The Social Psychology of Exercise and Sport 2005 (For Lydia

APPLYING SOCIAL PSYCHOLOGY
SERIES EDITOR: STEPHEN SUTTON

THE SOCIAL PSYCHOLOGY OF EXERCISE AND SPORT
The Social Psychology of Exercise and Sport

This new textbook examines the role that social psychology has in the
explanation of exercise and sport behaviour. It devotes considerable attention
to key social psychological issues within the two disciplines; health-related
exercise behaviour and the behaviour of competitive sport participants and
the spectators of elite sport. The
Rather than presenting a broad, superficial overview of diverse areas in
exercise and sport, the book focuses on a range of selected topics and provides
a comprehensive, in-depth and analytical coverage using social psychology as
a framework. It thoroughly examines how social psychological research and
Social Psychology
intervention has contributed to the understanding of key topics in exercise
and sport behaviour including:
• The social psychology of exercise and health of Exercise
• Social cognitive theories of exercise behaviour
• Exercise and the physical self
• Eating disorders in exercise and sport
• Emotion and mood in athletes and Sport
• Social psychology and motivation in sport
• Group processes in sport
• Aggression and crowd violence
The Social Psychology of Exercise and Sport is key reading for undergraduate
and postgraduate students on social or sport psychology courses and on
health-related or sports science courses. Illustrated throughout with practical
guidelines for researchers and practitioners, it is also a valuable resource for
professionals interested in understanding and changing the behaviour of
exercise participants and athletes.
Cover design: Kate Prentice
Martin Hagger is Lecturer in Social and Health Psychology at the University
of Essex. He is also a qualified Chartered Health Psychologist with the British
Psychological Society (BPS) and an accredited Sports Scientist (Psychology)
with the British Association of Sport and Exercise Sciences. He has diverse
research interests in the fields of health and social psychology.
Nikos Chatzisarantis is a lecturer in the School of Sport and Health Sciences

Hagger • Chatzisarantis
at the University of Exeter. His research interests span the fields of philosophy,
social psychology, and quantitative research methodology.
Martin Hagger and

ISBN 0-335-21618-8
Nikos Chatzisarantis
9 780335 216185
Social psychology of exercise and sport
Applying social psychology
Series editor: Stephen Sutton
Published titles
Richard P. Bagozzi, Zeynep Gürhan-Canli and Joseph R. Priester: The Social

Psychology of Consumer Behaviour
Mark Conner and Christopher J. Armitage: The Social Psychology of Food
Steve Sussman and Susan L. Ames: The Social Psychology of Drug Abuse
Social psychology of exercise
and sport
Martin Hagger and Nikos Chatzisarantis
Open University Press

Open University Press
McGraw-Hill Education
McGraw-Hill House
Shoppenhangers Road
Maidenhead
Berkshire
England
SL6 2QL
email: enquiries@openup.co.uk
world wide web: www.openup.co.uk
and Two Penn Plaza, New York, NY 10121-2289, USA
First published 2005

Copyright © Martin Hagger and Nikos Chatzisarantis
All rights reserved. Except for the quotation of short passages for the purpose
of criticism and review, no part of this publication may be reproduced, stored
in a retrieval system, or transmitted, in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without the prior written
permission of the publisher or a licence from the Copyright Licensing Agency
Limited. Details of such licences (for reprographic reproduction) may be obtained
from the Copyright Licensing Agency Ltd of 90 Tottenham Court Road, London,
W1T 4LP.
A catalogue record of this book is available from the British Library
ISBN 0 335 21618 8 (pb) 0 335 21619 6 (hb)
Library of Congress Cataloging-in-Publication Data
CIP data applied for
Typeset by RefineCatch Limited, Bungay, Suffolk
Printed in Poland by O.Z. Graf S.A.
www.polskabook.pl
Contents
Series editor’s foreword vii

Acknowledgements ix
Introduction 1
Part I The social psychology of exercise 5

1 Social psychology, exercise, and health 7
2 Social cognitive theories of exercise behaviour 15
3 From exercise intention to exercise behaviour and beyond 43
4 Exercise and the physical self 71
Part II The social psychology of sport 97

