult access to markets. Some villages have schools, health centres, electricity and a
good water supply while others do not. These and many other socio-economic and cultural differences
affect the context within which sexual relations take place and HIV is transmitted.

Villages differ in individual and institutional support for HIV/AIDS prevention, care and support programmes.
In some villages, influential local leaders strongly support sexual behavior change and condom distribution
programs. In other villages people are resistant to these programs and there is substantial stigma and
discrimination associated with HIV/AIDS. In some villages there are very active anti-AIDS clubs for youths,
strong PHLA organizations and effective orphan care NGOs while in others, these institutions are absent.
These differences in individual commitment to and institution support for HIV/AIDS programs affect the
sexual behavior of individuals, the use of condoms, the level of stigma and discrimination and the
transmission of HIV.

While the problem situation presented above is fairly clear, the possible and plausible reasons are many.
Each possible explanation presents a different research question. The unknown answers are problems.

In situations such as these, the researcher must devote considerable time and attention to identifying and
clearly defining the problem situation before any potential solutions to the problem can be tested. The aim
of clearly identifying and defining a problem situation is to focus the research on the most important aspects
of a problem that can be changed through a programme intervention.

Does a problem require research?

Whether a problem situation requires research depends on the following questions:

1. Is there a problem? There ought to be a perceived difference or discrepancy between what exists and
the ideal or planned situation

2. Does the problem require research? The reason (s) for this difference should be unclear (so that it
makes sense to develop research questions). In addition, there should be more than one answer to the
problem

3. Is the problem a programmatic problem? Think of ways in which the problem can be addressed by
the programme:
Is it possible to develop a feasible, effective and sustainable intervention?
What health improvements will result from solving the problem this way?
Is it an important problem? To whom is it important?

14
Problems and opportunities: Not all problems mean something is wrong with the program. Some
problems are opportunities, for instance: What is the best way to increase the size of this programme to
serve more people?

4. Is the problem a researchable problem? We ought to be able to answer the following questions:
Do I need to do research to solve the problem?
Can my question be answered correctly by common sense and experience?
Can I gent enough time, money and qualified persons to do research
Will those responsible for making decisions and taking action listen to the research results?

Non-research Research
1. A discrepancy exists between 1. A discrepancy exists between desired
desired and observed situation and observed
2. We know why the discrepancy 2. We don’t know why the discrepancy
exists exists
3. We know the best solution 3. Two or more potential solutions exist
and we do not know which is the best

Before we can analyze a problem and write a good statement, we must answer a more important question:
Is this problem researchable and worthy of our time and resources? Whether a problem situation requires
research depends on three conditions:

1. There should be a perceived difference or discrepancy between what exists and the ideal or
planned situation
2. The reason(s) for this difference should be unclear (so that it makes sense to develop a research
question; and
3. There should be more than one possible answer to the question or solution to the problem

Example:
In District X, sanitary conditions are poor (5% of households have latrines) and diseases related to poor
sanitation such as gastro-enteritis and worms rank highly in the burden of disease profile. The Ministry of
Health has initiated a sanitation project that aims at increasing the number of households with latrines by
15% each year. The project provides materials and the population should provide labor. Two years later,
less than half of the target has been reached.

Discrepancy: 35% of the households should have latrines and only 15% do have them.
Research Question: What factors can explain this difference?
Possible answers:
1. Service related factors: Not informing and involving the community, bottlenecks in the supply of
materials, differences in training and effectiveness of sanitary staff.
2. Population-related factors Community members’ lack of understanding of the relationship between
disease and sanitation or have a greater interest in other problems

b. Problem Prioritization

Very often, we find that health needs are infinite and communities have many concurrent problems.
Likewise, possible research problems are also many and the researcher has to decide on which is the most
pertinent problem to delve into, in the context of available time, resources and the most pressing priorities.

15
The researcher therefore has to prioritize the research problems to select the most appropriate one to
undertake.

Criteria for prioritizing health problems for research

Because research is intended to feed into decision-making to improve health and health care, the selection
and analysis of the problem for research should involve those who are responsible for the health status of
the community as well as the community itself.

Each problem proposed for research has to be judged according to certain guidelines or criteria. There are
usually several ideas to choose from.

Criteria:
1. Relevance: How large or widespread is the problem? Who is affected? How severe is the problem?
2. Avoidance of duplication
3. Feasibility
4. Political Acceptability
5. Applicability of possible results and recommendations
6. Urgency of data needed
7. Ethical acceptability

Activity:

Identify and state a health problem in your district/ work environment following the criteria above.

c. Problem Analysis

The researcher is often required to conduct research on a problem, which he or she is not very familiar
with. In the health domain for example, health workers and managers or the community members may be
much more familiar with the problem than the health worker – however, they too may never have given due
attention to the various aspects of the problem.

A systematic analysis of the problem, completed jointly by the researches and the stakeholders is a very
crucial step in designing the research because it:
1. Enables those concerned to pool their knowledge of the problem
2. Clarifies the problem and the possible factors that may be contributing to it
3. Facilitates decisions concerning the focus and scope of the research

Steps in analyzing the problem

Step 1: Clarify the viewpoint of the stakeholders/researchers in relation to the problem

Stakeholders often express areas of concern in broad or vague terms for example:

“Care of diabetic patients needs review”

“Outpatient services must be reviewed”
“Bypassing of peripheral facilities should be investigated”

16
Clarify the issues by listing all the problems in the area of concern, as they perceive them. Remember that
a problem exists when there is a discrepancy between “what is” and “what should be”. Therefore, the
perceived problems should be worded in such a way as to illustrate this discrepancy. For example, health
care managers and workers may determine that the general concern that “care of diabetic patients needs
review” includes the following problems:
- Insufficient awareness on diabetes and on self care measures among diabetic patients and their
relatives
- Insufficient peripheral facilities for the long-term follow-up care
- Excessive rate of readmissions among diabetics
- Inappropriate management of complications in diabetic patients
- High rate of diabetic complications
- Poor compliance of patients with therapy

Step 2: Further specify and describe the core problem

Try to identify the core problem and quantify it. In our diabetes example, the core problem may be: the high
rate of readmissions among diabetics
- Attempt to describe the nature of the problem; the discrepancy between “what is” and what you
prefer the situation to be in terms of readmissions
- Determine the distribution of the problem: Who is affected, when and where
- The size and intensity of the problem: Is it widespread? How severe is it? What are its
consequences? (Disability, death, waste of resources)

Step 3: Analyze the problem

After identifying the core problem, you should:

- Identify factors that may have contributed to the problem
- Clarify the relationship between the problem and contributing factors. It is helpful to form a
diagram, called the conceptual model. Draw the core problem in a double semi-circle at the centre
of a flip/blackboard and brainstorm on possible causes or factors contributing to the problem.
Remember to include the viewpoints from step 1.

Step 4: Identify other contributing factors

It may be possible to identify several “generations” of predisposing factors. It is desirable to continue

identifying underlying contributing factors until you reach basic factors that need to be modified to solve the
problem, and that can be modified within the existing context. This will facilitate the formulation of research
projects that can provide useful information for decision-making. This process of continued analysis will
necessitate several revisions or extensions of the initial analysis diagram. The final diagram should
encompass all the critical factors that may be contributing to the problem studied.

Attempt to organize related factors together into larger categories and develop your final draft of the
diagram. This final step will help you not to overlook important factors and will make it easier to develop the
data collection tools in a systematic way (see lesson on conceptual framework).

If your problem is complex and has many possible contributing factors, identify and demarcate the
boundaries of possible smaller research topics. If there is more than one possible topic, select the priority
topic for your focus and scope of research using earlier discussed criteria.

17
NB: The dissection of the diagram into different parts and selection of one part for research is not advised if
insufficient insight exists into the nature, relative weight and interrelations of the various factors contributing
to the problem. You would risk concentrating on marginal factors and coming up with marginal solutions.

For example, it is inadvisable to concentrate only on community factors or only on service factors to explain
underutilization of services if you do not know how these factors are interrelated and where the main
problem is. An exploratory study would then be indicated.

d. Statement of the problem

The first major section in a research proposal is “statement of the problem”. The statement of the problem
answers 3 cardinal questions:

1. What is the problem?

 Brief description of the socio-economic and cultural characteristics and overview of the health
care system in as far as these are relevant to the problem. Include a few illustrative statistics, if
available, to describe the context in which the problem occurs
 A description of the nature of the problem i.e. the difference between what is and what should
be or real vs. ideal
2. What is the magnitude of the problem?
 A description for the size, distribution and severity of the problem i.e. who is affected, where,
since when and what are the consequences for those affected and for the services
3. What are the main issues of interest and/or concern that need to be studied?
 Make and analysis of the major factor that influence the problem
 Describe any solutions that have been tried in the past and how well they have worked
 Discuss existing gaps in knowledge that need to be investigated
 Make convincing arguments that available knowledge is insufficient to fill the gaps and
therefore further research is needed.
 Describe the type of information expected to result from the project and how this information
will be used to help solve the problem

Remember:
1. Be coherent, have a logical flow and above all, be highly persuasive (argue in quantitative terms in
order to remain objective)
2. Keep it short, one page double spacing, one and a half pages at most; preferably one page

References:
1. Varkevessier C., Pathmanathan I., Brownlee A, (1995): “Designing and conducting Health Systems
Research Projects” Volume II, part 1
2. Polgar S., Shane T., (2001): “ Introduction to Research in the Health Sciences”
3. Wolski R., (2004): “Presenting a Proposal”.

Activity: Form groups of 3 members each. Using the problem statements you developed during the Field
Study I, spend sometime ranking and prioritizing the problems. Based on the ranking select one of these
problems for use in this exercise.

18
Write up a concise and persuasive statement of the problem to busy potential funders. Discuss the
statement with a colleague. Use 2-3 member discussion groups; you can discuss the problem on line.

Guideline: You will write the problem using the following structuring:

Paragraph 1: Brief explanation of the context

Statement 1: A brief statement of the context, with regard to the organization of service delivery or
what is on the ground
Statement 2: Follow up explanation of the context if any

Paragraph 2: Problem Description

Statement 1: State the actual problem in one sentence, clearly indicating the gap between “what is”
and “what is expected”.
Statement 2: State the magnitude of the problem; “Who is affected? When? Where?” Back this up
with a few statistics
Statement 3: State the consequences of the problem
Statement 4: Outline the possible explanatory factors

Paragraph 3: Main issues that can be researched

Statement 1: What has been attempted in the past till current, in relation to the problem?
Statement 2: What gaps still linger?
Statement 3: What gaps does this study intend to address?

19
Session 2: Designing a Conceptual Framework in Research
Introduction: One of the tools of Problem Analysis is the Conceptual Framework. In this session, we shall
discuss the rationale and elements of a conceptual framework.

Session Topics: The following topics will be covered:

a. Nature and Characteristics of a Conceptual Framework
b. Steps in analyzing the problem to develop a framework and Model

Session Objectives:
By the end of this lesson, the MPHO should be able to:
1. Describe the nature, rationale and components of a conceptual framework for a research
problem
2. Organise the different probable explanatory factors for a research problem into a conceptual
framework

Specific References:
1. WHO/IDRC HSR Manual Vol. II: Designing and Conducting Health Systems Research Projects.
Part 1: Proposal Development and Field Work
2. Dissertation Workshop CD from John’s Hopkins University; School of Public Health
3. Porter R.W. Prysor-Jones S: Making a Difference to policies and Programmes: A guide for
Researchers. SARA Projects/USAID July 1997
4. Leon Gordis: Epidemiology

a. Nature and Characteristics of a Conceptual Framework

What is a conceptual Framework? It is a broad theoretical construct or model linking concepts and sub-
concepts, hypotheses and sub-hypotheses that describe the problem. It identifies the factors that may
contribute to the problem and clarifies their relationship through diagrammatic links. It is synonymous with a
problem analysis diagram.

In disease related research, the model pertains to pathogenesis and logically connects several different
hypotheses which are causally related. For some research (especially descriptive or hypothesis generating)
it may be difficult to develop a conceptual framework. Theory adds great depth to the proposal but must be
plausible and acceptable.

Relationships in a conceptual framework

For cause-effect relationships, one-way arrows are used. For mutual relationships, two way arrows are
used. The core problem provides the central theme of the framework from which relationships emanate.
Factors are usually grouped e.g. service related, community-related etc.

Breadth of framework vs. focus

Many conceptual frameworks are too broad and only parts of it can be the subject of study. To investigate
all parts of a broad model would result in an excessively broad unfocussed project. If the specific objectives

20
are closely interdependent and the proposal is focused, then a conceptual framework is possible. The
conceptual framework services to keep the work focused.

Deciding the focus and scope

Usefulness of the information: Will the information collected on this problem help to improve health and
health care? Who will use the findings related to the factors that will be studied? How will the findings be
used?
Feasibility: Is it feasible to analyze all the factors related to the problem in the time available for the study?
Lack of Duplication: Is some of the information related t the factors that influence the problem already
available?

Rationale for a Conceptual Framework

 It enables those concerned to pool their knowledge of the problem
 It clarifies the problem and the possible factors that may be contributing to it
 It facilitates decisions concerning the focus and the scope of the research

b. Steps in analyzing the problem to develop a framework and Model

 Clarify the view points of the managers, health care workers and researchers in relation to the
problem
 Further specify and describe the problem
 Place the problem at the centre
 Highlight the different suspected factors linked to the problem
o A hierarchical pattern can be used
 Underlying factors (Distal) factors
 Proximal factors
 Immediate causes
o A categorical pattern can be used e.g.:
 Community factors
 Individual factors
 System factors etc.
 Draw lines to show the interrelationships; in epidemiology, most factors are neither necessary nor
sufficient; we expect more complex interrelationships between the factors
 Revise the concept to make it smarter
 Write a brief narrative to explain the salient relationships
 You can now use the framework to develop your research questions and objectives

Elements of a conceptual framework

Factor

Problem

Factor

21
Example: Analysis of factors causing high defaulter rate among TB patients

Factors
- Staff not trained
- Other service factors
- Other patients and community-related factors
- Little understanding of patients’ perceptions of TB treatment
- High defaulter rate of TB patients
- Inefficient staff distribution
- Patient does not understand requirement
- No adequate materials or guidelines on TB health education
- No systematic advice and counseling provided

Conceptual Framework

Other
service High defaulter
factors rate of TB
Patients

Other patient
and community Patient does
related factors not understand
requirement

Practical Exercise: Designing a conceptual framework in research

1. Form groups of three people: Buzz groups (Face to Face) or Chat groups (online)
2. For each group:
a. Each group member should retrieve the 2 problem statements that you developed for the
first field attachment. Spend some time ranking and prioritizing these problems. Bases on
your ranking, select one of these problems for use in this exercise and read the problem
thoroughly
b. Develop a refined problem statement
3. Place the core problem in the centre of a chart, a blackboard of an internet discussion
sheet/interface
4. Using personal knowledge, experience and available literature, each group member should silently
and independently write a list of contributory factors to the core problem

22
5. Display all contributory factors identified by group members on the flip chart or blackboard or
worksheet so that it is visible to all members. Encourage analytical thinking by asking group
members to determine which way the arrows should point between the various factors
6. Group the factors into categories. Pin or paste the factors around the core problem that you pinned
in the centre of the flip chart or blackboard. Move and re-group the factors around until the group is
satisfied with the resulting conceptual framework. Pay particular attention to the linkages and inter-
linkages. Revise as necessary
7. Narrow the scope and focus by considering whether the problem is too large to complete within the
available resources
8. Write 1 to 2 paragraphs of text describing you conceptual framework

23
Session 3: Development of Research Objectives and Variables

Introduction
Definition of Research: Research is a systematic collection, analysis and interpretation of data or
information, to answer certain questions or solve a problem.

Steps in conceptualizing a researchable problem:

 What is the problem?
 What is the available information?
 How do we want to carry out the research?
 What do we hope to achieve?
 How do we collect the information?
 Who will do that and when?

When we have these questions answered, we have to set our research objectives. In this session, we shall
describe the key issues involved in setting research objectives.

Session Topics: The following topics will be covered:

a. Formulation of research objectives
b. Why develop research objectives?
c. How should you state your objectives?

Session Objectives:
By the end of this lesson, the MPHO should be able to:
1. Illustrate the importance of research objectives
2. Formulate SMART research objectives for a given research problem

a. Formulation of research objectives

The most important aspect of the research proposal is the objectives. The objectives will guide the
researcher into:
 What type of information to collect about the research problem and what type of literature to
review?
 What type of methods to use to answer the different research questions?
 What types of variable to look out for and measure?
 What type of data analysis to conduct and how to present the findings?

Many researchers structure their literature review and their results section based on the objectives of the
study.
Objectives summarize what is to be achieved in the study. They should be closely related to the stated
problem. Objectives should be multiple – it is difficult to find a study where you have only one objective.
There should always be a general objective and specific objectives.

The General Objective:

24
It states what is expected to be achieved by the study in general terms. It summarizes the specific
objectives. It is often much related in wording with the title of the study.

Specific Objectives:
These further break down the general objectives into smaller areas for consideration. They systematically
address the various aspects of the problem and specify what one will do in the study, where and for what
purpose. Specific objectives should be SMART i.e.
S – Specific
M – Measurable
A – Achievable
R – Realistic
T – Time bound

However, in stating the specific objectives, time are always implicit i.e. we do not state the time repeatedly
for each.

b. Why develop research objectives?

 To focus the study (narrowing it down to the essentials)
 To avoid collection of data that is not strictly necessary for understanding and solving the problem
identified
 To organize and compartmentalize the study into clearly defined parts

How do we develop the different areas for setting objectives?

- Consider the conceptual framework
- The objectives start from the conceptual framework
- We consider the different ‘constructs’ that make up the conceptual framework
- Identify the sub-categorizations of the different constructs in the conceptual model
- Select the main themes that can realistically be studied within the resources you have (they should
answer your problem)
- Formulate the objectives
- NB: It may not always be possible to study all the constructs in a conceptual framework

c. How should you state your objectives?

- They should cover the different aspects of the problem and contributing factors in a coherent way
and in a logical sequence
- They should be clearly phrased in operational terms specifying exactly what you are going to do,
where and for what purpose
- Use action verbs that are specific enough to be evaluated

d. Research Questions

From the research objectives, the researcher may derive some research questions. The questions may be
singular or a number for each objective, but they must be related to the objectives

25
Session 4: Literature Review
Introduction: In formulating a good basis for research, we have to review available literature related to the
research questions and objectives. This helps us to solicit available information identify gaps in current
knowledge. In this session, we shall consider a few principles related to reviewing relevant literature and
citing references in a research protocol.

Session Topics: The following topics will be covered:

a. Definition and Rationale for a Literature Review
b. Stages in Literature Review
c. Information sources in a literature review
d. Organization of Literature review
e. Literature Citation and Referencing
f. Plagiarism

Session Objectives:
By the end of this lesson, the MPHO should be able to:
1. State the rationale for conducting a good literature review
2. Outline the stages of a literature review
3. Describe the different information sources for a literature review
4. Conduct meaningful literature review and organise the findings appropriately
5. Conduct appropriate literature citation and referencing
6. Differentiate literature review from plagiarism

a. Definition and Rationale for a Literature Review

What is Literature Review? Literature review is an assessment and narrative of what has been written on
a topic by accredited scholars. It places your work in the context of established work in the field. It provides
dimensions of current work and provides an up-to-date review of the topic. It is a recap of the information
and re-organization of that information, a summary and a synthesis of new ideas.

Why write a Literature Review?

 To increase your breadth of knowledge in your subject area

 To identify the current state of research in your field of study and identify trends
 To identify important works and authors in your area
 To identify gaps in the literature and areas that need further inquiry
 To critique work that has been done so as to discern gaps
 To demonstrate the command and understanding of your field of study
 To provide the background to and justification for the research undertaken
 To give your research a conceptual framework
 To discern possible explanatory factors that can be the basis for developing variables and
questions

Role of the Literature review

It helps to clarify what is already known
It identified gaps in current knowledge
26
Formulation of research questions and clarification on the limits and scope of the study
Describes the relationship of each work to other works
It places one’s original work in the context of others

b. Stages in Literature Review

Problem formulation: The topic, its component issues, theory and methodology
Literature search: Material relevant to the subject: Journals, books, government documents
Data evaluation: Literature that makes a significant contribution; it is important to select topical literature.
Moreover, it is preferred that literature is as recent as possible
Analysis and interpretation: Discuss findings and conclusions

Tips on conducting a literature Review

Use the librarian to get the database
Consult your supervisor regularly
Set deadlines and meet them
Give yourself time to read the material
Check the strength and weaknesses of the information collected
Keep full and accurate records
Be systematic in note taking under headings

c. Information sources in a literature review

There are many sources of information for a literature review. They include:

Text books: Most of these are available for commercial purposes and are governed by International Copy-
right law. You need to quote information from the most recent editions of these textbooks

Journals: many of the highly regarded journals have a lot of information and are a rich source of literature
citations. They also often have up-to date evidence based information on a particular topic. Unfortunately,
many of these journals and indeed their individual articles are only available on commercial terms and
accessing current articles is very expensive.

The Internet: The internet is an inexhaustible source of ready information and is one of the commonest
sources on citations. However, because of the vast amounts of information, one has to know how to
conduct a meaningful search, otherwise, a lot of the information posted on various sites in of low scholarly
value.

Other Sources: These include dissertations, compendia, annals, encyclopedias, and various data bases.
They also include several reports from health related agencies, ministries and other government
departments.

Open Course-ware: You ought to be aware that despite the fact that many quality journals are expensive
and are for commercial purposes, there are many open courseware sites from which you can obtain current
information for free, provided the sources are acknowledged. There are over 5000 health related journals
available for free and some on line.

27
d. Organization of Literature review

Introduction: Gives the central theme and organizational pattern

The body: It has the main discussion sources organized. It may be done:
 Chronologically by publication trends
 Thematically around a topic of issue
 Methodologically where the focus is on the methodology that the researcher is pursuing
 Objectives: We can organize our literature based on the anticipated objectives
 Conceptual Framework: We can use a thematic approach but based on key constructs in our
conceptual framework
Conclusions and recommendations: We may provide a summary of the review in relation to our
research problem

Structured Literature Review: The best method for presenting literature is by structuring it in a logical
flow, with subtitles. Omnibus literature does not bring out the issues clearly and is boring.

Writing up the review: We ought to ensure that we do not copy and paste everything as it appears in an
article. There are so many efforts world-wide that have been initiated to fight plagiarism. There is a limit to
the amount of words that can be copied verbatim from a cited document. We ought to interpret the
information presented and state it in our own way. Literature review means critiquing literature, not
transferring literature.

e. Literature Citation and Referencing

Citation of Literature
It is important that for all reviewed literature, citation is made. This enables us to avoid plagiarism – all
source of information must be acknowledged.

Methods of citation: There are two methods:

1. The Author-Date method: It is also called the ‘Harvard Referencing System’. In this system, we write
the author’s last name (surname) and year of publication. The entire citation can be bracketed in the middle
of the sentence (Mukasa, 2004) or at the end of the phrase or sentence. If the author is used to start the
sentence, we only bracket the year. E.g. Mukasa (2002) notes that…… If we have two authors, we cite
both e.g. Mukasa and Kakuyu (2004). If we have more than two authors, we indicate only the first or lead
author, followed by ‘et al’ e.g. Mukasa et al. This method of referencing is the one commonly used in your
field reports at MUSPH.

2. The Notation or numerical Method: This is also called the ‘Vancouver Referencing System’. We use
superscripts inserted in the literature. This method makes it less easy to refer to the authors and users are
increasingly complementing the reference listing with foot note insertions on every page that has a
reference. This method is preferred in journals.