5 Social psychology and motivation in sport 99
6 Athletes are emotional, too 130
7 Group processes in sport 160
8 Aggression and crowd violence 193
9 Conclusion 215
Glossary 227
Bibliography 233
Index 261
Series editor’s foreword
Social psychology is sometimes criticized for not being sufficiently ‘relevant’ to

everyday life. The Applying Social Psychology series challenges this criticism.
It is organized around applied topics rather than theoretical issues, and is
designed to complement the highly successful Mapping Social Psychology
series edited by Tony Manstead. Social psychologists, and others who take
a social psychological perspective, have conducted research on a wide range
of interesting and important applied topics such as drug use, consumer
behaviour, work, politics, the media, crime and environmental issues. Each
book in the series takes a different applied topic and reviews relevant social
psychological ideas and research. The books are texts rather than research
monographs. They are pitched at final year undergraduate level, but will
also be suitable for students on Masters level courses as well as researchers
and practitioners working in the relevant fields. Although the series has
an applied emphasis, theoretical issues are not neglected. Indeed, the series
aims to demonstrate that theory-based applications of social psychology can
contribute to our understanding of important applied topics.
This, the fourth, book in the series deals with exercise and sport. Both are
topical issues. As the authors note, the profile of regular exercise and sport has
risen in recent years, for different reasons: exercise, because of the links that
have been observed in epidemiological studies between low levels of physical
activity and risk of chronic diseases such as diabetes and heart disease; and
sport, partly because televised sporting events have become a hugely popular
form of entertainment. Social psychological research in the two domains has
addressed somewhat different research questions. With regard to exercise,
the key questions concern the psychological predictors and determinants
of exercise behaviour and how such information can be used to inform
interventions to increase physical activity. By contrast, social-psychological
approaches to sport have attempted to explain individual and team perform-
ance, again with implications for how performance can be enhanced. While
research in the exercise domain has focused on the behaviour of individuals,
viii Series editor’s foreword
work on performance in sport draws on the traditional social psychological

area of intragroup processes to explain team performance. Although the book
addresses the two topics separately, the final chapter provides an illuminating
comparison between social psychological research in the two domains, in
terms of themes, methods and theories. Both authors are active researchers in
the fields of exercise and sport and have an intimate knowledge and obvious
enthusiasm for their subject. In this book, they have succeeded in their aim of
demonstrating the past and potential future contribution of social psychology
to understanding exercise behaviour and sport performance.
Stephen Sutton
Acknowledgements
I dedicate this book to my parents, Mike and Elinor, and my brother Damian
and his wife Mary-Jane for their inspiration and tolerance.
Martin Hagger
I would like to express my thanks to Professor Nikolas Karanikolas from the
Anatolia College of Thessaloniki who kindled my creative instincts during my
teenage years. Most importantly, I would like to extend my gratitude to my
parents, Lazaros and Despoina Chatzisarantis, for support and encouragement.
The authors would also like to express special thanks to Professor Stuart Biddle
for his friendship, supervision, guidance, and encouragement. We would also
like to thank Dr. Elaine Duncan, Professor Nannette Mutrie, Professor Stephen
Sutton, Dr. Joanne Thatcher, and Dr. Mike Weed for their comments on earlier
drafts of this book.
Finally, the author and the publisher would like to thank the following for
granting permission to use material quoted in the text.
Figure 4.2: Source: Fox and Corbin (1989). Copyright © 1985 by Human
Kinetics Publishers, reprinted with permission.
Figure 4.5: Source: Sonstroem and Morgan (1989). Copyright © by Lippincott,
Williams, and Wilkins Publishers Inc., reprinted with permission.
Figure 5.3: Source: Vallerand and Ratelle (2002). Copyright © 2002 by
University of Rochester Press, reprinted with permission.
Figure 5.4: Source: Guay, Mageau, and Vallerand (2003). Copyright © 2002
by Sage Publications, Inc., reprinted with permission.
Figure 6.3: Source: Jones and Hardy (1990: 88). Copyright © 2002 by John
Wiley & Sons Ltd, adapted with permission.
x Acknowledgements
Figure 6.5: Source: Kerr (1985). Copyright © 1985 by Taylor and Francis Ltd,
reprinted with permission.
Figure 7.1: Source: Carron and Hausenblas (1998). Copyright © 2002 by Fitness
Information Technology, reprinted with permission.
Figure 7.2: Source: Carron (1992). Copyright © 1982 by Human Kinetics
Publishers, adapted with permission.
Figure 7.3: Source: Carron, Widmeyer, and Brawley (1985). Copyright © 1985
by Human Kinetics Publishers, reprinted with permission.
Figure 7.5: Source: Beauchamp (2004). Adapted with permission.
Figure 7.6: Source: Aiello and Douthitt (2001). Copyright © 2001 by American
Psychological Association, reprinted with permission.
Figure 8.1: Source: Silva (1980). Copyright © 1980 by Human Kinetics
Publishers, reprinted with permission.
Introduction
The profile of regular exercise and sport in society has risen in recent years. The
links between regular physical activity and physical and psychological health
grow ever stronger and the role of competitive sport as entertainment and
recreation is proliferating. Ever since Ralph Paffenbarger’s seminal paper on
the contribution of occupational physical activity to decreased mortality in
Californian dock workers, epidemiological research has consistently reported
a clear relationship between mortality rate and physical inactivity. Today,
physical activity features high on the list of priority health behaviours for
government campaigns aiming to improve health in industrialized nations.
Sport, on the other hand, has always been a conduit for a nation to express its
identity, but its increasing value for entertainment in the past two decades can
be quantified not only by the observed rises in mass participation, but also in
the earning power of and revenue generated by elite sporting individuals and
teams. As exercise and sport are behaviours conducted in social contexts, social
psychology has a significant role to play in understanding the motivation and
behaviour of people involved in both recreational exercise for health and
competitive sport. As social psychology is the study of human behaviour in
social contexts, much of the investigation into the factors that contribute to
exercise and sport behaviour, and the understanding of the relationships
among these factors have been conducted by applying theoretical approaches
from social psychology.
This book examines behaviour in sport and exercise from the point of view
of social psychology. Principally, the text aims to devote considerable atten-
tion to key social psychological issues within the two disciplines: exercise
behaviour for health reasons and the behaviour of competitive sport
participants and the spectators of elite sport. Rather than presenting a broad,
superficial overview of diverse areas in exercise and sport, the focus of the book
is on a narrow range of selected topics and serves to provide a comprehensive,
in-depth, analytical, and research-focused coverage using social psychology
as a framework. The aim of the book is therefore to provide a thorough
2 Introduction
examination of how social psychological research and intervention have

contributed to the understanding of key topics in exercise and sport
behaviour.
Social psychology, like many disciplines in psychology, has many branches
and sub-systems. This is not surprising, considering the multitude of questions
and social problems that social psychologists try to address and the diverse
methods of research inquiry that social psychologists adopt to investigate
these problems. Social psychology can be considered even more diverse than
other areas of applied psychology because it is informed by research from
other areas of the social and behavioural sciences (Bagozzi et al. 2002). In the
United Kingdom and the United States, social psychology has been treated as a
science and tends to adopt a positivistic approach that is driven by both theory
and hypothesis. The aim is to provide answers to research questions through
quantitative empirical methods that are based on the principle of disconfirm-
ation or falsification. This approach tends to be the dominant in the social
psychological literature, mainly because many of the mainstream peer-
reviewed journals are North American and tend to endorse this approach. This
approach has been labelled psychological social psychology because it
adopts the rigorous scientific approaches used by other sub-systems and
disciplines in psychology. The psychological social psychology tradition is con-
trasted with an approach that focuses on the effects of the broader social con-
text on social action. This sociological approach to social psychology examines
the effects of personal experience, meanings, language, culture, ideology, and
the material or physical environment on the ‘lived experience’ of individuals
in those contexts and, in particular, their relationships with others. The key
unit of analysis in this approach tends to be representations, stereotypes, and
cultural images and how they relate to people’s construction and interpret-
ation of the meaning they attribute to themself and others on the basis of
these broad social influences. Sociological social psychology tends to be pre-
dominant in Europe and adopts a relativist rather than absolute perspective
towards truth and meaning.
The research presented in this book is derived primarily from the psycho-
logical social psychology tradition and the quantitative, hypothesis-testing
methodological framework. This is primarily because it is the dominant trad-
ition in applied research in exercise and sport psychology. However, we aim to
provide a fair treatment of the various approaches and methods used in
research in this area and have therefore included research from the socio-
logical social psychology tradition that adopts a more grounded, theory-
building approach rather than a theory-testing one and adopts qualitative
methods to investigate research questions. We have excluded research from
the skill acquisition literature including motor learning, motor control, and
motor development perspective because most of the research in these areas has
focused on perceptual and neurological explanations of movement far
removed from social influences on human movement. Other exclusions
include more sociolocal approaches to the explanation of exercise and sport
Introduction 3
behaviour. We have focused on the individual as the unit of analysis, as is the