Please Refer to your Additional Resources Folder for detailed hand-outs that describe the two
systems of referencing

Referencing
We have to insert a sheet that gives a detailed description of all the references we have cited in the text.

28
We quote different types of references as follows:

Journals: Author(s), year of publication, title of article, journal serial number, volume number, issue
number and page numbers

Books: Author/Editor, year of publication, title, edition, place of publication, number of pages in the book

Electronic Information: Author/Editor, year of publication, article title, type of medium e.g. CD-ROM,
pages or length of or website plus the date when accessed

Dissertation: Name of author, year, title, indicate that it is a dissertation, town, university/Institution

Reference List: We may use alphabetical order for the author-date method or a numerical scheme for the
superscript method

References
Either the Boston (Author, Date method) or Vancouver method (citation – numbers are used within the text
with a full reference at the end of the document) can be used in referencing – stick to one type throughout
the write-up.

Example of the Boston method:

Wawer, Maria et al. 1997. Trends in HIV-1 prevalence may not reflect trends in incidence of mature
epidemics: Data from the Rakai population-based cohort, Uganda. AIDS 11: 1,023-1,030.

Gold, Marthe R. et al. (eds). 1996. Cost Effectiveness in Health and Medicine. New York: Oxford University
Press.

Example of Vancouver method:

“… they are positive benefits in added life and improved health, but the only record ordinarily kept in
morbidity and mortality statistics is the partial and negative record of death and of illness from certain
clearly defined types of disease, chiefly the more acute communicable diseases, which constitute only a
fraction of the total morbidity.1”

At the end of the document the full reference is made:

1. Frost WH: Rendering account in public health. Am J Public Health 15:394-398, 1925.

f. Plagiarism

A literature review in essence means that you read up available literature and formulate your own opinion
and description of your subtle findings. In doing this, you should not transfer directly what is written, but
instead, you should interpret it and write it in your own way. Whenever you obtain a piece of information
from a reference site, international copy-right law requires that you acknowledge the source, i.e. the
authors and the publishers. This is the basis for literature citation. In addition, international copy-right law
requires that in case you copy material from a source, you do not exceed 10% of the material from that
source.

29
Plagiarism occurs when you take credit for information that is not your own including failure to acknowledge
the sources. You are strongly warned against plagiarizing information directly from sites and failure to
acknowledge sources.

2.1.7 Extension Activities

Extension Activity 1: Discussion Forum Question

Extension Activity 2: Self- Assessment Quiz

30
 2.2 Unit 2: CONCEPTS UNDERLYING RESEARCH
II: METHODOLOGIES IN RESEARCH
2.2.1 Introduction to the Unit
In this unit, we shall build on the concepts in Unit 1 and continue looking at the principles underlying
research methods, this time focusing on methodological aspects to consider in preparation for a research.

2.2.2 Unit Outline

The following topics will be covered:
Session 1: Study Setting, Study Population and Study Designs
Session 2: Sample Size Determination
Session 3: Sampling Procedures
Session 4: Selection of Study participants
Session 5: Quantitative and Qualitative Data Collection Methods
Session 6: Quantitative and Qualitative Data Collection Tools
Session 7: Study Variables and Measurements
Session 8: Data Management and Quality Control in Research
Session 9: Ethical Considerations in Development of a Research Problem
Session 10: Workplan and Budgeting
Session 11: Alternative study designs

2.2.3 Instructional goal

The MPHO should be able to select, develop and use appropriate methodologies in conducting research in
a given research problem

2.2.4 Unit Objectives

By the end of this unit, the student should be able to:

1. Select and describe appropriate study sites and study populations in response to a given
research problem
2. Select appropriate study participants for a given research problem in a way that minimises
random and systematic errors
3. Select appropriate methodological procedures for data collection involving measurement of
variables for a given research question
4. Select and use appropriate alternative study designs where the research problem requires
alternative approaches
5. Set up appropriate mechanisms for ensuring proper quality control and data management in
research

2.2.5 Time Frame

2 WEEKS

2.2.6 Content

31
Session 1: The Study Setting, Study Population and Study Designs
Introduction: In preparation for conducting the study, the first methodological consideration is in describing
the study setting and the population to be studied. We have to describe certain specificities of this
population, so as to get a basis for deciding on the study design. In this session, we shall discuss key
aspects in describing the study setting, the study population and the design of a study.

Session Topics: The following topics will be covered:

a. The Study Setting
b. The Study Population
c. Overview of Study Designs

Session Objectives:
By the end of this lesson, the MPHO should be able to:
1. Describe the study setting for a given research problem
2. Describe the target population and accessible population for a given research design
3. Select appropriate study designs for given research problems

a. The Study Setting

The study setting refers to the context, characteristics and situation of the site in which the study will be
conducted. In some protocols, they only describe the ‘study site’ which is in essence part of the study
setting. It is important to give a description of the site and setting accordingly
The scope and Detail varies according to site
 It may be as small as a clinic
 It may be an institution (hospital with several clinics)
 It may be a local government or administrative area
Describe the characteristics of the setting
Describe the peculiarities of the setting
Example: For a study done in a district
 Geographical location, Population, Area, Economics
 BOD related to the study
 Key indicators (IMR, income, sex ratio, nutrition)
 Infrastructure for service delivery
Example: For a study done in a clinic setting Talk about
 The patient profile and flow
 Referral patterns
 Must be relevant to the general theme of the study

b. The Study Population

Population: A Universal/Complete set of individuals with a specified set of characteristics

Sample: Subset of a population
Characteristics depend of the study questions and objective
 It may be animate: People
 It may be inanimate: Health Centers

32
  It may be general (entire population)
 It may be specific (e.g. women of child-bearing are)

There may be different populations for different objectives

Defining characteristics may be: Clinical, Demographic, Temporal, Physical, Geographical etc

Study Populations have different levels of accessibility

The Target Population: It describes the large set of a sub-population of people throughout the world or the
geographic area to which the results will be generalized
 It is defined by clinical and demographic characteristics
 It consists of the Actual Population and the Accessible Population

The Actual Population: It is the entire set of the sub-population that we are interested in
E.g. All pregnant women with HIV

What’s an accessible population? The subset of the target population available for study or those that we
can actually reach for this study
E.g. Pregnant women who are HIV negative attending Kasangati HC in Nov 2005

c. Overview of Study Designs

After describing the study setting and the study population, it is logical to describe the study design that will
be used to access your target population, in particular, the accessible population. One of the challenging
situations for the MPH Officer is to select an appropriate study design. In summary, the epidemiological
study designs can be recapped as follows:

a. The Observational Designs: Here, the researcher does not manipulate the situation, but they simply
observe and make conclusions on different relationships they are observing. There are two broad
types:
a. The Descriptive Studies: Descriptive studies often investigate the suspected exposure and
outcome at the same point in time and usually generate univariate information that is the basis
for developing hypotheses. They are usually of several categories:
i. Case Studies: Usually follow up one case
ii. Case Series: Usually follow up a set of cases with interesting parameters
iii. Cross-sectional Studies: They usually study a set of parameters at the same point in
time and establish the prevalence of these parameters in the populations. Sometimes,
cross-sectional studies have been used for comparison purposes between suspected
exposure factors and disease and in essence analysed using analytical techniques.
However, the information generated should be interpreted in the context of
establishing hypotheses about possible associations and not causality. The analysis
may be within one sample if a prevalence formula is used, or it may be in two or more
comparison groups (a cross-sectional comparative study) if a formula for comparison
groups is used.
iv. Ecological Studies: These are variants of cross-sectional studies applied to
populations. They are also called ‘correlational studies’ in that they only establish
relationships between factors but may not readily establish the dependent and
independent variables because there is no temporal relationship.

33
 b. The Analytical Studies: Analytical studies involve the establishment of a temporal relation
ship between the exposure and the outcome. They can therefore be used in inferring causality,
because we know that the exposure factors preceded the outcome of interest. However, they
are observational in that the exposure is not determined by the researcher, but by
circumstances, the researcher being a mere observer. They are broadly of two types:
i. Cohort studies: Cohort studies involve identification of an exposed group and an un-
exposed group and following them forward in time to establish the association
between the exposure and disease
ii. Case-Control studies: In case control studies, we start with the people with the
outcome of interest (cases) and people without the outcomes (controls) and we
investigate their past to establish the association between past exposure factors with
their current condition.

b. The Experimental Designs: In experimental studies, the researcher applies an exposure to a study
sample and follows them forward in time to establish the association between the exposure and the
disease. They are similar to cohorts, but the difference is that here, it is the researcher who applies the
exposure and may even manipulate it. As a result, they cannot be used to investigate putatively
harmful exposures. There are broadly two types:
a. Clinical Trials: The exposure is applied to individuals in a clinical or laboratory setting
b. Field Trials: The exposure is applied to communities or sub-populations as the unit of
application.

Refer to your notes in Applied Epidemiology I for a detailed revision of the different Epidemiological
Study Designs

34
Session 2: Sample Size Determination

Introduction: What is a sample? It is a sub set of the study population. Many times it is necessary to study
a sample rather than the whole population because of the following reasons:
1. Feasibility: We cannot identify all population units
2. Cost: Implications in terms of money, personnel and other resources
3. Theory: We know from theory that a well taken representative sample is sufficient to give the
required answers

Researchers are often concerned about the representativeness of the sample. This affects the external
validity of the research findings. There are two major considerations:
1. How many should be selected? (Sample Size)
2. How should they be selected? (Sampling Procedure)
We need to take into consideration:
- The design of the study
- Units of measurement of Primary Objective variable (Continuous or categorical/proportions)
- Estimated difference between study groups that is worth noting (for studies that are comparative)

Session Topics: The following topics will be covered:

a. Prevalence Studies
b. Comparative studies
c. Cluster Surveys
d. Sample Size in Studies of Sensitivity and Specificity
e. Alternative ways of determining the sample size

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Compute appropriate sample sizes for different types of study designs
2. Employ alternative ways of determining sample sizes where appropriate

REFERENCES:
1. Bernard Rosner (1990): Fundamentals of Biostatistics (Harvard University Press – 3rd Edition
2. Jones SR, Carley S and Harrison M (2003): An introduction to power and sample size estimation.
Emmerg. Med J 20(5): 453-458.

Refer to your notes in Applied Biostatistics I and revise the rationale and derivation of the most
common formula for sample size determination for crossectional studies.

Determination of sample size depends on the study design and the nature of the main outcome variable we
want to measure.

a. Prevalence Studies

Considerations include:

35
 - Anticipated Prevalence of the problem under study: This is often based on prior studies or on a
pilot study. We can also use an arbitrary consideration of 50%
- Precision: It is the difference between estimated and actual prevalence that the researcher is
willing to tolerate. It is denoted as  .

If the outcome of interest is a continuous variable:

The Sample Size is given by:

 Z 2 2 
n 2 
  
Where:

Z = The standard normal deviate at 95% confidence (1.96)

 2
= Variance of the outcome of interest
2 = Maximum error between the estimated and actual population mean the investigator is willing to
allow. This is also known as the precision

NB: The sample Size “n” is assumed to be selected using a random mechanism.

If the outcome of interest is a categorical variable, reported as a proportion

This formula was proposed by Leslie Kish in a 1964 paper. It was also reviewed by Daniel (1999).

Leslie Kish (July 27, 1910 – October 7, 2000) was an American statistician of Hungarian descent. He
fought in the International Brigade in the Spanish Civil War.

His main work is the book Survey Sampling (1965), where he summarized and developed the current
theory of sampling and survey design. He was also a noted teacher at the University of Michigan and
on regular visits to the London School of Economics.

A common mistake by students (especially as a result of ‘plagiarizing’ the sample size formula
is to refer to ‘Kish Leslie’ as ‘Kish and Leslie’ – Kish and Leslie is one person!

For these types of studies, the Sample Size is given by:

 Z 2 PQ 
n 2 
  

Where:

Z = The standard normal deviate at 95% confidence (1.96)

P = Estimated prevalence of the problem under study
Q = 100% - P (or 1-P)
2 = The precision or maximum acceptable error the investigator is willing to accommodate

36
NB: Some books or guides suggest that if it is not possible to come up with a good estimate for P, one may
set P equal to 0.5 to yield the maximum sample size (Daniel, 1999, Lwanga and Lemeshow, 1991).
However, some anecdotes argue that this should be done with caution.

If the population is considered to be “finite” (Number limited and known)

The calculations so far are valid if the calculated Sample Size is smaller that or equal to 5% of the
population under study ( n / N  0.05 ) (Daniel, 1999). If the proportion of Sample to Population is larger
than 5%, we need to use the formula by Daniel (1999) as follows:

NZ 2 P(1  P)
n 
 2 ( N  1)  Z 2 P(1  P)

Where:
n = Sample Size with a finite population
N = Population Size (Estimated Population from which the sample is drawn)

Other Anecdotes show that we can adjust the calculated Sample size directly using the formula:

 n 
Adjusted Sample size is: n   
1  n 
 N

Where: N = Estimated population from which the sample is drawn

n = The previously calculated sample Size
n = Adjusted sample size

b. Comparative studies

Considerations include:
- Anticipated variance if the outcome is continuous
- Anticipated prevalence if the outcome is categorical
- Error the investigator is willing to accommodate (or precision)
- Desired level of significance

1. Formula for case-Control Studies

If it is a comparative study, with two groups and the outcome of interest is categorical, then the sample size
is given by:

 2  R 1 
 ( Z  Z (1 ) )  R  p(1  p) 
   
 ( P 0  P1 ) 2

 

Where:
37
Z = The standard normal deviate at 95% Confidence
Z (1  ) = The Z-value corresponding to a power of 80%
R = Estimated Odds Ratio of the outcome in exposed vs. the un-exposed
P0 =Estimated probability (prevalence) of the outcome amongst the un-exposed
P1 = Estimated probability (prevalence) of the outcome amongst the exposed
P  RP0
p = Weighted average of the two proportions, given as: p  1
1 R

2. The Modified Schlesselman Formula for Case Control Studies

According to Schlesselman (1982), the sample size for Case-Control Studies depends on 4 Parameters:

1. The relative frequency of exposure among controls in the target population, P0

2. Estimated Relative Risk (or Odds Ratio) associated with exposure
3. The desired level of significance, 
4. The desired study power, 1-

The formula for a comparative study with equal number of controls and cases is:
 2 pq ( Z a  Z1  
n  
 ( P1  P0 ) 2 

Where:

1
p = ( P1  P0 )
2

q = 1 p

P0 R
P1 = (the relative frequency of exposure among cases).
[1  P0 ( R  1)]

P0 = Estimated exposure rate (proportion exposed) among controls.

R = Hypothesized relative risk (estimated by the OR) associated with exposure

Z = Standard Normal Deviate at 95% Confidence (1.96)

Z (1  ) = Standard Normal Deviate at 80% Power (0.84)

3. Formula for follow-up comparative studies (Cohort Studies)

Again we have to consider if the main outcome of interest (or measure of effect) is continuous or
categorical. If outcome of interest is continuous: the formula is given by:
38
 2(Z  Z1 ) 2  2 
 
 2 

Where:

Z = The Standard normal deviate

 = The level of significance (Usually 5%)
1  = The Power of the study (Usually fixed at 80%)
2 = Variance of the outcome of interest
 = Anticipated effect difference/effect size worth detecting

NB: The sample size obtained is per each study group

If the outcome of interest is categorical and is reported as a proportion, the formula is the same as above:

 2(Z  Z1 ) 2  2 
 
 2 

However,

 2  {P0 (1  P0 )  P1 (1  P1 )

And:

  (P0  P1 )

Where:

P0  Anticipated proportion (or incidence) of the outcome in the control group

P1  Anticipated proportion (or incidence) of the outcome in the exposed group

The formula is therefore:

 2(Z  Z1 ) 2 {P0 (1  P0 )  P1 (1  P1 )}

 
 ( P0  P1 ) 2 

c. Cluster Surveys

The commonest Formula used is the one by Bennet. It determines the number of clusters required in a
cluster survey. The overall Sample Size is determined using the formula for crossectional studies.
However, determination of the number of clusters is determined based on the formula:

39
 P(1  P) D
C
S 2b

Where:

C = Number of clusters needed

P = Estimated Prevalence of outcome of interest
D = Design Effect
S = Level Precision
b = Estimated number of respondents per cluster

The Design effect D is given by: 1  b  1 . As a result of sampling at two levels, we introduce a design
effect likely to affect the randomness of our sampling. The formula therefore attempts to compensate for
the effect.

NB: “We would expect that that in all cluster surveys b should be known a priori, since it is decided and
selected on the basis of convenience (Kumar and Indrayan, 2002)”.

Reference: Kumar R and Indrayan A, (2002): A nomogram for single-stage cluster-sample surveys
for estimation of prevalence: International Journal of Epidemiology, 2002, 31: 463-467

d. Sample Size in Studies of Sensitivity and Specificity

There are several formulae for Sample Size determination in studies of sensitivity and Specificity.

1. In the absence of a Gold Standard

One formula for these studies is as follows:

 4Z 2 ( P(1  P) 
n   1 / 2 
2 
 (l (S  C  1) 

Where:
l = Level of significance
Z = Standard normal deviate at 95% Confidence (1.96)
S = Estimated Sensitivity
C = Estimated Specificity
P = S  (1   )(1  C )

2. When there is a gold standard

Sample size needed for adequate sensitivity

TP  FN
NS 
P

40
  SN (1  SN 
But: TP  FN  Z 2 X  )
 W 
2

Where:

TP = True Positives
FN = False Negatives
Z = Standard normal deviate at a given level of significance
SN = Lowest sensitivity acceptable
W = The confidence interval
P = Presumed disease prevalence in the study population

e. Alternative ways of determining the sample size

There are other alternative ways of determining the sample size that you should explore on your own. They
include:

1. Use of statistical tables of Sample Size:

These tables require that you know the different parameters for in-put into the sample size
determination. These parameters are then used to check the statistical tables for corresponding sample
sizes. Some of the parameters include:
a. The Desired Precision
b. The desired power
c. The estimated effect size worth noting
d. The desired level of confidence.

The tables enable us to determine different sample sizes at different levels of precision and power,
after which we can choose the most appropriate one for our research context.

2. The WHO Method

The WHO has developed a method for cluster surveys based on the power calculations at different
levels of precision. This methods was primarily used for immunization coverage surveys but has now
found popular use in other cluster surveys for Public Health Action, including nutritional and dietary
surveys as well as other crossectional cluster surveys.

Based on power calculations, it has been determined that a good sample size for these surveys is
derived by selecting 30 clusters of 7 respondents each. Some modifications have used 30 clusters of
12 respondents.

3. Software
Several statistical packages can be used for sample size determination e.g. stat-calc in Epi-Info. These
softwares require that you know the desired input parameters before an estimate can be made.

41
Session 3: Sampling Procedures

Introduction: One of the major purposes of doing research is to infer or generalize from a sample to a
larger population. The population refers to a large set or collection of items with something in common. A
sample is a sub-set of a population. After determining the appropriate sample size, we have to select the
study subjects using a sampling strategy. The sampling strategy must be designed in a way that does not
introduce bias in itself. In this Session, we shall examine the different sampling procedures available in
epidemiological research.

Session Topics: The following topics will be covered:

a. Rationale and Reasons for Sampling
b. Probability Sampling Methods
c. Non-probability sampling
d. Bias in Sampling

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Explain the rationale and justify the reasons for sampling in research
2. Select and apply appropriate probability methods for sampling in a given research design
3. Select and apply appropriate non-probability methods for sampling in a given research
design
4. Describe the different biases that are likely to occur in sampling and how they can be
minimised.

a. Rationale and Reasons for Sampling

What is sampling? It involves the selection of a number of study units from a defined study population. A
good sample needs to be representative i.e. it should have all the important characteristics of the
population from which it is drawn.

Schematic:

Problem

Study Population

Study Unit

Sample Units

42
Advantages and reasons for sampling
1. A sample can be studies more quickly than a population
2. It is less expensive
3. A census (study of the entire population) is impossible in most situations
4. Usually, sample results are more accurate
5. In sample survey, more time and resources are spent in training personnel for a smaller scope of
work
6. If properly selected, probability methods can be used to estimate the error
7. A sample can be selected to reduce heterogeneity; a sample of the population with more
representative characteristics is more appropriate than the entire population in studying certain
aspects of a disease or health related conditions e.g. a study among a high-risk sub-population

Bigger things are not always better!

Preparation for Sampling: Before applying a sampling procedure, the investigator must determine the
sample size appropriate for the study design selected. This determined the power of the study.

Methods of sampling: Methods of sampling are either:

1. Probability methods
2. Non probability methods

b. Probability Sampling Methods

Probability methods involve random selection procedures to ensure that each unit of the sample is chosen
on the basis of chance. All units of the study population should have an equal or at least a known chance of
being included in the sample. The best way to ensure that a sample will lead to reliable and validly
representative inferences is to use probability samples i.e. the probability of being in the sample is known
for each subject in the population. 4 common probability methods are in use in biostatistics:
(a) Simple Random Sampling
(b) Systematic Sampling
(c) Stratified Sampling
(d) Cluster Sampling
(e) Multi-stage sampling

1. Simple Random Sampling: Every subject has an equal opportunity of being selected. We may use a
table of random numbers or a computer generated list of random numbers. Each element is accorded a
serial number and the entire series is called a ‘sampling frame’. We then proceed to sample. Sampling
can be ‘with replacement’ or ‘without replacement’. Sometimes, when we are in the field, we may not
use random numbers, but other field techniques that give a semblance of randomness.

Advantages: It is simple

Disadvantages: Tedious for large samples; often times, we are not able to get a listing of the entire
study population so as to make a sampling frame

2. Systematic Sampling: In this method, every item is selected. K is derived by dividing the number of
items in the frame by the desired sample size; e.g.

3400/200 Radiographs = 17; therefore, every 17th X-Ray

43
 However, in order for it to be a probability sample, we must select the beginning number in the
sampling frame by random using a table of random numbers or other probability technique.

Advantages: Less time consuming and easier to perform

Disadvantages: There is a risk of bias as the interval may correspond with a systematic variation in
the sampling frame

3. Stratified sampling: The population is first divided into relevant strata (sub-groups based on given
attributes) and a random sample is selected from each stratum. By strata, we mean categorizations
that are homogenous for a given attribute e.g. sex (male/female; Married/not married). Strata ought to
be mutually exclusive (non-overlapping). Sampling from strata is associated with a ‘design effect’ and
therefore, the sample size calculation has to accommodate this.

Advantages: Allows a relatively larger sample from a smaller population. It also allows equitable
representation of important strata

Disadvantages: There is a potential for unequal sampling fractions; it is important to ‘adjust’ for it when
generalizing findings.

4. Cluster sampling: Population is divided into clusters and a random sample is taken from each cluster.
Clusters are mainly based on spatial attributes e.g. districts, sub-counties, villages and households.
The elements of a cluster however are not homogenous.

Advantages: Allows a relatively larger sample from a smaller population. It also allows equitable
representation of important strata

Disadvantages: There is a potential for unequal sampling fractions; it is important to ‘adjust’ for it when
generalizing findings.

5. Multistage sampling: This is a combination of methods. It involves sampling at different levels. The
sampling units are first the bigger units, either strata or clusters. Consequently, we then sample from a
population of sub-clusters or sub-strata within the bigger units. There after, we may sample individuals
from the sub-clusters. The different levels of sampling are referred to as stages. Multistage sampling
usually involves a mix of different sampling methods at different levels. It may be:
a. Multi-stage cluster sampling
b. Multi-stage stratified sampling

Advantages: It allows a very large and diverse population to be studied. A sampling frame is not
required for the entire population – we only need a frame for the clusters or strata at the different
stages. The sample is easier to select than a simple random sample of similar size.

Disadvantage: There is a larger probability of non-representativeness. We therefore have to adjust for

the ‘design effect’ and this increases the sample size.

44
Other probability techniques:

Random Assignment: In this approach, subjects are first selected fro inclusion based on certain criteria.
They are then assigned to intervention or control groups randomly. Random numbers are used. These
designs may be single or double blind. We can see here that the initial selection is not random – we select
people with particular attributes purposively.

Probability proportionate to size: This is applied in stratified sampling. We divide up the sample size and
allocate it to the different strata based on the proportionate size of the stratum in the general population.

c. Non-probability sampling

In such methods we select a sample purposively. The probability that a person is selected is not known.
Sometimes, we sample based on a purpose and sometimes not. These methods find wide application in
qualitative methods, but can also be applied in quantitative methods when generalisability is not at stake.
They however often reflect biases (selection biases); they do not fulfill the requirements of randomness
needed to estimate sampling errors. They include:
1. Convenience samples
2. Selection on purpose (purposive sampling)
3. Quota Samples:

1. Convenience samples: We select a sample depending on ease of accessibility. The method is for
convenience’s sake i.e. the study units that happen to be available at the time of data collection.

Advantage: Simplicity and resources

Disadvantages: May not be representative of the population, hence affecting generalisability.

2. Selection on purpose (purposive sampling) : We select a sample based on the purpose for which
we want information

Advantage: It may increase internal validity of findings and allows for exploratory depth. Information
may be useful for formation of hypotheses.

Disadvantage: It is open to bias (surveillance, selection and information biases) and findings may not
be generalisable.

3. Quota Samples: We allocate a quota to the different strata or clusters. The method ensures that a
certain number of sample units from different categories with specific characteristics appear in the
sample, so that all the categories are equitably represented.

Advantage: Useful when probability samples are unlikely to provide the desired balance of study units

Disadvantage: Non-representation and lack of generalisability.

Although these methods have less external validity, is done correctly, and tagged to answer specific
problems, they may have a high internal validity. They are more robust when the main issue is exploratory
depth of the experiences of a few individuals known potentially to harbour a ‘wealth of experience’.

45
d. Bias in Sampling

Definition: Bias in sampling is a systematic error in sampling procedures that may lead to a distortion in
results of the study. It includes:

(1) Non-response: Some sampled respondents fail to respond

(2) Self-selection bias (studying volunteers only or mainly). These may behave differently because of their
inherent motivation
(3) Sampling registered patients only: They are likely to differ systematically from the population e.g. some
referral hospitals and units
(4) Missing Cases of short duration: In studies involving short duration illness, some cases may be missed
(5) Seasonal Bias: Seasonal variations in the variable under study with regard to incidence and prevalence
(6) ‘Tarmac bias’: Selection of study areas and participants due to their easy accessibility along main
roads

Addressing Biases in Sampling

1) Data collection tools need to be simple

2) Data collection tools need to be tested
3) Non-response due to absence – need for follow-up
4) Non response due to refusal to co-operate – an extra study for non respondents may be
considered; full information may facilitate consent. We can replace the non-responders
5) Anticipating the problem and solving it during the process or avoiding it

46
Session 4: Selection of Study participants
Introduction: The selection of study participants is an important aspect of research and if not done
correctly, has the potential to bias the study and introduce confounding. In this Session, we shall examine
the different considerations in selecting study participants and in particular, we shall contrast between
inclusion and exclusion criteria and see their justification in research.

Session Topics: The following topics will be covered:

a. Goals in recruitment of study participants
b. Selection Criteria
c. Special Considerations in selection
d. Selection Biases

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Apply appropriate inclusion and exclusion criteria in selecting study participants for a given
research design
2. Apply specific selection criteria in Observational and experimental studies, mindful of the
situation specific considerations
3. Describe the different selection biases likely to occur in research and how they can be
minimised

a. Goals in recruitment of study participants

 Ensure that subjects are representative of the population

 Specify a sample size of subjects that can be studied
 At an acceptable cost in time and money
 Large enough to control random error
 Representative enough to control systematic error
 Select from the target and accessible population
 Begins in the design phase with choosing populations and sampling methods wisely
 Ends with implementation
 Guard against errors in applying entry criteria
 Monitor adherence to these criteria as the study progresses

Concerns in selection of study participants

- Non-response in observational studies
- Minimizing Refusal to Participate
- Improving efficiency and attractiveness of the study
- Invasive and uncomfortable tests
- Allaying anxiety and discomfort
- Providing incentives e.g. reimbursing costs of transportation
- Providing the results of tests
- Language barriers use

47
b. Selection Criteria

One of the means of increasing the precision of study designs is in using selection criteria for participants.
These are of two types:
- Inclusion criteria
- Exclusion criteria

Exclusion Criteria
Define the conditions that should not be included in an intended study
Normally based on potential to confound findings, accessibility or ethics
The criteria vary depending on study objectives
The following are typical;
 Concomitant disease conditions that may affect our observation of interest
 Treatments that interfere with evaluation of new treatment
 Signs/symptoms/test results leading to uncertainty about the diagnosis
 Safety and ethical reasons e.g.
o Patients with life threatening disease
o Patients with disease that has known treatment
o Pregnant and breast feeding women if the therapeutic agent may affect them
o People with known allergies
o People with a condition that may interfere with the metabolism of the drug under test
o People with prior exposure to the drug under study
 Patient characteristics indicating that:
o Such patients are likely not to comply
o May not be traced (no address)
o May need interpreters
o May not understand instructions
o Severe alcoholics and drug abusers
 Other criteria to minimize confounding

Inclusion Criteria
Conditions one must meet in order to be considered for participation
Idea is to have a ‘homogenous non-contaminated, non-confounded population’
Example: When we are conducting a study to determine the correct effect of a new treatment, we may
certain consider some medical standards
e.g.: susceptibility or risk of infection, non-presence of antibodies etc

NB: Inclusion Criteria are not the opposite of exclusion criteria. Inclusion criteria refer to standards that cut
across all study subjects, while exclusion criteria refer to possible confounding situations that have been left
out of the study.

c. Special Considerations in selection

The different epidemiological study designs have different considerations in selection of participants and
different issues specific to the particular design. They include:

Case-control studies
 Selection of cases
 Selection of controls

48
  Marching
Cohort studies
 Selection of exposed populations
 Selection of the non-exposed
Experimental studies
 Selection of subjects
 Allocation of subjects to intervention groups and randomization
 Crossovers that are planned or un-planned

Please refer to your notes in Applied Epidemiology I for a more detailed revision of the different
issues in selection of participants for the different types of study designs.

d. Selection Biases
 Selection may introduce confounding
 Self selection bias
 Referral bias (Berkson’s Bias)
 Diagnostic Bias
 Misclassification biases
 Differential and non-differential misclassification

Refer to your notes in Applied Epidemiology I for a revision of the different types of selection
biases

49
Session 5: Quantitative and Qualitative Data Collection Methods
Introduction: Do we use Qualitative or Quantitative Data collection methods? The decision depends on
the kind of study questions we are trying to answer. For studies in which we want to generalise findings to
an entire population, we often take the same measurements from all respondents in a sample – we derive
statistical parameters and may test hypotheses about these associations. These are Quantitative methods.
On the other hand, if we are interested in exploratory depth and understanding all dimensions of a
phenomenon, then we use qualitative methods. In many studies you will be conducting, however, both
types of methods may be used and as a result, they may complement each other. In this session, we shall
examine the different Quantitative and Qualitative methods.

Session Topics: The following topics will be covered:

a. Characteristics of Quantitative methods
b. Characteristics of Qualitative Research
c. Comparison of Qualitative and Quantitative Research
d. Quantitative Methods of Data Collection
e. Qualitative Methods of data collection

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Differentiate qualitative methods from quantitative methods
2. Appraise different research questions to justify the use of quantitative and quantitative
methods
3. Describe and contrast the different qualitative and quantitative methods of data collection

Definitions
Data: A collection of items of information
Quantitative Data: Data in numerical form such as counts and measurements
Quantitative Data collection Tools: Questionnaires and Checklists for Quantitative Data Collection
Qualitative Data: Data in descriptive format, not in form of numbers

a. Characteristics of Quantitative methods

 They are based on measures of quantity/frequency

 Their findings are described in numbers rather than words
 A pre-determined series of questions
 Questions are asked in the exact same way to every respondent
 This approach has long dominated social investigations
 Mainly for National planning and program evaluation
 Surveys attempt to measure social opinions and behaviors (KAP Surveys).
 They are attractive
 They allow statistical inference from relatively small sample to large population

When to use quantitative methods

This approach should be used:

 When the subject matter is clearly defined and familiar

50
  When measurement problems are minor or have been resolved.
 When there is need to relate findings to the broader socio- cultural settings that are already
sufficiently understood.
 When detailed numerical description for representative sample is required
 When repeatability of measurements is important
 When generalization of results and comparison across populations is desired

When not to use quantitative approaches

 When exploratory depth is required
 When there is no baseline information available of factors
 When detailed linkages between factors are required
 When detailed individual expertise and experiences are required
 When the interest is to understand all possible variations of a given phenomenon

b. Characteristics of Qualitative Research

Qualitative Research embodies four approaches:

Explorative flexibility: Deep exploration of a few experiences
Iteration: Changing tools with experiences
Triangulation: Collation of findings with other experiences
Contextualization: The context is part of the study

Qualitative Research useful for:

 Exploring health issues of which little is known
 Identifying local perceptions of priorities
 Identifying relevant interventions and target populations
 Investigating feasibility, acceptability and appropriateness of new health programs
 Developing appropriate IEC activities and materials
 Identifying problems and solutions for on-going interventions
 Complementing Quantitative data in M&E studies
 Designing more valid survey instruments; modifying questions in survey tools

When to use qualitative approaches

 When subject matter is unfamiliar
 For exploratory research: when relevant concepts and variables are unknown or their definitions
are unclear
 For explanatory depth: when you want to relate aspects of behavior to a wider context
 When a meaning rather than frequencies is sought
 When flexibility of approach is needed to allow for discovery of the unexpected
 For studying selected issues, cases or events in depth & detail

c. Comparison of Qualitative and Quantitative Research

Characteristic Qualitative Research Quantitative Research

PURPOSE Describe; Understand Predict causal relationships
LOGIC Why? How? Asked What? Tested
ROLE OF THEORY Inductive Deductive
INSTRUMENTATION Broad theoretical models; post Theory driven hypotheses of
51
 hoc development of relationships among variables
relationships
SAMPLING OF SUBJECTS Availability; purposive; Exclusion/inclusion criteria
theoretical; snow balling established in advance;
random assignment;
representative control group
Characteristic Qualitative Research Quantitative Research
INTERVENTION Observational only; Carefully designed and
documenting effect of manipulated by researcher
presence in setting
DATA COLLECTION MEANS Immersion; trained Structured interview
observation; in depth instruments with predetermined
interviewing; open or semi response categories
structured tools
DATA ANALYSIS Verbally descriptive and Numerical descriptions and
usually non-numerical statistical
ROLE OF RESEARCHER Must establish relationship with Must be blinded to conditions;
subjects standard interaction with
subjects

Selection of the best approach to use

How do we select which approach to use? This depends on:

 The type of research questions
 The depth of data you want to collect on each individual variable
 Whether your variables are defined or not
 Generalisability Vs. Depth of experiences
 The form you want the data to take (Numerical or textual)
 Your data analysis plan
 Your research plan (Amount of staff and time to conduct the research)

In most instances however, we use a blend of both methods so that we can triangulate the findings and
also gain deeper understanding of the phenomena we are trying to study, especially in cross-sectional
studies.

d. Quantitative Methods of Data Collection

They include:
1. Face to Face interviews
2. Mail/Postal Questionnaire
3. Telephone interview
4. Self administered questionnaires
5. Unobtrusive methods:
a. Documents review
b. Observation

52
1. Face-to-face Interviews

- It is the commonest method

- They are interviewer administered

Advantages
 Interviewers can probe fully for responses
 Interviewers can clarity ambiguities
 Inconsistencies can be checked
 More complicated and detailed questions can be asked
 No literacy requirements for respondents
 Questions can be asked in predetermined manner
 Response rates are usually higher
 The allow a combination with open ended question
 Allows for examination of body language

Disadvantages
 Expensive and time consuming
 Interviewer bias
 It is difficult to ask embarrassing questions

2. Postal Questionnaires and Self Administration

- A Questionnaire is sent to an individual

- The individual fills it by themselves
- The individual then sends the questionnaire back
- It may be hand delivered or by post
- It may be supervised or not supervised if hand delivered

Advantages
 Eliminates interview bias
 Useful for sensitive topics

Disadvantages
 Questions need to be straight forward and simple
 Study population needs to be 100% literate
 Response rates for questionnaires may be low
 Consultations and reporting bias
 No opportunity for probes
 Respondents can read all the questions before starting to respond
 Difficulty in supplementing information with qualitative data

3. Telephone interviews/Online interviews

- They involve speaking or other forms of interaction to the interviewee

- The mode of communication is electronic
- They may have equal accuracy to face-to-face interviews

53
Advantages
 They are economical in relation to time
 No traveling is required
 We can interview people at a great distance

Disadvantages
 People in lower socio-economic classes may not be accessible
 Suitable for brief questionnaires
 Expensive if lengthy
 Suitable for non-sensitive topics
 Challenges of IT development
 Less complete information
 Inability to observe body language

4. Structured Observations

- Here, we observe persons, activities or organizational units

- We check for set criteria
- We make use of a checklist to record the observations
- We may interview a person for sections of the checklist if our study unit is an organization

Advantages
 They reflect people’s behaviour more accurately than their reports of behavior
 Non-disruptive
 It may be easier to access data sites since no human contact is required
 Useful for observations on quality of patient management and standards
 Relatively inexpensive

Disadvantages
 People may modify behaviour when they know they are being observed
 No opportunity for explanations

5. Document reviews

- We may use a checklist to review a set of records and extract secondary data

Advantages
 Eliminates interviewer bias
 Cheap
 Useful for retrospective cohort studies (it shortens the time)

Disadvantages
 It requires complete records
 It requires accurate records
 The range of variables on which information is recorded may change with time

e. Qualitative Methods of data collection

54
Key methods used in Public Health include:
A) Focus Group Discussions
B) Key Informant Interviews
C) In-depth Interviews
D) Secondary Data Analysis
E) Participant Observation
F) Narratives

For a detailed revision of these methods, please refer to the course: Social and Behavioural
Determinants of Health

55
Session 6: Quantitative and Qualitative Data Collection Tools
Introduction

Quantitative Data Collection Tools are prepared with a view to statistical analysis
They include:
1. Structured Questionnaires
2. Observation Checklists

Qualitative Data Collection tools on the other hand are prepared to allow for iterative flexibility and
contextualization. They include:
1. Key Informant Guides
2. Semi-structured questionnaires
3. Focus Group Discussion Guides
4. In-depth Interview and Case Narrative Guides
5. Other tools

The Qualitative tools have already been described in the Course: “Social and Behavioral Determinants of
Health” and we shall not describe them again. Please refer to your notes for this course for revision.

In this session, we shall delve into the key considerations in designing ‘Quantitative Data Collection
Tools”

Session Topics: The following topics will be covered:

a. Questionnaire design
b. Validity
c. Sequencing of questions
d. Types of Questions
e. Other considerations in Questionnaire design
f. General structure of the Questionnaire

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Describe the key considerations in design of a good questionnaire
2. Use appropriate cues in designing valid questionnaires for use in answering given research
questions
3. Select and incorporate meaningful questions that measure the intended variables in a given
research design

a. Questionnaire design

Definition: A collection of questions and statements that help the investigator to determine the extent to
which respondents hold a particular attitude or perspective or knowledge or practice

The questions must have inherent meaning and value

They must be valid (measure correctly what they are intended to measure)
56
Each question is an epidemiological test
One question or a set of questions may be needed to measure each variable

Fundamental Issues

What do we want to know?

Which questions will get me all the possible answers?
If not all answers can be got, what other additional techniques do I need e.g. direct observation, FGDs,
KIs?

The Process of questionnaire design

Start from the specific objective

Develop variables
Develop the sub-variables for composite variables
List the variables and their sub-variables
Develop questions to measure each variable
The questions must have measuring value
Select the relevant questions
Discard the irrelevant questions

b. Validity of questions

There are 3 forms of validity:

 Face validity
 Content validity
 Criterion validity

Face Validity
When formatting questions or statements, they must have Face Validity
They must seem likely to yield information of real relevance to what the researcher wants to investigate or
measure
For Example:

The Question “Do you describe your self as poor?” is likely to be confounded by subjectivity. This question
may therefore not be valid for measuring “Socio-economic Status”

To effectively assess socio-economic status, we may ask a set of screening questions?

- What is your income per year?
- Which type of house do you stay in?

Content Validity
When formatting questions or statements, they must have Content Validity
The questions must take into account all the content elements of a variable

For Example:
The variable “Satisfaction with medical care” must have a composite scale to measure satisfaction

The set of questions may then take care of: Experience with doctors, experience with nursing staff, waiting
time, cost of care, convenience and several other issues
57
Criterion Validity
When formatting questions or statements, they must have Criterion Validity
They should aim at getting answers as close as possible to the gold standard

For Example:
When measuring age:
Asking for the Date of birth may be more valid than asking for age
When measuring age, we need to ask for “age in completed years”

Qualities of a good questionnaire

Sensitivity- Ability to identify correctly those who have a particular condition

Specificity- Ability to identify correctly those who do not have a particular condition
Questions must be non-threatening and anonymous
They should ask for an answer in only one dimension
They should accommodate all possible answers
There should be questions to cover all variables
Options should be mutually exclusive

For situations in which people have more than one experience, we ought to use checkboxes
The questions should not presuppose a certain state of affairs
Questions should not imply a desired answer
Does not use un-familiar words or abbreviations
Does not ask the respondent to order or rank a series of more than 5 items
It avoids useless questions
It incorporates checking questions to enhance validity for specific questions where biases are likely

Attributes of good questions

Simple, short and specific

They must deal with a single variable or sub-variable at a time
They must have face validity
They must have content validity
They must have criterion validity
Respondents must be expected to have the answer
They must not be offensive
They should be fair
Avoid negative questions or terms because they confuse respondents
Avoid leading questions because they introduce bias
Questions should convey the same message to all respondents
Avoid questions joined by articles or prepositions “and” and “with”
Joined questions are called double barreled questions and answer two variables

c. Sequencing of questions

Consider
- The Pattern
- The Rigour
- The Logical flow
58
Pattern
Start with simple questions
Follow with interesting questions that will capture attention
Move on to the more difficult ones
Embarrassing questions ought to come last

Rigour
Start with an opening question
Inter-sparse them with follow-up questions
You may then insert probes until the subject is depleted
Then more to the next question

Logical flow
Try to group questions according to subject
The questions should flow naturally or logically
Do not jump to the end of a given scenario then run back
The best questionnaires are structured according to subtitles
There are two approaches to subtitling:
According to Chronology of events, so that there is temporal flow
According to Outlay of Objectives, so that

1. Organizing Questions according to chronology of events

What event follows another?
How do events unfold to complete the story?
What are the main milestones in the process?
Now categorize the different events into milestones
These will form the subtitles of the questionnaire

Example:

Sub-title a. Events surrounding progression of illness

Main Question 1: When did you first feel ill?
Follow Up Question: What were the complaints?
Probe: Which of these was the most disturbing complaint?

Main Question 2: Did you inform your spouse?

Follow Up Question: If yes, after how many days since the start of your symptoms?
Probe 1: Were you hesitant to tell him/her?
Probe 2: If yes why?

Sub-title b: Health Care Seeking

Main Question 1: When did you first seek treatment?
Follow Up: Where did you go for the first consultation?
Probe: Did you get treatment there?

Main Question 2: Did you seek treatment elsewhere?

Follow-up: If so, where?

Etc, etc until the story is exhausted

59
2. Organizing questions according to outlay of objectives

Objective 1: To assess health system factors affecting up-take of VCT

Question 1: Is there a HC offering VCT in your locality?

Follow-up: If yes, what is the distance to this Centre?
Probes:
Don’t ask: Is this distance long and Costly
Instead ask: Is this distance long?
What is the transportation cost to this HC?
Don’t Ask: This is costly, isn’t it?
Instead Ask: Is this costly for you?

Objective 2: To Assess Individual Perceptions affecting up-take of VCT

Question 1: Have you ever gone for VCT

[This is a very important Question because it is a screening question; for prevalence studies and
comparative studies, it may be the basis for comparison, it collects information on the independent variable
– Uptake of VCT]

Follow-up: If not, why?