tendency in the psychological social psychological approach. While we
recognize the influence of overarching social factors such as age, socio-
economic status, culture, and ethnic background, these are treated as
peripheral to the influence of personal variables such as personality, beliefs,
emotions, expectations, and judgements on exercise and sport behaviour.
This book is divided into two parts: the social psychology of exercise (Part I)
and the social psychology of sport behaviour (Part II). Part I consists of four
chapters and will focus on the application of social psychological theory to the
explanation of exercise and physical activity participation. The key issues
covered in Part I include the links between exercise, and physical and psycho-
logical health (Chapter 1), the social cognitive theories that have attempted to
explain exercise behaviour (Chapter 2), the theories that aim to convert exer-
cise intention and motivation into exercise behaviour (Chapter 3), and the
role of the physical self in exercise behaviour and psychological disorders relat-
ing to self-perceptions (Chapter 4). We have adopted a step-by-step approach
to the understanding of social psychological theories and investigations in
exercise and how they can lead to interventions to promote exercise behaviour
in the general population. We initially introduce the prevailing theoretical
approaches, the empirical research from the social psychological literature
that has provided support for these theories, and provide a comprehensive
overview on the basis of this research as to how interventions can be designed
to change exercise behaviour in the general population.
Part II consists of four chapters and aims to provide coverage of key issues in
sport-related behaviour. The key areas covered are social psychological
approaches to motivation in sport (Chapter 5), the social psychology of emo-
tion and anxiety in sport (Chapter 6), group processes and social influence on
sport performance (Chapter 7), and aggressive behaviour and crowd violence
in sport contexts (Chapter 8). Again, we aim to introduce the reader to the
pertinent theories adopted by social psychologists to explain key social psy-
chological behaviours in sport, particularly sport performance. At each stage
we aim to provide a series of key target variables and highlight the strategies
that sport psychologists and coaches can adopt to enhance performance in
athletes and sports performers. In each chapter we provide an executive sum-
mary of the pertinent points in the chapter along with some key annotated
readings.
A common limitation of many books that cover psychological research in
both the exercise and sport domains is that they fail to draw any conclusions
on the commonalities between the two areas. This has the effect of reinforcing
the notion that approaches to these areas are entirely different and exercise
psychology and sport psychology should be rigorously compartmentalized
into two entirely different disciplines. Chapter 9 aims to break this tradition
and draws together the common themes and differences in the social psycho-
logical approaches to exercise and sport behaviour. The aim of Chapter 9 is to
point out to the reader that many of the approaches to exercise and sport from
4 Introduction
an applied social psychological perspective have as many similarities as they

do contrasts. In this chapter similarities are covered in terms of themes,
methods, and theories. In terms of themes, prediction of behaviour, affective
outcomes, and social influences are common to both areas. The common
methods adopted to investigate these predictions include cross-sectional
studies, longitudinal, cross-lagged panel designs, experimental designs, and
qualitative approaches. Theories of intention and motivation are common to
social psychological research in both areas. Finally, to illustrate some of the
contrasts inherent to social psychological approaches to exercise and sport, we
review the potential conflict between sport for competition and sport for
health purposes.
There is a glossary of key words and phrases at the back of the book. The first
occurrence of each word in the glossary is shown in bold in the text.
Part I
The social psychology of exercise