If yes, what motivated you to go for VCT?

d. Types of Questions

1. Questions may be open ended

They allow for a multitude of responses
They allow for many individual experiences
They are post-coded or categorized after the study
A few options may be given, and an open option “Others (Specify)” is provided for
Used in “Semi-structured” questionnaires
They allow for limited qualitative perspectives
All options should be mutually exclusive

2. Questions may be closed

The options are specified
Options are often limited
All options should be mutually exclusive
Useful when all options are known
Easy to analyze
They are the rationale for “Structure”
All options should be mutually exclusive

Non-mutually exclusive issues

We may ask respondents the different types of food they usually eat
60
This may require multiple options
In this case, each food is considered a sub-variable itself
We use “check-boxes”
In the computer, checkbox is considered as a separate variable

d. Other considerations in Questionnaire design

Pre and Post Coding

Structured questionnaires ought to be pre-coded

We assign a code for each question, as a separate variable
We then assign codes to each option within a question
Very often, it may be advantageous to enter numbers alone in statistical software
For open questions, we need to do post coding
If you use numbers to enter your data, do not lose your coding sheet

Translation of questions

Sometimes questions may need to be translated before hand

This limits interviewer bias especially in KAP studies
The translator ought to be linguistically competent
If not translated, then translations should be thoroughly discussed when training data collection assistants

e. General structure of the Questionnaire

A. Identification Section
B. Background Characteristics
C. Main Body
D. Conclusion

A. Identification Section
Questionnaire Number or official code
Interviewer’s Identification
Date
Time of start
Self introduction
Consent Form

B. Background Characteristics
Interviewee number
Demographic Characteristics
Socio-economic characteristics
Other Background Characteristics

C. Main Body
General titles
Subtitles
Questions
Follow-up questions
Probes
61
D. Conclusion
Allowance for interviewee queries
Concluding remarks
Time of end

62
Session 7: Study Variables and Measurements

Session Topics: The following topics will be covered:

a. General Variable Types
b. Other types of Variables
c. Re-phrasing and Operationalising Variables
d. Measurements

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Describe and contrast the different types of variables
2. Employ appropriate strategies in operationalising variables into measurable parameters
3. Conduct accurate research measurements using appropriately calibrated or validated
measurement tools
4. Select relevant parameters to design appropriate tools to measure composite variables like
‘quality of care’

Variable: A characteristic of a person, object or phenomenon that can take on different values.

a. General Variable Types

The common types of variables are:

 Numerical and Categorical
 Dependent and Independent
 Confounding or intervening
 Background and Socio-demographic

Numerical and Categorical Variables

Numerical Variables e.g. age, wt, distance, income

Can be discrete (e.g. number of drinks, age in completed years)
Can be continuous (e.g. weight, height, waist size)

Categorical variables e.g. color, disease outcome, type of food

Dichotomous (yes/no) (e.g. death, CHD, DM)
Nominal (no order) e.g. ethnicity, occupation
Ordinal (ordered rank) e.g. severity of disease

Dependent and independent variables

Dependent Variable
Used to measure or describe the problem under study
Variables that change as a result of the level of the independent variables
The ‘depend’ on the independent variables
Nutritional status depends on infant feeding practices
Practices depend on knowledge and attitudes
63
Synonyms: effect, outcome, consequence, result, condition

Independent variables
Variables that are manipulated or treated in a study
Variables that affect the measured effect of the dependent variables
Used to describe or measure the factors that are assumed to cause or influence the problem
Synonyms: predictor, cause, input, predisposing factor, antecedent, risk factor, attribute,
Their variation may affect the outcome measure of the dependent variable
NB: Despite being independent, some of them may affect each other
This is the basis for confounding, collinearity and effect modification
Remember: They can be hierarchical:
 Immediate causes or proximal factors
o E.g. PEM is caused by inadequate feeding
 Underlying factors
o E.g. PEM is underlied by lack of knowledge on feeding
 Distal or Background factors
o E.g. PEM is underlied by low education of the girl child

b. Other types of Variables

Confounding or intervening variables

Variables that should be studied because they may influence or “confound” the effect of the
independent variable(s) on the dependent variable(s)
Background variables
Variables that are so often of relevance in investigations of groups or populations that they should
always be considered in the study
E.g.: sex, age, ethnic origin, education, marital status, social status

c. Re-phrasing and Operationallising Variables

Rephrasing factors as variables

We have to move from Factors in the conceptual framework to Variables
We need to rephrase the factors highlighted in the conceptual framework into variables Examples include:
 Long waiting time – Waiting time
 Absence of drugs – Drug availability
 Lack of supervision – Frequency of supervisory visits
 Poor knowledge of the signs, causes and consequences of TB – Knowledge of the signs, causes,
and consequences of TB

Operationallising variables
Some variables cannot be measured directly
Examples include:
Performance of a programme
Client Satisfaction
Quality of care
Very often, these variables are either in non-specific forms
Or require a composite of attributes to be measured
We can operationalise these variables

64
By creating measurable options from the variables

Examples
Level of knowledge: This cannot be measured directly, but a series of questions can be set, and a scale
developed e.g.
0-3 correct answers is poor
4-6 correct answers is reasonable
7-10 correct answers is good

Nutritional Status of Under-5 year Olds: We can choose appropriate indicators such as:
Weight in relation to age
Weight in relation to height
Height in relation to age
Upper arm circumference

d. Measurements

The Measurement Spectrum

After deciding the exposure/outcome of interest, measurement includes:
 Instructions for applying the method for measuring the variable
 The measurement method itself
 Delineation of steps of preparing the collected data for analysis
 Quality control procedures (i.e. making sure you get what you meant to get)
 Improving Precision and Accuracy of Variables and Reducing Bias
 Standardize methods

Important considerations before measurement:

 Pretest the measurement tools
 Refine/automate instrument
 Train and evaluate staff
 Timely editing, coding and correcting of forms
 Multiple measurements where needed
 Validate against “gold standard”
 Blind to exposure status
 Institute quality control measures during data collection, processing and analysis

Forms of Measurement
 Some variables require direct measurement
 Some measurements are in-form of questions
 Some variables are composite

1. Some variables require direct measurement

Examples: Height, weight, blood sugar
However, describe in detail how the actual measurement is going to be done
Examples:
 Height to be measured using a standard height meter, with the shoes removed, and the patient
standing up right. To the nearest 0.1 Meter
 Weight will be determined using a freshly re-calibrated weighing machine, of the Salter or Seca
brand, with the patient lightly clothed. To the nearest 0.1 Kg
65
  The waste circumference will be measured from a horizontal plane based at the mid-point between
the lower costal margin and the iliac crest, with the waistline unclothed, and using a tape measure
 The hip circumference will be measured in a horizontal plane between the inter-trochanteric lines,
using a tape measure.

2. Some measurements are in-form of questions

 Questions must have validity
 There are 3 forms of validity:
o Face validity
o Content validity
o Criterion validity

3. Some variables are composite

Examples:
Quality of life
Health related quality of life
Economic status
 They require the use of a battery
 We can use an existing battery
 We can construct a new battery
 New Batteries must be validated
 Validation is by Factor Analysis

66
Session 8: Data Management and Quality Control in Research

Session Topics: The following topics will be covered:

a. Definitions and Rationale for Data Management and Quality Control
b. Quality Control

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Defend the rationale for good quality control in research
2. Describe the quality control measures that can be incorporated in the pre-study phase, the
data collection phase and data analysis phase of research

REFERENCES:
1. Hulley SB, Cummings SR : Designing Clinical Research: An Epidemiological Approach
2. Bernard Rosner (1990): Fundamentals of Biostatistics (Harvard University Press – 3rd Edition)

a. Definitions and Rationale for Data Management and Quality Control

Quality Control: A series of systematic steps taken to ensure that data or information generated from
research is reliable, complete, accurate, error free and above all, reproducible, using the same methods.

Data Management: A systematic series of steps taken to ensure data is complete, accurate and error free,
ready for statistical analysis. It is just one component of the quality control measures in research and
perhaps one of the most important

Need for Quality Control in research: Quality Control assures quality and consistency in data. However,
it often receives little attention by young researchers. Data collection is a repetitive process and therefore
requires quality control guidelines to standardize procedures throughout the conduct of the research. In
absence of these procedures, we may have the likely problem of:
- Missing data
- Inconsistency in data collected
- Erroneous data

b. Quality Control

Quality control occurs

1. Before the study
2. During the study and
3. After the study

1. Pre-study Quality Control Steps

Quality control begins in the planning stage. It is important to define the quality control measures in the
study protocol, and develop a manual or guidelines for each procedure and operations, especially for large
studies.
67
 It describes in detail and step by step how each procedure is to be accomplished (e.g. identification of
study subjects, how measurements are to be made, how to complete the forms, etc)
 If specialized procedures are used, prepare Standard Operating Procedures (SOPs) for each
procedure; this helps to reduce variation and changes in the way measurements are conducted
 Standardize data collection instruments and forms
o Develop a suitable ID system for study subjects – preferably in consultation with person to
manage the data
o Pre-coded answers are recommended and preferred wherever possible (develop appropriate
coding)
o Collect only the information needed; do not overload questionnaire forms
o Prepare computer data entry formats (choice of appropriate software important)
o Pre-test your data collection tools so as to validate each question and possible responses
o Revise your tools to incorporate more valid questions and remove redundant ones
 Identify qualified personnel
o Very often, researchers are biased by familial ties, cultural issues and convenience
o It is important to define the organizational structure, the responsibilities and the duties
 Train study personnel
o This helps to reduce variation in quality of measurements
o The purpose of the sessions is to orient personnel to the research objectives and study
procedures

2. During Study Quality Control Steps

 Supervision
o The investigator or designate must provide or ensure adequate supervision
o The investigator must ensure or check on completeness and accuracy of completed data
collection forms. This should preferably be on a regular basis
o The investigator should ensure that procedures are being followed as defined in the study
protocol or SOP Manuals
 Regular meetings with study personnel
o These help identify any problems cropping up
o They boost morale of study personnel and assures them of the PI’s interest in what they are
doing – this is very important
 Performance review
o Interim reports on target attainment – this may give early warning on things going wrong
 Laboratory Procedures
o Pay attention to labeling of samples
o Blinding procedures
o Procedures and assays should be conducted as those described in SOPs

3. After Study Quality Control Steps

 Data Management
o Manual and electronic data storage procedures must go hand in hand
o For large data bases, develop databases
o Pay attention to data cleaning; by cleaning, we refer to data entry errors, inconsistent entries
and missing entries
o Regular back up of data (have at least two copies of data)
o Safe storage of scripts and electronic storage devices)
o Post code the open ended questions
68
 o Safely store your master coding sheets
 Quality checks in Data Analysis
o Frequencies, counts
o Logic checks (Checking questions, consistency checks)

69
Session 9: Ethical Considerations in Development of a Research Problem

Introduction: Ethical considerations in research are a very important aspect of the process. Research,
especially that involving human subjects has been the source of many historical violations of human rights.
Researchers ought to incorporate rigorous measures to ensure that study subjects are protected. In this
session, we shall look at the strategies for making ethical provisions in research.

Most of the theory underlying Research Ethics is contained in the course: Health Ethics and Law;
please refer to these notes to strengthen your understanding of the subject
This session will mainly delve into the aspects of application of these principles in a research setting

Session Topics: The following topics will be covered:

a. The basis of Research Ethics
b. Principles of research ethics and strategies for their application in Research
c. Key Methodological Ethical Considerations

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Defend the rationale for incorporating ethics in research
2. Describe the strategies that can be used in ensuring that the different principles of research
ethics are adhered to in a research process
3. Describe the key methodological applications of ethical principles as would be incorporated
in a research protocol

Firstly, let us look at the basics:

a. The basis of research ethics

Definition: Ethics is the science of criteria, norms and values for human action and conduct. Research
Ethics determines the norms and values for the conduct of research.

Vulnerable populations: There is a special role of ethics in protecting vulnerable populations. These
populations include: Women and children, the disabled, refugees, prisoners, the poor and the
disadvantaged. Very often, research participants may become vulnerable people owing to the research
itself and its benefit vis-à-vis its dangers.

Are research participants in developing countries vulnerable?

 Many research participants are illiterate and language barriers exist
 The process of informed consent is sometimes violated
 They have limited or no protection from IRBs
 Less stringent trials regulations from the authorities
 Ease of convincing local researchers and politicians
 Lack of health care infrastructure to support research
 Different standards of care applied

70
History of research ethics: Important milestones in the development of research ethics include:
1. The Nuremberg Code
2. The Declaration of Helsinki
3. The Belmont Report
4. Common Rule
5. CIOMS
6. International Conference on Harmonization
7. National Bioethics advisory committee

b. Principles of research ethics and strategies for their application in Research

The key principles are:

1. Respect of persons
2. Beneficence
3. Justice
4. Non-malefiscence

1. Respect of persons
Foundation:
- Autonomy and self-determination
- Notions of human dignity
- Consent can be revised at any time and any stage in the process
- Organizational consent where appropriate
Strategy:
- Information about the study must be given to the participants
- Participants should give free and informed consent
- Consent can be revised at any stage of the process
- Organizational consent where appropriate

2. Beneficence
Foundation:
- Benefit to the research participant
- Weigh risks versus benefits
- They ought to include physical, mental and social well-being
Strategy:
- Demonstrate that the study will minimize harm and maximize benefit
- There ought to be no short-term individual benefits
- Harm (including inconvenience) needs to be minimized
- In experimental studies, the study needs to be terminated if benefit is very obvious before
the end of the study

3. Justice
Foundation:
- Notably distributive justice
- Equal distribution of risk and benefits between communities
- Equitable recruitment of participants
Strategy:
- No specific and deliberate exclusion of population sub-groups
71
 - Minimal harm and maximal benefits in the study

4. Non-malefiscence
Foundation:
- Absence of harm to the research participants
- Protection of participants
- Deliberate harm should not be tolerated at all
Strategy:
- Respect for privacy and confidentiality
- Assurances that all information will be held confidential
- Key informants are more prone to breaches because of their small numbers
- There is need to maximize anonymity
- Termination of the study if harmful effects are observed

b. Key Methodological Ethical Considerations

The Ethics section of the research proposal ought to describe the following:
 Review process
 Informed consent
 Confidentiality Safeguards
 Risks
 Benefits

1. Review process
We ought to make a note on which body will review the proposal and how it will be approved and passed.
The review process may include:
 Institutional (Higher Degrees Research and Ethics committee of MUIPH). There may be other
Institutional Review Boards depending on the respective institution that has jurisdiction over the
area of study. Examples of these include: The JCRC Board, the Medical Research Council Board,
Makerere University faculty of Medicine Research Board and others related. These Committees
assess the proposals on the following:
o Scientific review for merit
o Ethical review for compliance
 National (In Uganda, the National Council for Science and Technology). Other countries may have
other institutional arrangements. These agencies offer:
o Authorization to conduct the study
o Identity cards for investigators
o Overall approval for the study.

2. Informed consent
We should indicate how consent will be sought and how information shall be given prior to the consent. We
ought to indicate ‘full informed consent’. If our study will deal with minors, we ought to indicate who will
consent on their behalf, and how we shall solicit their own ‘assent’. We also have to explain how we are
going to safeguard the rights of vulnerable people e.g. the disabled, the aged etc. We ought to discuss
special ethical dilemmas, especially those regarding minors that have no guardians.

3. Confidentiality Safeguards
We should describe how we are going to maintain confidentiality and safeguard the information given to us,
especially from particularly vulnerable groups

72
4. Risks
We should be able to discuss the risks involved

5. Benefits
Potential benefits have to be evaluated and communicated.

In addition, all data collection tools ought to have a consent form attached (in the appendices) clearly
indicating that there will be signed consent

73
Session 10: Workplan and Budgeting

Session Topics: The following topics will be covered:

a. The Workplan
b. The Budget and budget justification

Session Objectives:
By the end of this Session, the MPHO should be able to:

a. The Workplan

A workplan is a schedule, chart or graph that summarizes in a clear fashion the various components of a
research project and how they fit together. It includes:
- The tasks to be performed
- When the tasks will be performed
- Who will perform the tasks
- The time each person will spend on them

The Gantt Chart

It is a planning tool that depicts graphically the order in which various tasks must be completed and the
duration of each activity. A bar that extends over the number of days, weeks or months that the task is
expected to take indicates the length of each task.

Example of a Gantt Chart

Tasks Person April May June July Aug Sept Oct

Responsible
Finalize proposal Research Team
Liaise with district Research Team
authorities
Train Research PI
Assistants
Pretest tools PI and RAs
Collect Data PI and RAs
Analyze Data
Report writing and
dissemination

Purpose of a workplan

a) A tool in planning the details of the project activities and later in the budgeting of activities
b) A visual outline or illustration of the sequence of project operations. It can facilitate presentations
and negotiations concerning the project with government authorities or other potential funding
agencies
74
 c) A management tool for the principle investigator and members of his/her team, showing what tasks
and activities are planned, their timing and when various staff members will be involved in various
tasks
d) A tool for monitoring and evaluation, when the current status of the projects is compared to what
has been foreseen in the workplan

When should a workplan be prepared and when should it be revised?

- The first draft of the workplan should be prepared when the project/research proposal is being
developed to that the schedule can be discussed easily with the relevant authorities
- A more detailed workplan should be prepared after the Pretest in the study area
- There should be no hesitation in revising workplans or preparing new ones after the project is
underway, based on re-assessments of what can be realistically accomplished in the period
indicated

What to keep in mind when preparing a workplan

- It should be simple, realistic and easily understood by those directly involved
- It should cover the preparatory and the implementation phase of the project as well as data
analysis, reporting and dissemination/utilization of results
- The activities covered should include technical or research tasks; administrative, secretarial,
training tasks and other support tasks
- The realities of local customs (local holidays, festivals etc) and working hours should be considered
when preparing the workplan
- Also, seasonal changes and their effects on travel, work habits and on the topic you are studying
(such as incidence of disease or nutritional status) should be kept in mind as the schedule is
planned

b. The Budget and budget justification

The study budget

A detailed budget will help you to identify which resources are already locally available and which additional
resources may be required. The process of budget design will encourage you to consider aspects of the
workplan you have thought about before and will serve as a useful reminder of activities planned as your
research gets underway.

Use the workplan as the starting point to prepare your budget. Specify for each activity in the workplan
what resources are required. For each resource needed, determine the unit cost and the total cost.

It is important to budget within the resources that you realistically hope to get, otherwise, failure to obtain
exorbitantly estimated resources may be a cause for invalidating your findings

75
Example of a Budget:

Budget Category Unit Cost Multiplying factor Total Cost

1. Personnel Daily allowance or per No. Of staff X No. Of
diem Days
Interviewers $10 5 X 5 = 25 $ 250
Supervisor $20 1X5=5 $ 100
Sub Total $ 350
2. Transport Cost per Km No. Of Km X No. Of
Days
Motorcycles $ 0.10 4 X 5 X 100 $ 200
Car $ 0.40 1 X 5 X 100 $ 200
Sub Total $ 400
3. Supplies Cost per Item Number

Pens $1.00 12 $ 12
Questionnaires $ 0.20 120 $ 24
Sub Total $ 36
GRAND TOTAL $ 786

If more than one budget source is envisaged (e.g. MOH and a donor) it would be useful to indicate in the
budget which source will pay for each cost. Usually, a separate column is used for each funding source.

The Budget Justification

It is not sufficient to present a budget without explanation and detail on the rationale for the different
provisions. The budget justification follows the budget, as an explanation note justifying briefly, in the
context of the proposal, why the various items in the budget are required. It is important to make sure that
you give clear explanations concerning why items that may seem questionable or those that are particularly
costly are needed and discuss how the complicated expenses may have been calculated. If a strong
budget justification has been prepared, it is less likely that essential items will be cut during the proposal
review.

Session 11: Alternative Study Designs

Introduction: In actual PH practice, we often use alternative methods and study designs that differ from
the common epidemiological study designs. These methods are based on purposiveness, convenience,
and allocativeness. The main focus is not generalisability but deeper understanding. They may be the basis
for designing interventions. They may also be the basis for evaluation of programmes. They provide an
opportunity for rapid assessment. They are also useful in resource constrained environments. In this
session, we shall examine briefly some of the more common alternative methods.

Session Topics: The following topics will be covered:

a. Why Alternative Study Designs?
b. The Alternative Study Designs
76
Session Objectives:
By the end of this lesson, the MPHO should be able to:
1. Justify and explain the rationale for use of alternative study designs in research
2. Select and use appropriate alternative designs to answer relevant research
questions, especially the systems analyses

a. Why Alternative Study Designs?

These study designs differ from the usual epidemiological study designs especially in the fact that they use
alternative ways of investigating research phenomena. In the context of the epidemiological study designs,
they are mainly variants of crossectional study designs. They find particular use in systems research and
are increasingly used by program managers in evaluation of projects and programs. They include:
1. Qualitative Research Designs
2. Quasi-Experimental methods and non-randomized studies
3. Expert opinions
4. Systems analyses
5. Rapid Rural Appraisal
6. Qualitative methods
7. Costing studies and Economic Appraisal

b. The Alternative Study Designs

1. Qualitative Research Designs

Qualitative research methods are a form of alternative study designs that can be used in your studies.
Details of these methods have already been described in this manual and on other courses.

A detailed review of Qualitative Methods can be found in your notes for the course: Social
and Behavioral Determinants of Health

Many aspects of Public Health can be investigated using purely qualitative methods or a blended approach
using both qualitative and Qualitative methods and ‘triangulating’ the results to enhance the discussions
and hypotheses.

2. Rapid Appraisals
 Rapid Assessments of the situation at present
 Used in formative stages of programs
 We completely ignore the tedious processes involved in structured research
 We collect information directly, on the areas of interest
 We collect information conveniently from target populations (purposive)
 Useful for involving communities in designing interventions based on their priorities
 PRA is mainly qualitative in nature

Advantages of PRA
77
  High community participation; a bottom-up planning
 Non hierarchical; individual contribution is important
 Short duration and low cost
 Allows for flexibility and is normally not superficial
 Allows for progressive learning during fieldwork
 Little or no statistical analysis applied
 Multi disciplinary teaming
 Best for learning and understanding rural community opinions, behaviors and attitudes
 Employs a variety of enriching methods
 Formal preparations and questionnaires minimized

Short comings of PRA

 Difficulty in finding the right team
 Rushing may create superficiality
 Difficulty in identifying the marginalized
 Failure to involve entire community
 Representation by early adopters/leaders hence low representativeness
 Lack of rapport with the community
 Over generalization based on too little information

Considerations in PRA
- Allow for flexibility in the design; change the design depending on emergent themes
- Avoid unnecessary detail and over collection of information
- On the spot analysis – analysis concurrent with the data collection
- Endeavour to include the marginalized; the quiet and the remote
- Learn progressively as events unfold
- Share findings with participants
- Link the research with action

PRA Methods: The major research methods/tools are:

 Semi-structured interviews (Interview Guide)
 Focus Group Discussions (Discussion Guide)
 Key Informant Interviews (KI Guide)

Other PRA Methods include:

Records Analysis: Secondary data reviews; review of routine records
Ranking: Preference ranking, Pair-wise ranking, Direct Matrix ranking, Ranking by Voting, Wealth Ranking
Analytical Groups: Analysis Group discussions, Innovation Assessment
Mapping: Participatory mapping, Historical mapping, Future (visioning) mapping, Mobility mapping, Social
Resource Mapping and Transect walks
Calendaring: Seasonal calendar, Historical seasonal calendar
Other PRA Methods
Trends: Time trends, Historical Profile, Livelihood analysis
Diagramming: Flow causal diagram, Systems diagram, Venn/Institutional diagram, Potential Partner
matrix, Pie chart, Histogram
Observation: Participant observations, Non-participant observations
Groups: Workshops, Group walks, Group interviews, Group discussions

78
Case experiences: Stories, Oral histories, Case studies and portraits, Proverbs, Verbal diaries, Memory
books

4. Quasi-experimental studies
 They look like experiments
 However units of application not randomized
 Not only to test new intervention but to do the intervention itself as a public good
 Described as “before and after” assessments
 We assess the same study unit before and after the intervention
 Commonly used in evaluating intervention programmes

5. Systems Analyses
 Special forms of situational analyses
 They are in reality crossectional studies
 Employ a mix of qualitative and quantitative methods
 Mainly used for evaluation of programmes and projects
 Evaluation may be formative or summative
 They track project from inception to conclusion
 They study systems and assess their functional contribution to success, failure or delay of a
program
 Analysis of systems is related to:
o Inputs
o Processes
o Outputs
o Outcomes
o Impacts

In systems analyses, we use indicators:

Inputs Input Indicators
Processes Process Indicators
Outputs Output Indicators
Outcomes Outcome Indicators
Impacts Impact Indicators

6. Costing Studies and Economic Appraisals: Costing studies often use un-conventional methodologies
in deriving fixed costs, marginal costs and other related computations. Economic appraisal methods like
cost-benefit analysis (CBA), cost-utility analysis (CUA) and cost-effectiveness analysis (CEA) also use
specific methods that will be discussed in detail in the course: Health Economics.

79
 2.3 Unit 3: QUALITATIVE RESEARCH METHODS

2.3.1 Introduction to the Unit

A central part of Public Health is research. Public Health Research Methodology embodies two
complementary approaches namely, Qualitative and Quantitative methods. While it is recognised that
quantitative methods are useful for the generalisation of findings from statistical samples to populations, it
is also important to note that they do not give an in-depth understanding of context as an explanation to
people’s experiences. Therefore, in order to fully appreciate socio-cultural and normative environments,
and the richness of people’s unique experiences (to understand why people behave the way they behave),
the behavioural sciences extensively employ qualitative research methods to evaluate behaviour. This unit
will delve on various applications of these methods in Public Health inquiry especially noting the role of
behavioural sciences in Public Health

2.3.2 Unit Outline

The following topics will be covered:
Session 1: Introductory concepts
Session 2: Qualitative Research and its role in Epidemiological Inquiry
Session 3: Social Science Methods Appropriate In Public Health
Session 4: Rapid Assessment Procedures
Session 5: Design of Qualitative Data Collection Tools
Session 6: Qualitative Data Analysis
Session 7: An in-depth view of Triangulation

2.3.3 Instructional goal

The MPHO should be able to design, implement and analyse qualitative research on behavioural and
socio-cultural aspects of health

2.3.4 Unit Objectives

By the end of this unit, the student should be able to:
1. Report on basic concepts that link the Social Sciences to Public Health and research
2. Elucidate the role of qualitative research in Epidemiological inquiry
3. Explain in detail, the social science methods appropriate in Public Health
4. Set up, implement and evaluate a Rapid Appraisal for a social intervention of Public
Health Importance
5. Synthesise and employ effective KI, FGD and Semi-structured Interview Guides in
collection of qualitative data
6. Conduct an analysis of qualitative data transcripts
7. Employ qualitative techniques to triangulate quantitative methods as a means of
increasing validity
8. Write, publish and critique scientific literature

2.3.5 Time Frame

1 WEEK

2.2.6 Content

80
Session 1: Concepts in Social Health Research

Introduction: This Session will introduce you to the key concepts that link the behavioural sciences to
Public Health
Session Topics: The following topics will be covered:
a. Definition of key terms and concepts
b. History of social science in Public Health work
c. Public health and Research
d. Methods utilized within the social sciences appropriate for public health

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Define the key components terms in the social/behavioural sciences
2. Define the key terms that link the social/behavioural sciences to Public Health
3. Analyse the use of research as a tool of Public Health
4. Summarize in broad categorisation the methods used in Public Health research

a. Definition of key terms and concepts

Students to make a write up on the definition of these concepts using available literature:
a) Anthropology
b) Economics
c) Political science
d) Psychology
e) Social work
f) Sociology

b. History of social science in Public Health work

The social sciences are very important in public health. Below, one finds some of the major components
that are used in public health.
 Social epidemiology
 Medical anthropology
 Health psychology etc.

c. Public health and Research

Definitions and aims of Public Health: Public health may be defined as activities that society undertakes
to assure conditions in which people can be healthy and in which people can prevent, identify and counter
threats to the health of the public.
It is the science and art of preventing disease, prolonging life and promoting health through
organised effort of society. The effort includes: Community mobilization and participation, Public
Health Education, Health Policy formulation, evidence based decision and action research.
Definition of Health: The state of complete Physical, mental and social well-being and not merely the
absence of disease or infirmity (WHO 1947).
Substance of Public Health: It consists of organised community efforts aimed at the prevention of disease
and the promotion of health. It links many disciplines but rests on the scientific core of Epidemiology.