1
Social psychology, exercise, and health
This chapter will introduce some of the key concepts relating to physical
activity and health and outline the problems faced by applied social psycho-
logists in the field of physical activity and health. The aim is to provide suf-
ficient background knowledge of the health, social, and economic problems
presented by a sedentary population in industrialized nations. In addition, it
will give an insight into research techniques such as descriptive epidemiology
as a means to evaluate the extent of the epidemic of physical inactivity and
obesity in industrialized countries. Subsequently, the focus will be on the
importance of applied social psychology to inform and drive interventions to
increase the levels of physical activity behaviour in sedentary populations. In
addition, the different social psychological theoretical approaches to physical
activity behaviour will be introduced.
What is physical activity?
People often talk about sport, exercise, and physical activity in an eclectic,
unstructured manner, and occasionally use the terms synonymously. In
everyday life, it seems, the understanding of the distinctions between these
forms of physical endeavour is often unclear. It is therefore essential that these
terms are formally defined before embarking on a discussion of the importance
of physical activity to health and how social psychology can lend itself to an
understanding of these behaviours. Physical activity is typically used to refer
to all types of movement that expends energy, regardless of features such
as type, location, mode, and intensity. Formally, Pate et al. define physical
activity as ‘any bodily movement produced by skeletal muscles that results in
energy expenditure’ (1995: 402). It can therefore be considered an umbrella
term under which other more specific forms of physical activity fall. Exercise
and sport are therefore subtypes of physical activity. Exercise generally refers
to structured physical activity whose purpose is to incur a health benefit such
8 Social psychology, exercise, health
as losing weight. Examples of these types of activity are jogging, cycling,

swimming, rowing, and walking. Sport is another form of physical activity, but
is far more structured than exercise, has specific sets of rules, and generally
involves competition with other people. Exercise can, however, also be less
structured such as that associated with certain occupations or for transport
such as walking to work. Importantly, exercise has a number of properties that
determine the degree of health benefit it can provide, namely, type, intensity,
frequency, and duration. Exercise that involves and stresses large muscle
groups (known as ‘gross’ motor tasks) over a long period of time and is of a
vigorous intensity tends to be favoured by exercise promoters because this
kind of exercise puts a high degree of stress on the cardiovascular system and
has the potential to increase the daily energy expenditure of the individual. If
undertaken frequently enough, such activity can assist in ameliorating risk
from cardiovascular disease, can help promote skeletal and psychological
health, and can reduce the risk of other illnesses such as cancer and type II
diabetes, and may prevent other conditions that may compromise health
such as obesity. Social psychological approaches to understanding exercise
participation will be the subject of the first part of this book.
Evidence for the effect of physical activity on health
Physical activity and chronic disease in adults

Low levels of regular physical activity in the populations of industrialized
nations have been implicated in a number of chronic diseases. These diseases
include various forms of cardiovascular disease, obesity, type II diabetes, cer-
tain types of cancers, and hypertension. Cardiovascular disease, in particular,
is a serious health issue. Recent statistics indicate that cardiovascular disease is
the largest single cause of death in many industrialized nations. For example,
cardiovascular disease accounts for approximately 931,000 deaths per year in
the United States (CDC/NCHS 2001) and 238,000 deaths per year in the
United Kingdom (Petersen et al. 2004). In addition, treatment for such diseases
puts a substantial burden on public health services. For example, treatment
for coronary heart disease, the most prevalent type of cardiovascular disease,
costs the UK National Health Service £1.75 billion per year (Liu et al. 2002).
Alongside cardiovascular disease, incidence of obesity, a clinical term for
excess adipose or fat tissue, is increasing. Obesity also presents a health risk
because it is implicated in many health problems such as cardiovascular dis-
ease, diabetes, colorectal cancer, and hypertension. Obesity is defined using a
measurement that accounts for a person’s height and weight, known as the
body mass index (BMI). This is the ratio of a person’s body mass to their height
squared (weight/height2) and a BMI greater than 30kg/m2 is considered obese.
In the United States, 27.5 per cent of men and 33.4 per cent of women are
considered obese (Centers for Disease Control and Prevention 2002), while in
Social psychology, exercise, health 9
the United Kingdom 22 per cent of men and 23 per cent of women are
classified as obese (National Health Service Health Development Agency
1996). This evidence indicates that cardiovascular disease and obesity are
substantial problems to public health in industrialized nations.
There is a general consensus that multiple key risk factors contribute to the
likelihood of an individual contracting cardiovascular disease, namely, age,
gender, a family history or heredity of stress, tobacco smoking, elevated
blood cholesterol (hypercholesterolemia) or more specifically a serum lipid
profile that has a disproportionately high ratio of low density lipoproteins
(LDL), hypertension or high blood pressure, obesity, type II diabetes, stress,
and physical inactivity (American Heart Association 1999). Clearly, age,
gender, and heredity are demographic variables that are unalterable. However,
many of the other risk factors can be affected by behaviour change and
research has indicated that regular physical activity of the relevant type, inten-
sity, duration, and frequency can substantially alter the levels of these risk
factors and thereby reduce the risk of cardiovascular disease (Wannamethee
and Shaper 2001).
Physical inactivity has been isolated as a risk factor for cardiovascular disease
(Wannamethee and Shaper 2001). Epidemiological studies have shown that
low physical activity and fitness levels are related to all-cause mortality in the
general population (e.g. Paffenbarger and Hale 1975), although physical fit-
ness is controversial as a correlate of cardiovascular and other disease risk
because it only indirectly relates to physical activity participation. Further
epidemiological research has revealed some of the mechanisms behind these
links. Research has shown that regular vigorous physical activity lowers
blood cholesterol and produces a more favourable serum lipid profile, reduces
hypertension, and can help control and reduce the severity of type II diabetes
(Wannamethee and Shaper 2001). Physical activity, along with dietary
manipulation, can also help maintain a healthy level of body fitness and has
been shown to be effective in reducing obesity (Wannamethee and Shaper
2001). This evidence indicates that regular physical activity has profound
effects on cardiovascular health and can reduce the risk factors associated with
cardiovascular disease.
In addition to the ameliorating effect of physical activity on cardiovascular
disease risk factors, research has also implicated physical activity in reducing
the incidence of certain cancers. Recent evidence suggests that regular
vigorous physical activity can reduce the risk of ovarian, breast, and colorectal
cancer (Courneya and Friedenreich 1997). Although researchers are unsure as
to the exact mechanism by which physical activity acts to reduce cancer risk, it
is thought that it may be its efficacy in maintaining a healthy body weight
and that exercisers tend to adopt more healthy behaviours such as eating foods
high in antioxidants.
Regular physical activity has also been implicated in maintaining skeletal
health, particularly in women (Branca 1999). Coupled with a diet that includes
adequate calcium supplementation, regular weight-bearing physical activity
can act as a preventive measure against osteoporosis – a chronic disease that

results in the demineralization of bones and an increased susceptibility to
fractures. Finally, an important effect of regular physical activity is its effect on
psychological and mental health. Regular physical activity has been found
to be an effective treatment for mental health and has been implicated in
positive well-being and self-esteem in the general population (Fox 1999).
Physical activity and young people