81
Definition of Epidemiology: It is the study of the distribution and the determinants of health related states
and their use in the prevention of disease and injuries in human populations.
Classic examples of epidemiological research include: Case-Control studies, Randomized Clinical trials,
community trials and cohort studies.
Foundations of Research: The primary purpose of research is to conduct scientific or scholarly
investigations into a phenomenon or to answer a question. It is therefore defined as the systematic
approach to problem solving. Good research recognises the norms of Science.
Definition of Research: Research is the organised quest for new knowledge based on curiosity or
perceived need. It may consist of systematic empirical observation or hypothesis testing.
The Norms of Science: Science is an objective logical and systematic method of analysis of phenomenon
of interest to permit the accumulation of reliable knowledge. (Lastrucci, 1963)
Objective: The idea of truly objective inquiry has long been understood as a delusion. Scientists do
hold that striving for objectivity is useful; in practice, this means constantly trying to improve
measurement to make it more precise and accurate and submitting findings for peer review.
Method: Each scientific discipline has developed a set of techniques for gathering and handling
data
Reliable: Something true in Kampala should be replicated in New York, if the conditions are the
same.
Categorizing Research: Research can be:
 Basic vs. Applied
 Quantitative vs. Qualitative approaches
Quantitative Research: It generates data from numbers.
Examples of methods include census taking, household surveys, administering structured
questionnaires, telephone interviews.
Qualitative Research: Generates data from words, observations of events, pictures etc.
Examples of methods include Participatory mapping, Free listing, Participant observations, Key
Informant interviews, Case narratives, Pile sorting, Case narratives, Focus Group discussions and
Timelines.
Public Health Research: As noted, Public Health is strongly associated with Epidemiology.
Epidemiologists most commonly use quantitative research methods to study issues related to human
disease. However, social and behavioural science using qualitative research methods is also an important
component of Public Health.

d. Methods utilized within the social sciences appropriate for public health.

In the social sciences, both qualitative and quantitative methods are used. However more emphasis is
laid on qualitative approaches which aim at in-depth information, thick description rather that numbers

82
Session 2: Qualitative Research and its role in Epidemiological Inquiry

Introduction: Under this Session you should be able to describe qualitative research, distinguish between
qualitative and quantitative approaches, and characterize qualitative approaches.

Session Topics: The following topics will be covered:

a. Public health and Research
b. Characteristics of Qualitative Research
c. Qualitative and quantitative methods
d. Rigor in qualitative research
e. Characteristics of Participatory Qualitative Research in PH

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Define research and discriminate participatory from non-participatory research
2. Contrast qualitative and quantitative research methods to elucidate their characteristics
and the rationale for their differences
3. Appraise the rationale and value of iterative flexibility in qualitative research
4. Evaluate qualitative and quantitative research methods to highlight the situations in
which each is applicable
5. Defend the complimentarity of quantitative and qualitative research methodology
6. Use probes to demonstrate how rigor is applied in strengthening the internal validity of
qualitative research

a. Public health and Research

Research: It is defined as the organized quest for new knowledge based on curiosity or perceived need. It
may consist of systematic empirical observation or hypothesis testing.
Participatory Research: It is research in which the subject of the study plays a prominent role in
some or all phases of the research. Qualitative Research is often participatory research, as the
respondents are democratically involved in the iterative process of reshaping the direction the
research should take. The study participants play a prominent role in some or all phases of the
research.
Public Health: It basically is the science and art of preventing disease, prolonging life and promoting
health through organized effort of society i.e. community mobilization, public health education, health policy
formulation, evidence based decision and action research.

b. Characteristics of Qualitative Research

Qualitative research is characterized by an approach which seeks to describe and analyze culture and
behavior of humans and their groups from the point of view of those being studied. Qualitative research
emphasizes on providing a comprehensive or "holistic" understanding of the social settings in which
research is conducted.
Iterative Flexibility: Qualitative research relies on research strategy which is flexible. The
flexibility of qualitative research allows researchers to reduce this bias by using interviewing
techniques & creating an interview environment which the respondent feels most comfortable. This

83
 approach allows discovery of unexpectedly important topics which may not have been visible to the
researcher at onset.
Qualitative Research embodies four approaches: Explorative flexibility, Iteration, Triangulation
and Contextualisation (Smith and Morrow, 1996)

Complementarily of Qualitative and Quantitative Research: Qualitative and Quantitative research are
complementary and are often combined in a way that maximizes the strength and minimizes limitations of
each. Qualitative researchers spend more time in community, develop rapport & identify potential sources
of contextual bias.
Some examples of methodological complementation: Qualitative research can facilitate
quantitative research by identifying factors appropriate to survey methods; on the other hand
quantitative research can be used to facilitate qualitative research by generalizing findings to a
large sample, or by identifying groups that warrant in-depth qualitative study.

Perspectives in Qualitative Research: The main perspectives in Qualitative Research are referred to as
the emic and etic. Anthropologists emphasize understanding culture, in context trying to discover how
people view their own situation and how they solve their own problems. This is the emic
perspective/approach. It is the insiders’ point of view. In contrast the etic perspective is based on
outsiders, policy makers, health workers view point about particular groups i.e. "Researchers" outsiders
point of view. Often projects fail because they do not take the community's own ideas and preferences into
account.

When qualitative approach is most appropriate to use? Qualitative Research is appropriate in the
following situations:

Box 1: When to use qualitative research

Appropriate use of qualitative research
When the subject matter is unfamiliar.
For exploratory research, when relevant concepts and variables are unknown or their definitions are unclear.
For explanatory depth; when you want to relate aspects of behavior to a wider context.
When a meaning rather than frequencies is sought.
When flexibility of approach is needed to allow for discovery of the unexpected, in-depth investigations of some
topics.
For studying selected issues, cases or events in depth & detail.

Qualitative research is especially useful for:

1. Exploring a health problem/issue of which little is known
2. Identifying local perceptions of health and development priorities
3. Identifying relevant intervention strategies and target populations
4. Investigating feasibility, acceptability and appropriateness of potential new health programs
5. Develop appropriate information, education and communication (EIC) activities and materials
6. Identifying problems in on going interventions & suggesting appropriate solutions to those
problems
7. Complementing Quantitative data in routine monitoring and evaluation studies by helping to
interpret Quantitative results
8. Designing more valid survey instruments by identifying topics for which survey research is
appropriate and by identifying relevant questions & their appropriate wording

Who should conduct Qualitative Research? The research team should be multi disciplinary e.g.
qualitative researcher & disease control specialist. This helps to identify and keep the study focused on
relevant issues. It is important to involve local community e.g. field guides, translators (participatory)
84
approach. The researcher should be willing to spend much time in the field. This helps to understand the
social context in which people live.

c. Qualitative and quantitative methods

Quantitative methods: They are based on measures of quantity/frequency. Their findings are described
in numbers rather than words. Quantitative research is one in which a pre-determined series of questions
are asked in the exact same way to every respondent. This approach has long dominated social
investigations for National planning and program evaluation. Surveys attempt to measure social opinions
and behaviors (KAP Surveys). They are attractive as they allow statistical inference from relatively small
sample to large population. Survey research includes personal interviews, mail Questionnaire, telephone
interview.
When to use quantitative (survey)? This approach should be used:
 When the subject matter is clearly defined and familiar
 When measurement problems are minor or have been resolved.
 When there is need to relate findings to the broader socio- cultural settings that are already
sufficiently understood.
 When detailed numerical description for representative sample is required
 When repeatability of measurements is important
 When generalization of results and comparison across populations is desired

Comparison of Qualitative and Quantitative Research

Characteristic Qualitative Research Quantitative Research
PURPOSE Describe; Understand Predict causal relationships
LOGIC Why? How? Asked What? Tested
ROLE OF THEORY Inductive Deductive
INSTRUMENTATION Broad theoretical models; post Theory driven hypotheses of
hoc development of relationships among variables
relationships
SAMPLING OF SUBJECTS Availability; purposive; Exclusion/inclusion criteria
theoretical; snow balling established in advance;
random assignment;
representative control group
INTERVENTION Observational only; Carefully designed and
documenting effect of manipulated by researcher
presence in setting
DATA COLLECTION MEANS Immersion; trained Structured interview
observation; in depth instruments with predetermined
interviewing; open or semi response categories
structured tools
DATA ANALYSIS Verbally descriptive and Numerical descriptions and
usually non-numerical statistical
ROLE OF RESEARCHER Must establish relationship with Must be blinded to conditions;
subjects standard interaction with
subjects

Appropriate application of Research methods in Public Health: How do we select which approach to
use? This depends on:
85
  The type of research questions
 The form you want the data to take (Numerical or textual)
 Your data analysis plan
 Your research plan (Amount of staff and time to conduct the research)

d. Rigor in qualitative research:

Definition: Rigor refers to applying procedures and process to increase the quality of data collection and
analysis

a) Probing
Probing is an important tool that helps to improve quality of data collection and analysis. Probes are
devices used to "elicit adequate answers from respondents and to structure and direct the interview". They
have a role in leading the respondent to answer questions more fully and accurately. They also reduce
irrelevant and ambiguous answers. Some of the probes include:
1. Silence probe: A brief pause of about 6 seconds as interviewer looks in face of respondent
expectantly. Silence should not be too long to embarrass the respondent.
2. Assertions of interest or understanding e.g. "I see", "I understand" which means that the
respondent is on right track but more information is needed.
3. Natural phrases/sentences e.g. anything else? What do you mean? Is it so!
4. Repeating the question. This is done if respondent looks hesitant.
5. Repeating the answer: In case the interviewer has not understood properly. This assures both
interviewer & respondent of the answer.

b) Rigor and the concept of Triangulation

Triangulation refers to the combination of at least two or more theoretical perspectives, methodological
approaches, data sources, investigators and data analysis methods. Triangulation promotes validity of
data findings by allowing the researcher to explore a phenomenon more fully whilst facilitating a variety of
methods to encourage comprehensive understanding and explanation. Researchers believe that the use
of different methods will yield a more valid, or at least a more complete picture of the research. [This
concept will be dealt with in detail in another Session]

e. Characteristics of Participatory Qualitative Research in PH

1. Focus of study participants
2. Emphasis on Depth rather than numbers
3. Flexibility
4. Patience and prudence
5. Complementary but not opposed to quantitative approach

86
Session 3: Social Science Methods Appropriate In Public Health

Introduction: This Session will highlight the methods used in qualitative research

Session Topics: The following topics will be covered:

a. The role of Social and Behavioural Science in Public health
b. Social Science methods used in PH Research
1. Why are qualitative methods used in PH research?
2. The Key qualitative methods used in Public Health
A) Focus Group Discussions
B) Key Informant Interviews
C) In-depth Interviews
D) Secondary Data Analysis
E) Participant Observation
F) Narratives
3. Other methods

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Defend the rationale for use of qualitative research in PH
2. Describe the key qualitative methods used in PH, highlighting their characteristics
advantages and disadvantages, and arguing their relevance
3. Set up and conduct a focus group discussion
4. Set up and conduct a Key Informant interview
5. Set up and conduct a participant and non-participant observation
6. Examine social phenomena to discriminate those in which a narrative analysis would be
most relevant
7. Make a summary of other qualitative methods useful in PH

a. The role of Social and Behavioural Science in Public health

Traditionally, social science research draws from qualitative methods. In addition to epidemiology and basic
and applied science, basic principles from psychology, sociology, anthropology and other social science
disciplines can be analysed in relation to causes, consequences and control measures for PH problems.
Rationale for Social Science in Public Health: Many of the issues that threaten the health of the public
have biological and behavioural components. Examples of global PH issues include: HIV/AIDS, war related
morbidity and mortality, Violence and Injury, SARS. When we work with populations, we work with people
whose situation, background and often the culture is different from our own expectation. Social/behavioural
research will help you better understand other people and avoid misunderstandings. Lack of understanding
of normative issues is frequently the cause of PH programme failures

b. Social Science methods used in PH Research

The social sciences employ both qualitative and quantitative research.

1. Why are qualitative methods used in PH research?
 They serve as formative research in designing quantitative studies
87
  To go into more exploratory depth on issues raised in quantitative studies, so as to develop a
deeper understanding
 To complement and triangulate the findings of quantitative research
 Why are qualitative methods used in PH policy?
 To improve participation of programme beneficiaries and health programmes
 To improve cross-cultural communication between programme beneficiaries and programme staff
 To improve quality of programme planning and management
 Range of Qualitative Research Methods used in Public Health

2. Examples of methods used in Public Health include: There are a variety of social science methods
that are useful in Public Health. This depends on the kind of public health issue one is investigating. They
include the following:

1. Focus Group Discussions (FGDs)

2. Key Informant Interviews (KII)
3. In Depth Interviews
a. Informal conversational interviews
b. Interview guides
c. Standardized open ended interviews
4. Secondary data analysis
a. unstructured qualitative case study approach
b. Structured quantitative content analysis
5. Observation
a. Participant observation
b. Clinical assessment
6. Narratives

A) Focus Group Discussions

Focus Group Discussions refer to a qualitative method that gathers people of similar backgrounds or
experiences to discuss a specific topic of interest to the researcher. The group participants who are
homogeneously composed are guided by a moderator who introduces the topic and keeps the discussion
lively. Usually, consensus is sought i.e. participants agree or disagree and a rich description is given.
During FGDs about 8-12 participants are gathered together at a convenient venue. They last approximately
2 hours. Specific topics are addressed to the participants by the moderator and noted are taken by a notes
taker. The proceedings are often recorded. A topic guide is used; this is mainly made up of open ended
questions although some may be closed ended to maintain the flow of the discussion. It is presumed that
the sum total of the group interaction and dynamics produces data cannot be gathered in one-on-one
interviews. FGDs elicit normative data and provide a broad understanding of how issues of interest apply to
the cultural groups or sub-groups represented in the focus group.
Open vs. Closed-ended questions
Open ended questions are as stated open and not restricted in options for answering. They generate a list
and usually explanations. An example is “Tell me about the health problems you have had in the past year”.
Closed ended questions generate a restricted answer, either yes/no or a small range of options and a
little explanation often on probing.

88
Box 2: Advantages and disadvantages of using Focus Group Discussions
Advantages
Quick information at less cost
Obtaining data from illiterate communities
Good at exploring attitudes and opinions
Researcher can be present at session
FGD participatory & well accepted method by the community
FGDs identify relevant and appropriate questions for individual interviews

Limitations
Limited generalisation
Moderator may influence participants
Limited value in exploring complex beliefs of individuals
Errors if there is no homogeneity
Researcher has less control (Flow)
Results are harder to analyze, interpret (avoid lifting comments out of context and sequence)
Limited number of themes (1-2 hrs)
Recording and transcribing difficult

B) Key Informant Interviews

An interview between a key informant and an interviewer proceeds flexibly, much like a dialogue. The same
key informant is usually interviewed several times, to discuss certain issues in depth. These can be issues
raised in the first interview or earlier in the study. The questions are open ended and the interviewer makes
an active effort at building rapport with the key informant. The interviewer can use an interview guide, but
does not need to follow the guide exactly. Guides can be developed and revised based on information
gained during earlier interviews or during earlier phases of the study. Usually, the interviewer explores
topics on the guide or related topics as the informant brings them up during the interview – there is iterative
flexibility.
Who is the Key Informant? KIs are individuals that the research interviewer comes to trust as
honest, reliable and forthcoming. They should be capable of providing in-depth information about
the topic of interest. Key Informants share basic characteristics of the intended participant
population and may be interviewed several times. Key informants are people who, because of their
position or experience, have greater knowledge of what is being investigated than the average
person.
Advantages and Disadvantages: This method of data collection is associated with several
advantages namely; small sample, have access to culture & values, have time & willingness to share, non
analytic, key informants develop relationship to researcher. However, some limitations have been notes
i.e. difficult to identify, may take advantage for political gains, and may insight false hopes to community

C) In-depth Interviews
These are face to face encounters between the researcher and the informant directed towards
understanding informant’s perspectives on their lives, experiences or situations as expressed in their own
words. As implied in the name, they have the advantage of long, one to one interaction which gives an
opportunity to hear in detail from one individual thereby giving a more complete picture of the experience
and context. In-depth interviews are open ended and flexible in nature and this allows the interviewee to
tell the story in their own words.

D) Secondary Data Analysis

This involves collection of data by reviewing documents. Documents can be medical records, reports etc.
One can generate quantitative or qualitative information. Advantages are accessible subjects (whether
living or dead), Non reactivity, longitudinal analysis (Documents can be studied retrospectively); low cost
89
compared to surveys, experiments and visit the author or interactive electronic communication system.
However there are limitations associated. These are, mostly documents are recorded for clinical,
administrative purposes, degree of bias if the documents have monetary implications e.g. taxes,
Incompleteness, non-availability, lack of standard format, coding difficult as documents are written for
various purposes, documents can not be supplemented with observation data, absence of information does
not mean absence of the phenomenon, often hand writing is difficult to read.

E) Participant Observation
It is a careful systematic observation of social and cultural events and activities. This is followed by
systematic recording of these observations. Quite often what is not said is as important as what is said.
Participant observation (PO) brings the researcher into direct interaction with the people and the activities.
It involves getting close to people and making them feel comfortable enough with your presence that you
can observe and record information about their lives. PO is an approach rather than a data collection
technique. It involves mainly unstructured observation and unstructured interviewing.

Advantages
Allows contextual data
Facilitates coop. with study subjects
Facilitates immediate data follow-up
allows wide range of data
data collected in natural setting
Allows discovery of complex relationships
Obtains data from non verbal behavior/communication
Facilitates analysis validity checks & triangulation
Discovers subjective perspective of organizational processes

Limitations
Cultural differences can cause data miss interpretation Dependent of small groups (KIs)
Ethical dilemmas
Difficult to replicate
Data subject to observer effect
Process can cause danger & discomfort to researcher
Dependent on researcher honesty& responsibility

F) Narratives
The narrative research approach has recently gained the attention of researchers in health professions as it
is perceived as an opportunity to tap into respondents’ or patients’ experience. A narrative is a story that
tells a sequence of events that are significant to the narrator and his or her audience. It has a plot, a
beginning and an end. The process of narrating experience is always set within a historical and temporal
frame that the teller brings to the story. The development of research narratives should therefore be
understood as a form of reflection upon an event. Every narrative describes a sequence of events that
have happened. The method is also described as an active (re)construction of events and experiences
which the narrator decides to include and tie them in a certain manner.
Case or event Narratives: It is a type of key informant interview used to record the actual
activities surrounding an event such as an illness, a birth or death. The interview focuses on the sequence
of events such as when the symptoms appeared for an illness, and what decisions a pregnant woman
made in preparation for a birth. It is a detailed in-depth itinerary of an entire story of events as they unfolded
about a particular main event of interest. The transcript for this interview may be as long as 60 pages. The

90
interview focuses more on understanding the experiences of one person rather than those of many – it is
therefore not uncommon to have only one or two people interviewed.

3. Other methods: There are so many other methods used in qualitative research. Most of these methods
are used in participatory research and rapid appraisals. A few of these methods will be highlighted later.
[These methods will be further expounded on in the Session: Rapid Appraisals]. A few are described below:
Participatory or social Mapping: It is a method of collecting geographically related information
such as water sources, health centres, markets, homes of key persons, landmarks, fields and schools and
representing them spatially. It is commonly one of the first activities done by a research team to ‘break the
ice’ and familiarize a community with your study. Information gathered helps the researchers (especially
when they are not familiar with the surroundings) to find their way around, plan activities and select
research participants.
Free-listing: An informant is asked to list all of the different components of an issue of interest (for
instance, all the different illnesses that children get or the main difficulties faced by persons in the
community. This method is used as a preliminary exploration of the issue of interest and is often carried out
at the beginning of a study.
Pile sorting: It is used to understand how issues of interest and their components are interrelated
in a culture by allowing informants to group together items (e.g. Childhood illnesses, community problems
according to their own categorizations.
Timeline (History line): It is used to gather time-related information such as the sequence for key
events in the history of a particular population. Informal groups of people, knowledgeable about the history
of the population under study are asked to use locally available materials to mark the events.

EXERCISE 2.2.1: AEFI Study

The Centers for Disease Control is planning to conduct a survey on perceptions Adverse
Events Following Immunisation (AEFIs) and how they affect future willingness to go for
immunisation. Though there is a general rough idea on the attendant phenomena, they are
not fully mapped and counted. They need a local contractor to undertake 3 phases of the
study as follows:
Stage 1: Qualitative study using the WHO 30 clusters of 7. The aim is to get an
understanding of the emergent socio-cultural phenomena, so as to use them to
design the survey questionnaires
Stage 2: Pre-testing the questionnaire and Iteration
Stage 3: Survey and Follow up survey for a sample of 1080 house holds
You have received the contact for Stage 1; describe how you would:
a. Set up and conduct a focus group discussion
b. Set up and conduct a Key Informant interview
c. Conduct a detailed debriefing (narrative) of one mother who had a “near miss”
experience with and AEFI

91
Session 4: Rapid Assessment Procedures

Introduction: When planning social interventions, we often need to assess the requirements of the target
community, so as to design relevant programmes. It may therefore be necessary to conduct a baseline
needs assessment. However, conducting a detailed empirical study may not be feasible; all we may need is
a rapid assessment of community needs – this assessment should be participatory, so that the
communities are involved in prioritisation. PRA tools have been developed just for this purpose. The beauty
in PRA tools is that they provide easily understood and simple methods of involving communities – they are
as diverse as they are interesting. This Session will consider the various methods available for the Public
Health professional.

Session Topics: The following topics will be covered:

a. What are Rapid Assessments?
b. Rapid Epidemiological Assessment (REA)
c. Participatory ‘Rural’ Appraisal and Community Diagnosis
d. Some individual PRA techniques

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Define Rapid Assessments and explain the rationale for Rapid Epidemiological
Assessment
2. Summarise the features and key considerations of PRA and appraise its advantages
3. Synthesise, facilitate and conduct an effective PRA debriefing interview for an important
health intervention
4. Paraphrase the features of at least 5 PRA techniques so as to demonstrate the
simplicity and wide methodological variability of PRA

a. What are Rapid Assessments?

Rapid assessments procedures are a quick way of collecting information through the use of focus group
discussions, Key informant interviews, case studies and observations. The data is quickly analysed for
appropriate planning/interventions. The desire to develop health education and other interventions to
improve programme effectiveness, and ensure compliance and acceptability has resulted in the demand for
methodologies and techniques which provide highly specific social and cultural information. Concomitantly
there has been a demand for cost effective and timely tools for program evaluation. Consequently there
has been recent interest in methods for rapid or focused data collection, resulting in the development of a
number of formal rapid assessment procedures including those of rapid rural appraisal (RRA) Rapid
Epidemiological Assessment (REA) and Rapid assessment Procedures for anthropological studies (RAP)
[Manderson and Aaby, 1992]. Rapid Assessment Procedures are important for operation research in
collecting baseline and evaluation data. These research procedures collects highly focused sets of
information. Fieldwork in done within limited duration and different methods of qualitative and quantitative
data gathering are used.

b. Rapid Epidemiological Assessment (REA)

REA were developed due to a need in epidemiology to develop techniques which yield highly reliable
quantitative data basically for purpose of programme evaluation. They started with the WHO programme
92
(EPI) and resulted into cluster sampling technique (30x7). They arose out of the need to adapt standard
epidemiological methods to collect information quickly and cheaply to plan new programmes and monitor
those in place

c. Participatory ‘Rural’ Appraisal and Community Diagnosis

1. What is PRA? Participatory Rural Appraisal is a collective learning process with partners using a series
of flexible participatory techniques to analyse a situation, carry out needs assessment, identification of
priorities for development activities, and to plan activities in general (etc.) in a systematic and intensive
fashion.
It has been used to help development workers better understand local situations and plan solutions
to problems with their local partners. It lays emphasis on participation of community members,
‘beneficiaries’ in identification, planning (and where required, ideally the whole project cycle). The
process not only contributes to capacity building and empowerment but also sustainability of
projects. It is mainly applicable to action-oriented research but could also be used for other kinds
of research.

Advantages of PRA:
It entails high community participation, a bottom-up approach; is informal and non hierarchical,
individual contribution is important.
Short duration
Low cost
Allows for flexibility and is normally not superficial;
Allows for progressive learning during fieldwork.
little or no statistical analysis is applied
Multi disciplinary team (ideally)
It is best for learning and understanding rural people’s opinions, behaviours, and attitudes.
PRA is mainly qualitative in nature; employs a variety of tools; formal questionnaires are avoided.

Dangers and Short comings of PRA:

Difficulty in finding the right team
Rushing may create superficiality
Difficulty in identifying the marginalised
Failure to involve the community
Lack of rapport with the community
Over generalization based on too little information etc

2. Features of PRA:
Triangulation is done in relation to:
 Composition of the team
 Sources of information
 The mis of techniques (tailored reinforcement of the study)
PRA is mainly qualitative in nature: It employs a variety of tools and formal questionnaire are avoided

3. Considerations when undertaking PRA:

 Allow for flexibility
 Avoid unnecessary detail and over collection of information
 On the spot analysis – analysis concurrent with the data collection
 Endeavour to include the marginalised; the quiet and the remote – to counter bias
93
  Learn progressively as events unfolds
 Share findings with participants
 Link the research with action

4. PRA Techniques
The major research tools are: Semi-structured interviews, Focus Group Discussions and key Informant
Interviews. However, the techniques are as diverse as possible. They include:
- Secondary data reviews, Direct Observation, Semi-structured Interviews, Group discussions
- Ranking: This includes – Preference ranking, Pair-wise ranking, Direct Matrix ranking, Ranking by
Voting, Wealth Ranking
- Analysis Group discussions, Innovation Assessment
- Mapping: This includes – Participatory mapping, Historical mapping, Future (visioning) mapping,
Mobility mapping, Social Resource Mapping and Transect walks
- Seasonal calendar, Historical seasonal calendar
- Time trends, Historical Profile, Livelihood analysis
- Flow causal diagram, Systems diagram, Venn/Institutional diagram, Potential Partner matrix, Pie
chart, Histogram
- Participant observation, , Workshops, Group walks
- Stories, Oral histories, Case studies and portraits, Proverbs

5. Key issues about facilitation and interviewing: The facilitators guide the interventions and are not
supposed to dominate. They should work towards:
 Promoting active participation that is non-directive
 Avoiding being rude or taking the “expert” role
 Ensure that all group members participate
 Probing and going below the surface of things
 Communication: Consider verbal and body language. Try to match wit the participants, be friendly,
and where feasible, sit in a circle and at the same level, so that there is full view between
participants; however, sit at the same level, so that the focus in on themselves and not you.
 Seek to motivate participants and not to manipulate or impose your ideas
Don’ts: Do not give orders, threaten, moralize, criticize or advise
Dos:
- Passive listening
- Acknowledgements
- Door openers and ice breakers
- Active listening
- Paraphrasing
- Use drawings where possible; use locally available materials
- Interview the diagrams, not the drawers
- Beware of the differences between men and women
6. Principally, PRA favours the following attitudes:
1. Participation
2. Respect of community members
3. Interest in what they know, say, show and do
4. Patience
5. Listening and not lecturing
6. Humility
7. The methods used result in community empowerment to express, share, enhance and analyse
their knowledge
94
7. Interview skills: In addition to the above
 Be prepared
 Introduce your self and the purpose of the meeting
 Watch your body language throughout
 Start with general questions/issues (to put people at ease) – ask one question at a time
 Mix questions with general discussion – make it a casual dialogue
 Avoid using closed ended questions, leading questions, complicated and offending questions; use
probes
 Use diagrams, symbols and other drawings
 Use simple language – avoid scientific words
 Restrain from interrupting; observe
 When interview is over, allow for questions
 Finish politely by thanking the participant
 Make full notes of the interview

d. Some individual PRA techniques:

1. Review of secondary sources: Here, we review published or unpublished reports, statistics, maps, etc.
It helps identify existing data, trends and gaps. However, do not spend a lot of time for this at the expense
of field work
2. Transect walks: It is an observational walk paying attention to people, activities, resources,
environments and other specific issues of interest. Decide on information needs and take notes on
observed issues during the walk. After the walk we may draft a crossectional spatial diagram. It has the
advantage of being highly participatory and relaxed; it enhances local knowledge and can be used by
illiterate people. It is used in validating findings of participatory mapping. Its weakness is that it may be time
consuming and requires graphic skills.
3. Semi-structured Interviews: This is one of the main PRA tools. A checklist is used but flexibly. It is held
with individuals, Key Informants, Groups and Focus Groups. They are less intrusive than a questionnaire
and encourage two-way communication. They are easier to administer and collect more enriched data.
They are relatively open and flexible. Their weakness however is in the need for experience and sensitivity
on the part of the interviewer. They require literacy, good communication and summarising skills. They also
require a grasp of the topic to be addressed.
Key Informant Interviews: They are conducted with people that have special knowledge on a
particular topic. Outsiders that live in a community are valuable informants.
Focus Group Discussions: They address specific topics in detail. They involve 6 to 12 people.
They have a facilitator and a notes taker. Their strength lies in the fact that interaction enriches quality and
quantity of information; they do exhaust the possible suggestions, experiences and scenarios. They are
therefore good at disclosing the range and nature of the problems and sometimes the solutions. Their
weaknesses include the fact that practice and experience is needed to moderate them. They require skill in
extraction and summary of findings.
4. Resource map: It is a map of the village or community with resources such as roads, schools, health
units, churches and fields. It is useful for discussion of local resources, their distribution, problems, and
opportunities. All maps are correct.
5. Service Maps: They are useful for analyzing what services are or are not available in an area. Their
strengths lie in their ability to give a broad view of the overall situation on the ground, encouragement of
two-way communication, helping the participants to see links, patterns and interrelationships in their
territory and allowing illiterate people to participate. Their weakness lies in their being subjective. They may
also be time consuming.

95
6. Preference Ranking: It is a quick method used to identify preferences of the group. It takes several
forms:

a. Modified Simple Preference ranking: It is similar to prioritizing in VIPP. The individuals vote on
each and allocate them their score, from which a total is computed and a rank.

Illustration:
Items to be ranked Score Rank
(Problems, institutions etc.) A B C D E
Safe water
Health Centre
Church
Etc…

b. Pair-wise ranking: This is useful in analyzing reasons for particular choices (The maximum of
options should be 7)
Illustration:
a. b. c. d. Weight Favourable reason Unfavourable reasons
a. a. a. a. 3
c. d. b. 0
d. c. 1
d. 2

c. Wealth ranking: Poverty and issues of human development are central in public health.
Communities are not homogenous but are differentiated especially on wealth lines. Wealth ranking is used
to identify the criteria that local people use to define wealth and poverty and how one becomes richer or
poorer in relevant communities.

Strengths of ranking in general: It is flexible and can be used in diverse situations and settings. It is a
close alternative to closed ended questions. It is interesting and amusing to the participant, and generates
commitment to action research. It collects information on choices and related reasons.
Weaknesses: Pre-testing may be required. Choices may be affected by highly subjective factors, therefore
requiring proper sampling.
7. Venn Diagrams (Also called Chapatti Diagrams) and Institutional Analysis: They are useful in
discussing community structures, their relationships, linkages and potential usefulness in contributing to
development activities.
8. Potential Partner Matrix: It is a more direct situation analysis. An example of a matrix is tabulated
below:

Illustration:
Potential Partner Strengths Weaknesses What can they Offer?

9. Problem Decision Trees: It is a dichotomous or multiple cascades of events from the root cause
(Roots) to the proximate causes (Trunk) to the underlying causes (Branches).

96
EXCERCISE 2.2.2
1. Define Rapid Assessments and explain the rationale for Rapid Epidemiological
Assessment; summarise the features and key considerations of PRA and appraise its
advantages and disadvantages
2. What key considerations would you put in mind when moderating a PRA debriefing
interview for an important health intervention?
3. You have been selected to moderate a PRA session in Kaberamaido District in which
you are to select priority interventions that will be the focus of a new World Bank funded
programme to improve long term child survival. These strategies must be integrated in
the available institutional framework. Describe 5 PRA tools that you would strategically
employ to optimise the community’s contribution. Your participants are the PDCs

97
Session 5: Design of Qualitative Data Collection Tools

Introduction: A key feature of qualitative data collection is that the information collection process changes
as emergent themes develop. Therefore, the tools used in qualitative data collection are not rigid – on the
contrary, they are used merely as guides on general topics of discussion. The finer details emerge as the
data collection process goes on. Sometimes the detail has to be elicited with the use of probes. This
Session will therefore highlight key considerations in the design of qualitative data collection tools.

Session Topics: The following topics will be covered:

a. Introduction: Recap of key Qualitative Data Collection Methods
b. Tools used for Qualitative Data Collection
c. Structure of Qualitative Data tools
d. Pattern of Qualitative Data Questions
e. A Matrix for development of Qualitative Questions

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Describe the general characteristics and structure of qualitative data collection tools
2. Explain the rationale for the framing of Qualitative data collection questions
3. Design and use a KI Guide for investigating a social intervention of PH importance
4. Design and use a Focus Group Discussion Guide for investigating a social intervention
of PH importance

a. Introduction: Recap of key Qualitative Data Collection Methods

 In-depth interviews: They deal with very few or one respondent at take a very detailed history of
what transpired, often in form of narratives that go through an itinerary of events as they unfolded
to the end. They are very useful for investigating the quality of care, or in case of dramatic events,
like maternal mortality, they can give a detailed picture of the delays at different stages of the
referral process
 Key Informant Interviews: They deal with individual experts who hold special positions and can
be given specific information on critical issues
 Focus Group Discussion: They deal with 6 to 12 persons selected according to homogenous
characteristics e.g. Gender, socio-economic characteristics, Education level and work experience.

b. Tools used for Qualitative Data Collection

They are referred to as a Guide and may be specified as:

- Focus Group Discussion Guide
- Key Informant Guide
They are guides because the nature of the discussion may change depending on emergent themes. They
can be altered accordingly. They provide a list of topics that will be covered and serve as a memory guide
to the interviewer or the moderator.

Points to note about qualitative data collection tools:

 They are planned in Advance
 They are used to identify issues, terms, perceptions and interpretations
98
  They cover a few general topics linked to specific objectives
 Theme areas are highlighted (most often the objectives) and a few questions identified under each
of them
 Unstructured interview guides are better as they allow respondents to identify and discuss issues
freely, and to express their opinions and views
 They should lend themselves to triangulation (have similar questions and themes to other tools)
 Debrief after the session with respondents and your team if nay

c. Structure of Qualitative Data Collection Tools

1. A Question should be:

 Specific
 Informal
 Non-judgmental
 Open ended
 Not suggestive of any answers
 Following in logical sequence from easy to complex and more personal
 As detailed as possible, thinking through all possible answers

2. Structure of the Data Collection tool

1. Key Identifiers – Title, type of tool, date, location, type and number of participants, duration of
interview and discussion
2. Introduction – Greeting, welcome, researcher introduction, explain goal and purpose of interview,
informed consent, request for participation and recording; ask participants to introduce themselves
(may or may not use names)
3. Questions – highlighted themes, main questions and probes. Relate to specific objectives
4. Conclusion/Closure
 Provide opportunity for participants to ask questions and express appreciation to
them
 The moderator or interviewer summarizes impressions, points or agreement and
disagreement from the discussion and interviews
 The participants then confirm, clarify and elaborate on these impressions of the
researcher

d. Pattern of qualitative data Questions

They are in three stages:-

1) Main Question:
i. It should be easy and non-threatening, to help establish rapport
ii. It should never yield a “Yes/No” answer only
iii. It should inspire continued interest
2) Follow-up question: It moves the interview into a deeper level by asking for more
information after the main question
3) Probe: This is also a form of follow-up question which elicits further information, views,
opinions and perceptions with or without specific reference to the topic e.g.
a. Can you further explain what you mean?
b. Tell me more about that
c. How do you feel about that
d. Please share with us your experience
99
 Always probe on why or why not on an issue. Do not be too aggressive when probing, as it
may be intrusive. Probes should be flexible and the interview should be open to asking
other probes that are not necessarily in the guide but stem from emergent issues in the
discussion. They are therefore the key iterative tools.

e. Example of a Matrix for development of Qualitative Questions

E.g. Subject: Effects of Malaria Medication

TOPIC MAIN QUESTION FOLLOW UP PROBE QUESTION
QUESTON
Knowledge Can you tell me what How did you come to When did you come
(Simple) you know about the know these side to know these side
side effects of Malaria effects effects?
medication
Source of Information Where did you get this What specific
information? information was
given?
Experience Have you ever What action did you Did your actions
(Personal and experienced any of the take against these work?
Complex) side effects you side effects
mentioned?
Opinions, What do you think
perceptions, benefits doctors do to reduce
and barriers some of these side
(Personal and effects?
Complex)

ACTIVITY 2.2.1: Refer to the Additional Resources folder for the resource Examples of
Qualitative Data Collection Tools for a glance at a typical KI and FGD Guide. You should now
be able to use both quantitative and qualitative methods in your Field Studies

EXERCISE 2.2.3
1. Design a KI guide that you will use in the CDC survey in Exercise 2.2.1; what main
sections did you consider?
2. Design an FGD guide that you will use in the CDC survey in Exercise 2.2.1; what
considerations did you put in mind in paraphrasing the questions?
3. In a quinquinneal social-behavioural survey on behavioral change to prevent HIV/AIDS,
it was observed that many women respondents withheld vital information regarding
fidelity and whether they ever contracted an STD. Describe how the strategic use of
probes would elicit the required information

100
Session 6: Qualitative Data Analysis

Introduction: Qualitative analysis differs from the quantitative one in that in the former, analysis begins
from the initial phase of the study is data collection up to report writing. It is important to note that
qualitative studies aim at understanding broader contexts; meaning of ideas & concepts take shape during
data collection; qualitative analysis begins with field activities and may lead to revisions / refinement in
research questions

Session Topics: The following topics will be covered:

a. Characteristics of qualitative data analysis
b. Principles of Qualitative Analysis
c. Basic Steps in Qualitative Data Analysis
d. Stages of Qualitative Data analysis
e. Fourteen stages of Qualitative data Analysis

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Paraphrase the key characteristics of qualitative data analysis
2. Describe in detail the principles underlying the rationale of qualitative data analysis
methods
3. Contrast the method of Content analysis with thematic and semiotic analysis and
elucidate their application in Qualitative Data analysis
4. Describe the processes involved in the 14-stage qualitative data analysis process,
evaluating their rationale
5. Employ open, axial and selective coding in qualitative data reduction for a given
qualitative data transcript
6. Use a master coding sheet to tally qualitative data themes and prepare a report

a. Characteristics of qualitative data analysis

 Analysis is not in the last phase of the research; it is concurrent with the data collection. The
process of analysis is systematic and comprehensive but not rigid.
 Data units are classified according to an organizing system that is predominantly derived from the
data itself.
 Categories and codes are tentative and can be changed as the research progresses
 The main intellectual tools are making comparisons and examining for outliers
 There is no “right” way to manipulate the data and no “right” results

b. Principles of Qualitative Analysis

 People differ in the experiences and analysis of reality – how participants define the question may
not reflect assumptions made by the researcher
o Learn to become aware of your own perspectives during the data collection process
o Note what is emmerging as explanations and check it against raw data
o Be open to surprising findings
o Always be ready to return to your study participants to understand better what you have
learnt
101
 o Actively seek out alternative explanations
o The important issue is not necessarily the commonest explanations – the important issue
is the wide range of observations that exhausts the suggestions and is sensitive to outliers
o It takes patience and perseverance to understand the world of your participants
o It is important to work closely with Research Assistants and Key Informants as they
understand the context and need to be sensitive to the changing detail.
 A social phenomenon cannot be understood outside its context
o Context is the physical, social, political and organizational/ individual characteristics that
affect or influence behaviour
o It is bought into analysis by considering how informants stories are shaped by their social
position, economic opportunities, religious convictions and how these are situated in
broader contexts above
o Through analysis of information from many sources, you will identify ways in which
different groups’ views converge or diverge and contextual factors leading to those views
 Theory both guides qualitative Research and is the result of it
o Qualitative analysis can be informed by theory or it can generate theory
o Qualitative data leads to a rich textured analysis and always the possibility of new
theoretical understanding
 Exceptional cases may yield insight into a problem or new leads into further inquiry
o Although analysis seeks common ground and consensus from all groups or individuals, it
is important to understand why differences occur
o Tracking exceptions is important
o At times contexts and circumstances cause these differences
 Understanding human behaviour emerges slowly
o A flexible and integrated approach is essential
o Beginning analysis as you collect data is as good as you start questioning and re-phrasing
for emphasis
 Reading: There is need to develop intimacy with the data
o Reading for content
 Are you obtaining the king of data and responses you expected
 Are the responses full and details or they are superficial
 Other aspects of the interview process (sitting, venue, composition etc.)
o Noting quality
 Check quality of transcripts e.g. non-verbal interactions etc.
o Identify patterns
 Identify patterns from different sources e.g. different kinds of participants
 Make Iterative changes
o You need to modify the data collection tool as you make observations
o This will help you to capture issues that were previously not speculated on

c. Basic Steps in Qualitative Data Analysis

The process involves 4 steps:

1. Coding: This involves identifying emmerging themes. They are just words that act as flags
Example: <Treatment seeking> can be <TRITSEEK>
<Pregnancy signs> can be <PREGSIGNS>
2. Labelling: This involves attaching labels and displaying the data
3. Summarising: Data summary in a matrix

102
4. Data Presentation: Interpretation and presentation of the information. This involves commentary text,
text quotes, taxonomies and flow charts

d. Stages of Qualitative Data analysis

The first steps: Identifying units of analysis

As in sampling, one of the first steps in qualitative data analysis is the identification of units of
analysis. Lofland and Lofland (1971, pg 71) describe a unit of analysis as “a tool to use in
scrutinizing your data”. The unit may be meanings, practices, encounters, narrative structures,
organisations or lifestyles. The unit will be usually decided on after analyzing previous research
on the topic, the theory to be used, and the issues for which the research is designed. Content
analysis, grounded theory and semiotic analysis all require a careful identification of which units
will be analysed. The focus group and not the individual participants may be the unit of analysis,
they for the key Informants, the individual KI may be the unit, For a case narrative, each
milestone may be a unit of analysis.

Content Analysis: Content analysis is different from Thematic Analysis (or Grounded Theory)
on the basis that it involves the identification of codes prior to searching for them in the data. The
principles are according to Kellehear (1993a):
 Develop categories, prior to searching for them in the data
 Select the sample to be categorized
 Count or systematically record the number of times the categories appear.
It is usually an initial step that leads to another type of qualitative analysis. An analytical coding
structure can be developed in a hierarchical order, so that participants’ responses are coded
against this structure. Hierarchically arranged and codes are prerequisites to content analysis.

Thematic Analysis (Grounded Theory): The major difference between thematic analysis and
grounded theory is that grounded theory includes theoretical sampling while thematic analysis
does not; the latter is therefore a subset of the former. According to the methods of grounded
theory, concepts, categories and themes are identified and developed while the research is
being done. The concepts are the basic units of analysis; these are them grouped and related to
form more abstract categories. Relationships between categories are then identified to develop
themes or what is referred to as “formal theory”. Additional constructs can be examined and new
codes added as new themes emerge. It allows the researcher to examine what different people
have said about various topics. As new themes emerge, all the relevant information is retrieved
and examined for further coding designations.

The practice of Grounded Theory: At the heart of grounded theory and thematic analysis is the
process of coding, sorting and organizing data. Whether done on paper or computer, analyzing
qualitative data involves coding pieces of text (Chunks) that are coherent in theme and then
collating all those that are coded in the same way – coding is clearly central to the analysis
process. Straus and Corbin (1990) describe three main coding procedures: Open Coding, Axial
Coding and Selective Coding.
Open Coding: It involves comparisons between events, actions and interactions. The
analyst searches for differences and similarities between events, actions and interaction and
applies conceptual labels to these, grouping them into categories. We tear apart and
observation, sentence or paragraph and give each discrete incident, idea, point, topic or event a
name that stands for a phenomenon. It is the first run at coding data. It aims at looking at the
data in new ways and angles to see new relationships between events or interactions and to

103
 develop new ways of dealing with these relationships. Miles and Huberman (1994, p.148)
describe a number of coding processes that are consistent with open coding including: Noting
patterns and themes and making metaphors. Metaphors are data reducing devices that
summarize a number of different aspects. They are pattern making devices that suggest new
relationships as a consequence of new imagery. They decentralize data and provide new
perspectives of the same set of themes and suggest different questions about data. “Their
inherent ambiguity is useful for avoiding preconceptions and positivistic assumptions of a fixed
world governed by preset laws”.
Axial coding: If open coding attempts to break down data and re-conceptualize it, axial
coding is about putting the data back together in new ways by making connections between
categories and making sub-categories or collapsing repetitions
Selective Coding: It is the process by which the emergent categories are unified around a
core category. The processes are similar to axial coding, but at a higher level of generalisation.
This provides the theoretical point of integration of the study.

Keeping a journal or research memos: It is very important for the researcher to make notes on
emergent issues and notable observations throughout the study. It is separate from all other
forms of recording e.g. the field notes, interview transcripts or files used in coding data. These
ideas may contain emergent themes that will be developed fully at the opportune time and may
be the guide to selective coding or forming theory. Keeping a journal facilitates thinking about
and beginning to write about the analysis and interpretation of qualitative data.

Use of computers in Qualitative Data Analysis: Computers are increasingly in use for
analyzing qualitative data. They assist in the process of coding and can be handy in cutting and
pasting categorizations under central themes. In general, they simplify the process. However,
the qualitative computer packages do not do the analysis – the only one that can detect
emergent themes and make sense of categorizations is the researcher. The computer merely
facilitates the permutations and combinations. Examples of packages include: EZ text (Free
software), Ethnograph and QSR Nvivo (For axial coding).

e. Fourteen stages of Qualitative data Analysis

These stages are routed in thematic analysis (grounded theory).

Stage One: Notes are made after the interview on topics that were talked about during the interview
Stage Two: Transcripts are read and noted are made through reading the general themes in the transcript
Stage Three: Transcripts are read again and as many headings as possible are generated to describe the
content (Open coding)
Stage Four: The emmerging categories are grouped in order ad some categories are collapsed together to
reduce their number
Stage Five: The new list of categories and sub-headings is worked through. Repetitions and similar ones
are harmonized
Stage Six: Two colleagues are invited to generate category systems, independently without seeing the
researcher’s list. The three categorizations are then discussed and adjusted
Stage Seven: Transcripts are re-read alongside the finally made list of categories and sub-headings to
estimate the extent to which categories cover the aspects of interviews. Adjustments are made
Stage Eight: Transcripts are read with lists of categories and subheadings and coded accordingly
Stage Nine: Each coded section of the interview is cut out of the transcript and collected together.
Stage Ten: The cut out sections are pasted against appropriate headings and subheadings
Stage Eleven: Selected respondents check the appropriateness (Does this quotation fit the category?)

104
Stage Twelve: All sections are files together for direct reference when writing up the findings. Audio
transcripts and notes written during interviews are filed together
Stage Thirteen: When sections are together, the writing begins; there is selection of examples from
various data under some sections.
Stage Fourteen: The researcher decides whether to link examples and the commentary with the literature
e.g using verbatim illustrations that give a feel of what was actually said in the data collection context.

The Master Coding Sheet Analysis Technique: It allows for logging all data on one sheet. It facilitates:
 Direct scan at a time
 Comparisons and outlier identification
 Intra and inter-group observations and comparisons
The method is easy to use and practical. It is also inexpensive.

Example of a Master Sheet Analysis Tool

Theme Male Female Totals

Why adolescent Urban Rural Urban Rural Total Resp. Total FGDs
Pregnancy FGD FGD FGD FGD
Lack of Money III I IIII IIII 12 4
Etc…Etc…

(For a diagrammatic presentation, refer to Ulin Priscilla et al 2002. P 139 fig. 6.1)

105
Session 7: An In-depth view of Triangulation

Introduction: Triangulation is a means of achieving rigor in qualitative research and is a tool often used to
enhance the internal validity of research (both qualitative and quantitative). It is an important bridge for
eliciting complimentarity for qualitative and quantitative research. In this Session, we shall examine the
principles of triangulation and further elaborate the method of methodological triangulation. It is important to
note however, that triangulation is not a scapegoat for trashy research methodology. Wrong methodology
cannot be salvaged by triangulation.
Session Topics: The following topics will be covered:
a. The paradigmatic approach to research
b. Qualitative and Qualitative techniques
c. Methods of Triangulation
d. Limitations of triangulation

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Distinguish the Rationalistic from the Naturalistic paradigm so as to appropriately link
them to qualitative and quantitative methodologies
2. Differentiate the types of triangulation and illustrate how they can be applied in a real
time social intervention
3. Evaluate the method of triangulation to highlight its limitations

a. The paradigmatic approach to research

Multiple Triangulation: It is an approach to research which combines various techniques. Before delving
into this approach, it is important to consider two paradigms that under-lie all research.
1. The Rationalistic paradigm: It stems from the branch of science known as positivism or
empiricism. It was the product of a group of philosophers that came to be known as the Vienna
Circle in the 1930s. They viewed man in empirical measurable terms and were interested in the
truth. They believed every variable in man could be allocated a number and counted. They strove
to explain nature through observation, the testing of scientific hypotheses and the development of
theories. The rationalistic approach is a reflection of these assumptions; it is a search for rules that
govern or explain reality. This type of research would allow health professionals to predict and
control behaviour.
2. The Naturalistic paradigm: These theorists share their assumptions with the most recent view of
science. These scientists view science as the process of understanding human behaviour. They
look at the individual as a whole and believe that reality is multiple, interrelated and determined by
context – not randomly. This approach assumes that individual share varying perceptions of a
given situation due to their history, which the world is constantly changing, that individuals interact
with the environment and that individuals interpret their experiences and make choices. This
approach has developed many branches, which the student should know about; the include:-
a. Ethnography
b. Phenomenography
c. Hermeneutics
d. Grounded theory

106
b. Qualitative and Qualitative techniques

From the above paradigms, there are two distinct methods of research which are known as quantitative and
qualitative research.
The Quantitative method searches for quantifiable or measurable data. The researcher remains
detached from the facts to prevent bias and attempts to control the events by isolating variables and
implementing controls. The method often involves reducing the subject matter under investigation into
parts, believing that learning about the parts will reveal facts about the whole.
The Qualitative methods relate to description and interpretation of what is actually happening.
Here, the researcher becomes more involved in the work. They are concerned with the dynamic nature of
the reality and attempt to achieve a holistic view of what is occurring, including subjective data. The
researcher is flexible, using whatever techniques they can to represent a picture of what is actually
happening.

Data produced in quantitative research lends itself to the rationalistic paradigm while that in qualitative
research lends itself to the naturalistic paradigm – some literature there fore uses these terms
interchangeably – this however is misleading to the new researcher. The new researcher tends to use both
approaches, to gain better understanding of phenomena. Data from quantitative measurements can be
compared with data from qualitative research to reinforce findings – this is called triangulation.

c. Triangulation
There are 5 main types of triangulation
1. Data triangulation (Sources triangulation)
2. Investigator (Analyst) triangulation
3. Theoretical (Perspective) triangulation
4. Methodological (Methods) triangulation
5. Multiple triangulation

1. Data triangulation (Sources triangulation): We use multiple sources of data that all have a similar
focus. The different approaches will increase the likelihood of discovery of negative cases. An example of
this is collection of data from health workers and patients, on their perception of quality of care in a given
hospital. The different perceptions can then be amalgamated
2. Investigator (Analyst) triangulation: More than one investigator is used in the study. This reduces the
potential bias presented by a single author.
3. Theoretical (Perspective) triangulation: This involves the use of several hypotheses or frames of
focus within the same data. This produces amore convincing result if the same data are tested against
other theories
4. Methodological (Methods) triangulation: This is the most commonly used type of triangulation. This
involves the use of more than one method of data collection. The researcher may use one method of data
collection in several different ways e.g. in a study of compliance to ART, one researcher used 4 scales that
included Clinic based pill counts, 30 day pill counts, 3 day recall and serological measurements of drug
concentration in blood. This is called “Within method triangulation”. It tests the reliability (repeatability) of
data. On the other hand we may collect data using several techniques and assess for commonalities and
points of dispersion – this is called “Across methods triangulation”. It increases the validity. It may include a
structured quantitative interview, key informants, and FGDs, all in the same study.

A combination of “across method” and “within method” triangulation is known as “between method”,
“holistic”, or “contextual” triangulation. It involves the use of several independent methods and it assumes
that if they are separate, they are not prone to weaknesses.

107
Regarding its merits, methodological triangulation is seen as an opportunity to enrich research findings and
deepen insight; in addition, findings from one method can be validated by another method. In
methodological triangulation, the implicit assumption is that the methods used are of equal epistemological
value and internal validity. Triangulation is therefore most useful in situations where contradicting findings
are welcome as enrichment and not perceived as a flaw. The primary aim is deepening insight and not
formal validation.
5. Multiple triangulation: It the use of more than one type of triangulation

d. Problems with triangulation

1. If not done well, it may increase bias

2. It is costly
3. It is time consuming
4. It requires more technical know-how
5. Triangulation is trans-paradigmatic and not restricted to one paradigm
6. If contradictions are found, which if the “true” finding? Triangulation leads to a problem on how to
deal with contradictions
7. How to decide what methods can be combined; the question is: How different can methods be so
that their combination will still be enriching.

NOTE: Triangulation is not only about enhancing validity and reliability but also enriching and deepening
insight through discovery of peculiar patterns often termed “Outliers”.

NOTE: There is wide agreement on how triangulation should not be used; it cannot cover up weaknesses
or flaws in research design. A study in which one method has been shoddily applied cannot be salvaged by
simply adding another method.

General Recommended Reading

Essential Reading:
Manderson L. and P. Aby (19XX): An epidemic in the field? Rapid assessment Procedures and
Health Research; Social Sciences in Medicine; Vol. 35 No. 7; Pergamon Press Limited

Patton, Michael Quinn (2002): Qualitative Research and Evaluation Methods. 3rd Edition. Sage
publication Inc. London

Bryman Alan (2001): Social Research Methods. Oxford University, New York

Dawson, S., L. Manderson, et al. (1992): "The Focus Group Manual." Methods for Social
Research in Tropical Diseases, 1992

Further Reading:
Morgan D.L. (1996): Focus Groups. Annual Review of Sociology, 1996; 22: p. 127-58

Khan, M.E., et al. (1991): The use of focus groups in Social and behavioural research: some
methodological issues. World Health Status-quo, 1991; 44(3): 145-9

Lee, J.A., SJ. Moore, and B.S. Cotiw-an (1999): Problems translating a questionnaire in a Cross-
cultural setting

108
2.2.7 Extension Activities

Extension Activity 1: Discussion Forum Question

Briefly describe the distinguishing features of qualitative research and its strategic value
especially in understanding health phenomena and triangulation to enrich research findings

Extension Activity 2: Self- Assessment Quiz

QUIZ 2.2.1
(For each question, follow the instruction given)

1. About qualitative research, the following are true except:

a) It embodies a holistic understanding of social phenomena
b) Iteration means that the research questions are revised during the research process, and the line of
questioning is emergent
c) Iteration means that the line of questioning is not rigid, and can delve into other areas side from the main
question
d) Flexibility means that we can change the tangent of questioning in a dynamic way
2. Qualitative Research is most appropriate in the following situation:
a) When repeatability if measurements is important
b) When generalizability of measurements is paramount
c) When the subject matter under study is clearly defined and familiar
d) When the subject matter is fuzzy, and we need to generate constructs and theory
e) None of the above
3. The following are common probes used in qualitative data collection except:
a) Reprimand
b) Silence
c) Repetition
d) Emphasis phrases
e) “Assertions of Interest”
4. The following are true (T) or false (F) about qualitative research questions; indicate appropriately
a) The question “Do you believe in Witchcraft?” is an open ended question
b) The question “Don’t you believe in your health workers?” is a leading question
c) The phrase “And what happened next?” is a repetition probe
d) The phrase “And what happened next?” is an assertion of interest probe
5. About Key Informants, one of the following statements is incorrect:
a) They are often authorities on the topic of interest
b) They are always in a managerial or leadership role
c) They often share similar characteristics as the participant population
d) Often, one Key Informant may be interviewed several times
6. The most commonly used PRA techniques are:
a) Semi structured Interviews, FGDs and Key Informants
b) FGDs, Key Informants and Ranking
c) FGDs, Participatory Mapping and Key Informants
d) Participatory Mapping, FGDs and Ranking
7. During the formative stage of a post vaccination survey, a Pre-survey Rapid Appraisal was conducted to develop
the scope variables to be considered in the actual survey. In one qualitative tool, the first set of guiding questions
went as follows:
A. Do you experience Adverse Events Following Immunisation in Your Village?
B. What common complaints do you find?
109
C. Mr. X, Please share with us what happened next?
a) Question A is called the “Opening Statement”
b) Question C is a follow-up question
c) Question A and B form the “Main Question”
d) Question C is an “Assertion of Interest” Probe
8. The qualitative data analysis process can be summarised into 4 steps. They are:
a) Coding, transcribing, summarising and presentation
b) Coding, labelling, summarising and Presentation
c) Transcribing, Collapsing, summarising and Display
d) Encoding, transcribing, presentation and Citation
9. Indicate True (T) or False (F), for the following statements:
a. In Content Analysis, codes are emergent in the data analysis process
b. In content analysis, codes are identified prior to searching for them in the data
c. In Semiotic analysis, codes and metaphors are collapsed into emergent themes, a process also called
Selective Coding
d. Metaphors are data reducing devices that summarise a number of differing aspects of a phenomenon
10. The following statements are True (T) of False (F) about Triangulation
a. Across Methods triangulation is a methodological triangulation technique that employs several methods of
data collection
b. Investigator or Analyst triangulation involves the use of several hypotheses or paradigms
c. In data triangulation, multiple sources of data are used and the data compared
d. Within Method Triangulation is mainly dome to increase reliability of data
e. Methods triangulation is the commonest form of triangulation used

110
 2.4 Unit 4: RESEARCH PROPOSAL DEVELOPMENT,
IMPLEMENTATION AND WRITING SKILLS
2.4.1 Introduction to the Unit
In the previous units, we have explored in details the key concepts in Research methods. These concepts
will help you in developing your research questions in to a complete protocol and in writing up your
research proposal. In this Unit, we shall explore the operationallisation of the principles gathered in the
previous sessions through development of a research proposal.

This unit is particularly important in the course Units “Field Study I, II and III, when you will be expected to
present field reports for short studies. It will also be important when you are developing your dissertation.

2.4.2 Unit Outline

The following topics will be covered:
Session 1: Research Proposal Format and Writing Skills
Session 2: Tips in implementing the research proposal

2.4.3 Instructional goal

The MPHO should be able to write and implement a research proposal, incorporating appropriate
methodologies to address specified research questions

2.4.4 Unit Objectives

By the end of this unit, the student should be able to:

1. Write a research proposal using appropriate methodologies
2. Implement a research process based on a research proposal

2.4.5 Time Frame

1 WEEK

2.4.6 Content

111
Session 1: Research Proposal Format and Writing Skills

Introduction: A research proposal is a written presentation of a plan indicating the strategy the researcher
intends to use in search of answers to the problem in question and the specific objectives identified. The
proposal ought to follow a standard logical format and in this session, we shall examine a recommended
format for your proposals and reports.

Session Topics: The following topics will be covered:

a. The Proposal Format
b. Steps in developing a research proposal
c. Components of a Research Proposal

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Outline and paraphrase the different components of a research proposal
2. Use appropriate and acceptable writing skills to formulate a good research proposal

a. The Proposal Format

 Title Page
 Table of contents
 Acronyms and Abbreviations
 Definition of Terms/Operational Definitions
 Abstract

 Introduction and Background to the study

 Literature Review
 Statement of the Problem
 Justification (Rationale, Significance)
 Conceptual Framework
 Hypotheses

 General Objective
 Specific Objectives
 Research Questions

 Methodology
o Study Site/ Setting
o Study Population
o Study Design

o Sample Size
o Sampling Procedures or selection of study participants
 Inclusion Criteria
 Exclusion Criteria
o Data Collection Procedures
112
  Data collection tools
o Quality Control

o Variables
o Data Handling and Management
o Data Analysis

o Ethical Considerations
o Study Limitations
o Dissemination of findings

 Budget and Budget Justification

 Workplan
 References
 Appendices
o Tools
o Other appendices

b. Steps in developing a research proposal

The flow chart below summarized steps of developing a research proposal.

Questions you must ask Steps you will take Important elements of each
step

What is the problem? Why Selection, analysis and - Problem identification

should it be studied? statement of the research - Prioritizing problems
problem? Why is it important? - Analysis
- Justification

What information is already - Literature and other available

available? Literature review information

Why do we want to carry out - General and specific

the research? What do we Formulation of objectives objectives (SMART)
hope to achieve? - Hypotheses

What additional data do we - Variables

need to meet our research - Types of study
objectives? How are we going - Data collection techniques
to collect this information? - Sampling
Research methodology - Plan for data collection
- Plan for data processing and

113
 analysis
- Ethical consideration
- Pre-test of pilot study

Who will do what and when? Work plan - Personnel

- Work-plan

How will the project be - Administration

administered? How will Plan for project administration - Monitoring
utilization of results be and utilization of results - Identification of potential
ensured? users

What resources do we need - Material support and

to carry out the study? What Budget equipment
resources do we have? - Money

How will we present our

proposal to relevant Proposal summary
authorities and potential
funding agencies?

114
c. Components of a Research Proposal

The Title Page: The title page contains basic information concerning your study, including the title,
the authors and other related information. The title ought to contain only the critical information
necessary to indicate what the study is about. It should be as brief as possible and avoid redundant
words. The title page also to contain information on the author, his or her identity and any related
wording. It ought to indicate the purpose for which the proposal has been submitted. The title or
should be brief and specific. It should carry the meaning behind the study i.e. it should be concise
but informative. Preferably 5 – 15 words, it should not exceed 20 words.

Table of contents: This is an outline of the contents of the proposal

Acronyms and Abbreviations: We make a list of acronyms and abbreviations, and describe them
in full. By abbreviating, we make a word shorter by omitting letters, e.g. Under-5s (Children under 5
years of age), Prof. (Professor). Acronyms on the other hand refer to summaries involving the first
letter of every word in a given phrase e.g. WHO (World Health Organization), PLWAs (People living
with AIDS).

Definition of Terms/Operational Definitions: Terms: These are words or phrases commonly

used but you think they are not easily understood. Operational definitions on the other hand are
words or phrases coined specifically for your research or those terms used differently from common
English.

Abstract: In IPH we use a structured abstract. It ought to be a brief summary, preferably one page,
and not exceeding 2 pages. It contains the following sections: Background, Study Objective,
Methods, and Expected Results.

Introduction/Background
It contains the general statements that introduce the subject we are dealing with. This section
enables the reader to appreciate the context and framework of the problem being studied. The
magnitude of the problem and the importance for studying it are stated. There are variations in the
style of writing the introduction and background section of a proposal depending on the requirement
(dissertation, manuscript or journal article) institution, funding agency or teacher of research
methodology. Commonly the introduction section introduces the study subject in the context of a
large-scale e.g. global and /or national situation, while the background focuses on the problem in
the context of the study area. The format adopted by Makerere University Institute of Public Health
is such that both the introduction and the background are in the same chapter with different
subtitles that reflect content of each subsection.

Literature Review

A detailed review of literature in relation to the subject we are dealing with. It ought to review the
salient components of the conceptual framework and inter-linkages between different factors. It
should try to collate information to discern what research has been done in the area of interest and
what gaps still linger. The best literature review is structured.

Careful search and choice of literature will facilitate a better analysis of the subject of interest and it
should lead to a well thought problem to be studied. Cite only patent literature to support your
statements regarding the importance of the subject to be studied, what is already known, what the
gaps are in terms of knowledge, study design and data analysis etc. This is usually presented as a
separate section in the dissertation because of the need to demonstrate depth and breadth in
understanding of the subject matter. In ordinary proposals and reports, it is part of the introduction
and discussion sections.

Statement of a problem

This usually introduces the problem to be studied and clearly brings out the gap in knowledge
between what is and what should be. Possible explanatory factors ought to be brought out as a
basis for conceptualizing the problem.

It is a concise and focused statement identifying the problems and subject gaps that will form the
basis for areas of research. It must be concise, brief, clear and capturing the reader’s interest
answering the questions:
 What is the problem?
 What is the magnitude of the problem?
 Who are affected?
 What is already known about the problem?
 What seem to be the major unanswered questions about the problem?
 The problem states that the answer to the research question is not known.

Rationale/Justification for the study

The rationale of a study and justification are the same. The justification of the study explains why it
is important to do carry out the study. It also describes the benefits of the study to the existing body
of knowledge, medical technology, service to the community and the health of the people and how
the study addresses the stated problem. This section attempts to answer the questions: Why does
this study need to be done? Why the urgency – why does it need to be conducted now? More
importantly, of what use will be the information generated from this study?

Conceptual framework
This is an examination of existing or self-formulated theories in relation to the objectives. It is an
explanatory framework. This is a scheme of concepts or variables that underlie the study and the
relationship among these various variables. The conceptual framework should be described to
enable the reader to understand the relationships among the various variables.

We prepare a problem analysis diagram showing the interrelationship between the different factors
and grouping constructs. This is called the conceptual model. We attach a brief narrative to
describe the model: this is called the conceptual framework.

Hypothesis
A hypothesis is a statement of the likely cause-effect relationship. It is stated in analytical and
experimental studies. Hypothesis statements should be well focused and indicative of the
researcher trend of thought and inquiry. The test of hypothesis involves collection and analysis of
data that may either support or fail to support the hypothesis. Descriptive studies generate
hypotheses to be tested in analytical studies. Cross-sectional studies may be analyzed like
analytical studies, in which case there may be several hypotheses generated, more like a “fishing
expedition”.

Study Objectives
State the general objective and a series of specific ones reflecting clearly what you want to
investigate.

116
The General Objective is the goal of the study. It states the overall question that has to be
answered. It often has a dimension that states of what use the information generated will be.

The Specific Objectives break down the general objective into thematic areas that will be
investigated. The objectives ought to be Specific, Measurable, Achievable, Realistic and Time
bound (SMART). In many cases, the time aspect is implicit.

Research Questions: Research questions arise from the objectives.

Methods (Materials and methods)

The Methodology is a detailed description of how the research will be conducted. This is a plan,
structure and strategy of the investigation conceived so as to obtain answers to the research
questions. The description of the method section should be detailed enough for any
researcher to replicate the study. In this section include the study area, study population, study
design, recruitment procedure, sample size, sampling procedure, study variables, data collection
and data management and analysis. Describe the materials/ equipments for the study. Give
detailed description of what will be done step by step. Describe any limitations or bias that may be
produced as a result of your methods and how you intend to deal with them. Specify any computer
programs that you intend to use in data analysis.

The methodology section has the following thematic areas:

Study Site/ Setting

Study Population
Study Design

Sample Size
Sampling Procedures and selection criteria
Inclusion Criteria
Exclusion Criteria
Data Collection Procedures
Data collection tools
Quality Control
Variables
Ethical Considerations
Data Handling and Management
Data Analysis
Study Limitations
Dissemination of findings

Ethical considerations
The methods and procedures to be used in the study must conform to the national ethical
guidelines. Permission must be obtained from the ethical review board. Informed consent must be
obtained if the study involves human subjects. Potential risks and benefits that may result from the
participation must be explained to the subjects before obtaining consent.

Dissemination
The utilization of research results is the goal of every research activity. The process of research is
not complete until the results are disseminated fully and every effort has been used to have them

117
used. Some of the strategies you can follow to ensure that the results of the study will be used
include:
 Involving relevant authorities, staff and community members in the selection of topic and
definition of the problem
 Listing the major types of recommendations expected from the study and identifying who
will be involved in their implementation
 Identifying the communication channels which already exist that can be used to
disseminate results
 Determining what written material should be prepared to keep relevant authorities involved
 Determining whether additional actions should be taken or mechanisms developed to
inform all parties concerned of the study results and obtaining their approval and
cooperation for the implementation of the recommendations.

Appendices:
The Work Plan
This is a brief schedule showing the chronology of activities, as they will follow each other to the
completion of the study. In addition to the components of the research proposal discussed above, a
work plan has to be developed addressing the questions: who will do what and when? The
researcher also needs to develop a realistic plan for project administration and utilization of results.
There is need to list the responsibilities of the principal investigator related to administration and
monitoring of the research project. A plan for administration and monitoring of the research project
allows for orderly and accurate purchase and procurement of equipment, payment of bills and
preparation of financial reports. It allows the researcher to foresee the need for funds and make
timely request, to avoid unwanted breaks in the implementation of the project and allows the
researchers to devote most their time to technical and scientific aspects of the project.

Budget and Budget Justification: We ought to prepare a detailed line item budget, showing the
expenditure items and their costs. We then write a justification for the different budget items.
Another important question to be answered is: what resources are needed to carry out the study
vis-à-vis what resources are already available? The researchers will need to develop a realistic and
appropriate budget, which will help in identifying the resources already available and the additional
resources that will be required. The budget will specify for each activity in the work plan what
resources are required. It is not sufficient to present a budget without an explanation. The budget
justification follows the budget as an explanatory note justifying briefly, in the context of the
proposal, why the various items in the budget are required. If a strong budget justification has been
prepared, it is less likely that essential items will be cut during budget review.

The Data Collection Tools and Standard Operating Procedures: Your proposal should have the
data collection tools appended, so that they can be critiqued by other readers. All tools,
qualitative and quantitative, ought to be attached.

118
Session 2: Tips in Implementing the Research Proposal

Session Topics: The following topics will be covered:

a. The Field Work Phase
b. Overview of Data Collection Techniques

Session Objectives:
By the end of this Session, the MPHO should be able to:
1. Implement a research proposal, using methods specified in the proposal
2. Select a mix of data collection techniques to allow methods triangulation in a given study
design

a. The Field Work Phase

Having completed the research proposal, you are now ready to conduct the study. The framework
on the following page guides you through the steps of fieldwork.

Steps in the Field Work Phase

Questions you must ask Steps you will take Important elements of each step

Will managers and health Administrative and motivational - Briefing: obtain permission
staff provide support? preparation - Form research groups

Are formats and instruction Preparation for data collection - Logistic preparations
manuals ready? Have data - Pre-testing and revision of
collectors been trained? tools
- Training
-Arrangements for supervision
and quality control

Is data collection on Data collection - Checking; sorting

schedule?

For qualitative data, are more Data processing and - Coding and preparation of
(or different) data needed? preliminary analysis manual for data processing by
computer or
- Preparation of master sheets for
manual processing of data
- Categorization of qualitative
data

119
Activities to be carried out between proposal development and data analysis and report writing
consist of field operations and data processing. These include
 Briefing managers and health service staff regarding the project
 Obtaining the necessary permission to collect data
 Identifying and obtaining the resources (manpower, materials etc) needed to collect data.
 Reviewing the availability of subjects and information and organizing logistics for data
collection.
 Training interviewers, data collectors and supervisors.
 Refining, pre-testing and revising the research instruments for data collection.
 Collecting the required data
 Processing the data.

b. Overview of Data Collection Techniques

Data collection techniques allow us to systematically collect information about our objects of study
(people, objects and phenomena) and about the setting in which they occur. In the collection of
data, we need to be systematic i.e. organized. If the data are collected haphazardly, it will be
difficult to answer our research questions in a conclusive way. Various data collection techniques
can be used such as:
 Using available information
 Observing
 Interviewing (face-to-face)
 Administering written questionnaires
 Focus group discussions
 and other data collection techniques.

Data Collection Techniques and Tools

Data Collection Techniques Data collection Tools
Using available information Checklist, data compilation forms

Observing Eyes and other senses, pen and paper, watch,

scales etc.

Interviewing Interview schedule, checklist, questionnaire, tape

recorder
Administering written questionnaire Questionnaire

Advantages and Disadvantages of various data collection techniques

Technique Advantages Disadvantages
Using available Inexpensive because data is Data are not always easily
information already there. accessible.
Permits examination of past Ethical issues concerning
trends. confidentiality may arise.
Information may be imprecise or
complete.
Observing Gives more detailed and context- Ethical issues concerning
related information. confidentiality or privacy may arise.
Permits collection of information Observer bias may occur (observer
on facts not mentioned in the notices only what interests him/her)

120
 questionnaire. The presence of data collector can
Permits tests of reliability of influence the situation observed
responses to questionnaires. Thorough training of research
assistants is required
Interviewing Suitable for use with illiterates The presence of interviewer can
Permits clarification of questions influence responses
Higher response rate than written Reports of events may less
questionnaires complete than information gained
through observation
Small-scale flexible Permits collection of in-depth The interviewer may inadvertently
interview information influence the respondents
Exploration of spontaneous Open ended data are difficult to
remarks by respondents analyze

Larger-scale fixed Easy to analyze Important information may be

interview missed because spontaneous
remarks by respondents are usually
not recorded or explored.
Administering written Less expensive. Cannot be used with illiterate
questionnaire Permits anonymity and may result respondents
in more honest responses There is often a low response rate.
Does not require research Questions may be misunderstood.
assistants
Eliminates bias due to phrasing
questions differently with different
respondents.

Importance of combining different data collection techniques

From the discussion of different data collection techniques, their advantages and disadvantages
above, it becomes clear that they can complement each other. The skilful use of a combination of
techniques can maximise the quality of the data collected and reduce the chance of bias.

Structured questionnaires that enable the researcher to quantify pre- or post categorized answers
to questions are an example of QUANTITATIVE research techniques. The answers to questions
can be counted and expressed numerically. Quantitative research techniques are used to quantify
the size, distribution and association of certain variables in a study population. Both quantitative
and qualitative techniques are often used within a single study.

Qualitative research techniques involve the identification and exploration of a number of often
related variables that give INSIGHT into the nature and causes of certain problems and into the
consequences of the problems for those affected.

Researchers often use a combination of flexible and less flexible research techniques. They
produce qualitative information, which is often recorded in a narrative form.

Flexible techniques include:

 Loosely structured interviews using open ended questions
 Focus group discussions (FGD)- a group discussion of 6-12 persons guided by a facilitator,
during which group members talk freely and spontaneously about a certain topic.

121
  Nominal group technique- a group discussion technique useful when one wants to obtain a
consensus from a group where decision-making can be usefully guided by the perception
and opinions of the various group members.
 Delphi technique- used in a situation where a group needs to reach a consensus over an
issue that is highly value-laden. However, the groups usually don not meet for discussion
but communicate by means of questionnaire.
 Life histories- this technique allows people to tell stories, which provides incite in what they
consider insight for example patterns of reproduction, women feelings about marriage and
contraception.
 Scales- highly structured interview. A sequence of questions is set and highly
standardised. Scales have been invented to measure complex concepts such as health,
depression, intelligence, fear etc.
 Case studies- involve detailed investigations of a few people, a community or a particular
situation. Usually a number of methods for collecting information are used simultaneously.
 Essays- used to explore the (hidden) values and aspirations. Essays may be analyzed to
determine differences in beliefs concerning perceived causes of illness, rational for health
related behaviour and the like.
 Participatory research- the boundaries between research and health programs are blurred:
through the implementation of the research, it is expected that conditions influence in the
health system will change. All phases of the research are planned and conducted together
with the researchers and target population together.

Extension Activities

Extension Activity 1: Discussion Forum Question

Extension Activity 2: Self- Assessment Quiz

122
 2.5 Unit 5: DATA ANALYSIS AND REPORT
WRITING
2.5.1 Introduction to the Unit

After collecting data from a research process, the researcher ought to analyze the data an present
their findings in an orderly and scientific way. The research proposal is them modified into the
research report. This unit guides the learner through the compilation, analyzing and writing up of
the findings of the study. Study findings are presented in the results section of the report.

2.5.2 Unit Outline

The following topics will be covered:
Session 1: Data Analysis
Session 2: Research Report Writing
Session 3: Dissertation, Abstract and Manuscript Formats

2.5.3 Instructional goal

The MPHO should be able to analyse data and prepare a scientific report for a given research
design

2.5.4 Unit Objectives

By the end of this unit, the student should be able to:

1. Select appropriate methods for analysing data to answer a given research question
2. Write a research proposal
3. Outline the features of an academic research dissertation and a scientific manuscript for
publication

REFERENCES
For further reading on data analysis, read Varkevessier CM, I. Pathmanathan and A. Brownlee,
(1991). Designing and conducting health systems research projects. Part 2. Data analysis and
report writing. IDRC. Ottawa, Canada.

2.5.5 Time Frame

½ WEEK

2.5.6 Content

123
Session 1: Data Analysis

Introduction: Data analysis is a very important aspect of the report writing process. You will have
to analyze your data objectively and present the important findings. Data Analysis is in three
stages:

1. Univariate (or Univariable) Analysis

2. Bivariate (or Bivariable) Analysis
3. Multivariate (or Multivariable) Analysis

In this session, we shall re-cap our earlier knowledge on these methods.

Session Topics: The following topics will be covered:

a. Univariate Analysis
b. Bivariate Analysis
c. Multivariate Analysis

Session Objectives:
By the end of this Session, the MPHO should be able to:

a. Univariate Analysis

Univariate analysis is the first step of data analysis. In Univariate analysis, we conduct an
exploration of the data and analyze for descriptive statistics for the main variables in our data.
Univariate analysis may be the only analysis necessary and may provide adequate results for some
studies, especially the short studies involving cross-sectional studies that are purely descriptive or
those that have small sample sizes. The main interest here is to determine prevalence of important
characteristics in a population and possibly formulate hypotheses.

For numerical variables:

We compute summary parameters like:
 The means
 The modes
 The medians
 The standard Deviation and variance
 The range and inter-quartile ranges

We present these mainly using a narrative text, but we may also resent summary tables where the
parameters and variables are many.

For categorical variables

We compute proportions, mainly in the form of percentages. We present them in frequency tables
that show the frequencies and percentages. Most variables you will deal with in your study reports
are categorical variables and it is important that you learn how to present the findings.

124
Use of Figures and Graphs in data presentation

It is important to use appropriate figures to present your data. Apart from frequency tables, other
sections of the variables may be presented in an enhanced way using graphs and figures.
Examples include:
- Frequency polygons
- Histograms
- Bar Charts
- Scatter Plots
- Line Graphs
- Pie Charts
- Box and Whisker plots
- Many other forms

Refer to your notes in the course: Applied Biostatistics I for revision of the different methods of
data presentation and summary

b. Bivariate Analysis

In many of your studies, it may be necessary to conduct a level of Bivariate analysis, to compare
the relationship between the dependent and independent variables. In such cases, we have to
conduct tests of association, report statistical significance and make conclusions concerning the
relationships between variables. There are different scenarios possible:

a. Comparison of a multi-categorical variable with another variable that is either binary

or multi-categorical:
a. Here, we use the chi-square test. An example is when we compare level of
education (Multicategorical) with religion (multi-categorical) or sex (binary)

If the dependent variable we are comparing is the same, it is even better to present
these in the same table e.g.

Education Level X2 p-value

Variable Primary Secondary Tertiary
no % no % no %
Religion
Catholic
Protestant
Moslem
Pentecostal
Other
Sex
Male
Female

We do not have to present a separate table for sex and one for religion, comparing
with education level.

125
 As you can see, we need a data analysis plan to predetermine what univariate
analysis we shall run and what Bivariate comparisons we shall make.

b. Comparison of a binary categorical variable with another binary categorical variable:

a. Here, we have two options. We may use the chi-square test, nut we can also use
Odds Ratios or Relative Risk, and reporting statistical association using either the
chi-square p-value for the chi-square or the confidence interval or p-value for the
measures of risk (OR and RR). Again, if we have a common dependent variable,
all the comparisons with independent variables can be summarized in the same
frequency table;

An example is as follows:

Nutritional Status Odds CI of OR or p-

Malnourished Not Ratio value
Variables Malnourished
No % No %
Sex
Male
Female
Age Group
Above 5 Yrs
Below 5 Yrs
Parent
Present
Yes
No

c. Comparison of a numerical variable with a multi-categorical variable:

a. Here, if the multicategorical variable is binary, we use a t – test. An example is if
we are comparing the mean heights of students in two schools. The computer tells
us the mean difference, the t – statistic and the p-value or confidence interval of
the difference. The t – test may be an independent samples t – test like in the
situation above, or it may be a paired t – test if the two samples are comparing
means before and after in the same sample e.g. if we have malnourished babies
and we take their weights and determine the mean. Then we feed them for 3
weeks and take their weights again, so that we can compare the change in mean
weights.
b. If the multicategorical variable (grouping variable) is more than two categories, we
conduct several t – tests, or we can use a procedure called Analysis of Variance
(ANOVA) which we shall see in Applied Biostatistics II

d. Comparison of a numerical variable with another numerical variable with one of

them being discrete and the other continuous
a. Here, we may use linear regression analysis. An example is if we are comparing
the relationship between class-grade (Senior 1 to Senior 6) with height. This
procedure you will learn in Applied Biostatistics II
e. Comparison of a numerical variable with both of them being continuous variables

126
 a. Here, we compute correlation coefficients as the test for statistical significance.
This you will also learn in Applied Biostatistics II

c. Multivariate Analysis

Sometimes it may be necessary to conduct multivariate analysis (or adjusted analysis) to control for
confounding and assess for interaction and effect modification. These techniques will be introduced
to you in Applied Biostatistics II

127
Session 2: Research Report Writing

Session Topics: The following topics will be covered:

a. Formative Aspects of the Report
b. Summative Aspects of the Report

Session Objectives:
By the end of this Session, the MPHO should be able to:

a. Formative Aspects of the Report

By this stage of the research report, you will have written down the formative aspects of the
research report. These are all contained in the Research Proposal. All you have to do is to up-date
the proposal by changing the different sections in two ways:

1. You will up-date the aspects that changed during the actual implementation of the
research. A few aspects of the proposal may have changed according to the circumstances
that evolved during the actual implementation of the data collection
a. The literature review may have changed
b. The research objectives and questions may have been slightly altered
c. Certain aspects of the methodology may have changed
d. Certain variables may have evolved while some may have been altered during the
data collection process
e. Other limitations of the study and ethical issues may have been encountered other
that those that were anticipated in the proposal
f. Quality control procedures and data collection techniques may have been altered

You have to up-date all these in the proposal, which has now turned into the formative part of
the research report.
2. You need to change the tense of the proposal. The proposal is written in future tense, while
the report is written in the past tense.

b. Summative Aspects of the Report

After up-dating the report, you will have to write the Summative aspects of the report, including:
The results section, the Discussion, the conclusions and the recommendations.

Results

The results should be presented in form of text, tables, charts and graphs as applicable. When and
where possible, results should be quantified and a measure or indicator of their measurement error
or uncertainty, such as confidence interval should be given. Avoid non-technical use of technical
words in statistics like “random”, “normal”, “significant”, “correlation” and “sample”. Define those
terms where necessary.

128
Scholarly value of Your Write up

At the graduate level, you are expected to make a write-up using a powerful combination of the
different skills you have learned so that your report reflects the skills.

A scheme for presenting your results

- Always start by describing the background socio-demographic characteristics of your study

population
- The do on to describe other findings in relation to your research objectives
o For each sub-theme, you may start by the univariates
o Then go to the comparisons
- Always start with an opening statement, followed by some core findings. You can then
present a table or graph, and them some description of the key findings from the table
- In the descriptive narrative after a table or figure, do not repeat all the data presented in
tables because this is a waste of time – the table is intended to present these better. Just
mention the peculiar findings or key observations
- Always introduce your tables and graphs and after them make some observations – this
prevents them from looking ‘orphaned’
- Intersperse your quantitative findings with qualitative findings if you have any, and observe
to what extent they relate.

Discussion
Under this section, bring out and emphasise only the results that are important to the study and
conclusions that follow from them. Relate the study findings to other study findings. Indicate the
implication of study findings. If there are limitations, indicate the areas for future research.

A Scheme for discussing your results

- It is always better to use a structured discussion with themes mirroring those you used in
the results and related to your objectives
- Start by a summary of the key finding
- Then go on to explain the finding in relation to your expectation, your qualitative findings,
the available literature and other studies that have been done
- Then make a scholarly deduction
- Move on to the next theme

Conclusions and Recommendations

Link the conclusions of the study with the objectives of the study. Recommendations, where
appropriate, should be included.

MPH requirements:
During the MPH training, you will be required to write four field reports of short studies and a
dissertation.

129
Session 3: Dissertation, Abstract and Manuscript Formats

Session Topics: The following topics will be covered:

a. Format for the Dissertation
b. Format of a structured Abstract
c. Preparing a manuscript for publication

Session Objectives:
By the end of this Session, the MPHO should be able to:

a. Format for the Dissertation

 Title page
 Supervisor’s page to be signed
 Declaration (optional)
 Dedication (optional)
 Acknowledgements
 Table of Contents
 Acronyms and abbreviations
 Operational definitions
 Abstract (150 – 300 words)

1.0 Introduction and Background (with sub titles that reflect content of each subsection)
2.0 Literature review (with subsections)
3.0 Statement of the Problem, Justification and Conceptual Framework
4.0 Study objectives
5.0 Methodology
5.1 Study Site/ Setting
5.2 Study Population
5.3 Study Design

5.4 Sample Size

5.5 Sampling Procedures or selection of study participants
5.5.1 Inclusion Criteria
5.5.2 Exclusion Criteria
5.6 Data Collection Procedures
5.6.1 Data collection tools
5.7 Quality Control

5.8 Variables
5.9 Data Handling and Management
5.10 Data Analysis

5.11 Ethical Considerations

5.12 Study Limitations
5.13 Dissemination of findings

130
6.0 Results
7.0 Discussion (Limitations pointed out as part of the discussion of results)
8.0 Conclusions
9.0 Recommendations
 References
 Appendices/Annexes

b. Format of a structured Abstract

(150 - 300 words)

For a proposal, the abstract should have the following subsections:

 Background
 Objectives
 Methods

For a research report, the abstract has the following subsections:

 Background
 Objectives
 Methods
 Results
 Conclusions and recommendations

c. Preparing a manuscript for publication

Each publisher will have different requirements. However the generic format includes:

1. Title page
2. Abstract and Key Words
3. Introduction
4. Methods
5. Results
6. Discussion
7. References
8. Tables
9. Illustrations (Figures)

UNIT SUMMARY
We have gone over the generic format of a research proposal, discussed data collection and
fieldwork and the additional sections for writing a research report. The particular format you would
use will depend on what your client wants. Some clients in calling for proposals will specify which
sections and topics it should have. However, the generic format above has the generally agreed
parts a research proposal and research report should have.

131
 2.5.7 Extension Activities

Extension Activity 1: Discussion Forum Question

Extension Activity 2: Self- Assessment Quiz

132
 3.0 ADDITIONAL RESOURCES

3.1 TEXT DOCUMENTS FOR ADDITIONAL READING

Text Document 3.1.1:

3.2 GLOSSARY OF TERMS

PRA: It is a collective learning process with partners using a series of flexible participatory
techniques to analyse a situation, carry out needs assessment, identification of priorities for
development activities, and to plan activities in general in a systematic and intensive fashion.

Public Health: Activities that society undertakes to assure conditions in which people can be
healthy and in which people can prevent, identify and counter threats to the health of the public.

Qualitative Research: Generates data from words, observations of events, pictures etc.

Quantitative Research: It generates data from numbers.

Research: It is the organised quest for new knowledge based on curiosity or perceived need. It
may consist of systematic empirical observation or hypothesis testing.

Variables: They are the operational forms of constructs. They state how a construct is to be
measured in a specific situation. It is important to keep in mind that VARIABLES should be
matched to CONSTRUCTS when you are identifying what needs to be assessed in the evaluation
of a theory-driven program

3.3 REFERENCES

1. Varkevessier CM, I. Pathmanathan and A. Brownlee, (1991). Designing and conducting

health systems research projects. Pt1. proposal development and fieldwork. Pt2. Data
analysis and report writing. IDRC. Ottawa, Canada
2. Kakitahi J.T. (1998): Students’ guidelines for dissertation and research proposal write-up.
Makerere University Kampala.
3. Makerere University Institute of Public Health, (2004): MPH Programme Hand book
Kampala.
4. Polgar S. and SA Thomas (2000). Introduction to Research in the Health Sciences.
Churchill Livingstone. 4th Edition. Toronto.
5. Fisher AA, JR Foreit (2002). Designing HIV/AIDS Intervention studies. An operations
Research Handbook. Population Control. Washington DC.
6. Alex Ojacor (2004). Format for Research proposal, thesis/ dissertation writing; A guide to
carrying out scientific research today. Kampala.
7. Hulley SB; Cummings Designing Clinical Research;
8. Bernard Rosner (1990): Fundamentals of Biostatistics; Harvard University Press

133
3.4 ANSWERS TO QUIZ QUESTIONS

QUIZ 2.2.1
1. c
2. d
3. a
4a. F 4b. T 4c. F 4d. T
5. b
6. a
7. d
8. b
9a. F 9b. T 9c. F 9d. T
10a. T 10b. F 10c. T 10d. T 10e. T

134
3.5 INDEX OF URLs FOR INTERNET RESOURCES

3.6 INDEX OF DISCUSSION FORUM QUESTIONS

You are encouraged to participate in the discussion forums that have been pre-planned for you in the
semester. Through these forums, you will be able to exchange information with the moderators and
fellow students and gain a deeper understanding of the material you have read. There will be at-least
one discussion forum for each course, with a number of questions drawn from each course unit. You
are requested to post your discussion points to the board for other members to share. Please be brief
and to the point. You may discuss only one question or a number of them depending on where you feel
motivated. You may also post a discussion point that is outside the set questions, provided you have
HOT points to share. These forums will enhance the “virtual” classroom environment and facilitate you
to learn at the same pace as the others.

There is a detailed outline of the schedule of these forums that will be handed to you at the beginning of
the semester, under the resource: SEMESTER SCHEDULES. At the precise times indicated
(Modifications in the schedule may from time to time be communicated to you by the Moderator), the
discussion will be activated and you will be called upon to contribute; this will be a “silent” online call –
you are requested to remain alert, and regularly check the forum platform or your internet mail-box for
the call. For each course the discussion will run for an entire week.

The forums will be hosted at the MUSPH e-learning tool on the site http://courses.musph.ac.ug . Please
sign up and register your unique identity in the forum. The following is a summary of the questions that
are up for discussion in the Discussion Forum for this particular Course:
1.

2.

3.

4.

5.

POST YOUR REPLY NOW:

Post your reply now to one or more of these issues and attend the Forum; you will discover the
unique learning experience from sharing knowledge in this interesting resource!

3.7 INDEX OF ADDITIONAL RESOURCES FOLDER

3.8 INDEX OF SELECTED LECTURE NOTES

135
3.9 SUMMATIVE EVALUATION OF THE INSTRUCTION PROCESS

Summative Evaluation of the instruction will be conducted using the following means:
1. Progressive Assessment in form of Hand-in Assignments
2. Progressive Assessment Test – To be done during the Face – to – Face Sessions
3. The University Examination
4. An optional Post-test
5. A Course Post Evaluation Questionnaire

Progressive Assessment – Hand-in Assignments

These assignments should be handed in by the time of sitting for the progressive assessment test at
the institute of public health. They will be marked and will contribute to the final progressive score. The
number, nature and timing of assignments will be determined by the Course Coordinator. Some of
these assignments may be directly included in these materials by the Course Coordinator. An index of
them is listed below:

TITLE COMENTS
1.
2.
3.
4.
5.

Post-test
It is optional for you to attempt this Post-test. It will assist you to gauge your grasp of the material after
the instruction process. The test is contained in the Additional Resources Folder.

136
 Post Evaluation

We are in need of your feedback on the quality and content of these materials. It will be valuable to the iterative
process of their further improvement. For this purpose, we have attached a questionnaire to gauge your
perception of the design and conduct of this course and to link this to your understanding of the subject matter.
This questionnaire has been introduced to you previously. It contains two parts: a Post Evaluation of the
materials for the previous semester and a pre-evaluation of the materials you expect for this semester. Please
note that this is not a progressive assessment or exam, and will not contribute to your final mark. It should be
completed at the beginning of the semester. Please fill in the Evaluation questionnaire for this semester; make
comments as requested, and send it as an e-mail attachment or hardcopy to: Dr. Roy William Mayega –
Instructional Designer/Materials Editor – MPH Distance education Programme: e-mail:
de_materials@musph.ac.ug .

137