Epidemiological and clinical studies have indicated that the occurrence of dis-
eases such as cardiovascular disease and other illnesses linked with low levels
of physical activity in adults is only a manifestation of behavioural patterns
that have paediatric origins. Initial evidence that cardiovascular disease had
paediatric origins was gleaned from autopsies on American infantrymen killed
in action in the Vietnam War that showed the beginnings of atherosclerosis (a
precursor of some forms of cardiovascular disease) in their arteries (McNamara
et al. 1971). Research has shown that the risk factors associated with cardio-
vascular disease in adults are present among children and adolescents and
that these risk factors tend to track into adulthood (Webber et al. 1983). In
addition, rates of childhood or juvenile obesity are increasing with surveys
indicating incidences of 20 per cent in the United States (Flegal 1999) and 9.2
per cent in the United Kingdom (Bundred et al. 2001), figures that represent
substantial increases on previous years. The detrimental health effects of
obesity are the same in adolescents as they are for adults such as hypertension,
elevated blood lipids and high levels of LDLs, and increased risk of diabetes.
Indeed, a study of obese children revealed that 97 per cent had three or more
cardiovascular disease risk factors (Parker and Bar-Or 1991). Importantly, it
seems that obese children and adolescents are likely to become obese adults
and this highlights the need to control this through diet and exercise.
Indeed, research in the past two decades has established the importance of
physical activity to children’s cardiovascular health (Sallis and Patrick 1994).
There is evidence to suggest that children who exhibit higher levels of physical
activity and fitness are less likely to have high levels of cardiovascular risk
factors, although there is only limited evidence that physical activity is related
to hypertension and serum lipid profiles of young people (Raitakari et al.
1994). Aside from the amelioration of cardiovascular disease risk factors,
increased physical activity in children is also associated with other health
benefits. Higher levels of regular physical activity are associated with lower
levels of juvenile obesity (Parker and Bar-Or 1991). Physical activity has been
shown to promote skeletal health and aerobic fitness, have a positive effect on
variables related to psychological health such as depression, anxiety, stress
hostility, anger, and intellectual functioning and can also enhance self-esteem
and overall well-being (Sallis and Patrick 1994). Such findings have compelled
researchers in the behavioural sciences to further investigate the antecedents
of children’s physical activity behaviour.
Guidelines for physical activity

Converging evidence from these studies has highlighted the importance of
physical activity to health and has forced exercise scientists to propose guide-
lines for physical activity and nutrition. National bodies and advisory groups
have published position statements and guidelines to appropriate levels of
physical activity to reduce risk of the diseases that are associated with low
levels of activity and to maintain positive health and well-being. For adults,
numerous guidelines have been issued stating the appropriate frequency,
intensity, type, and duration of physical activity necessary for good health and
these have been the focus of various campaigns to promote physical activity.
The content of these guidelines is relatively uniform across nations and
research bodies. The consensus drawn from these guidelines and recom-
mendations from the United States (e.g. Byers et al. 2002) and the United
Kingdom (e.g. Department of Health 1996) is that individuals should engage
in continuous physical activity of at least moderate intensity for a period of 30
minutes on most days of the week, and preferably engage in bouts of higher-
intensity vigorous physical activity on some days of the week. In addition,
similar guidelines have been issued for special populations such as the elderly
(American College of Sports Medicine 1998) and adolescents and children
(Sallis and Patrick 1994).
The physical inactivity epidemic

Given the abundance of evidence to support the links between regular and
vigorous physical activity and cardiovascular, skeletal, and mental health,
research has also focused on the descriptive epidemiology of physical
inactivity among populations in industrialized nations. Such an investigation
aims to outline the extent to which people attain the recommended levels of
physical activity associated with good health. Results from such studies have
consistently indicated that there is an epidemic of inactivity among these
populations. Evidence from national health surveys in the United States
(Centers for Disease Control and Prevention 2003) and the United Kingdom
(National Centre for Social Research 1999) show that approximately 30 per
cent of people did not participate in any physical activity. In England, only 37
per cent of men and 25 per cent of women meet the recommended guidelines
for physical activity (Joint Surveys Unit 1999). A survey of 15,339 consumers
from the 14 European Union states (approximately 1,000 adults from each
member state) found a large variation across national groups in the percentage
of people who engaged in no physical activity, ranging from 1 per cent to
47.6 per cent (Institute of European Food Studies 1999). However, the percent-
age of people engaging in regular physical activity ranged from 92 per cent in
Finland to 60 per cent in Greece. In addition, there is also evidence to suggest
that the majority of children in the United States (Centers for Disease Control
and Prevention 2003) and the United Kingdom (Cale and Almond 1992) do
not engage in sufficient physical activity of the type, intensity, duration, and
frequency likely to bring about health benefits.
Alongside these statistics that seem to indicate low levels of regular physical
activity among many industrialized nations, surveys have also suggested that
the majority of people believe that physical activity is important to health (UK
Health Education Authority and Sports Council 1992) but only about 50 per
cent of people surveyed agreed that they needed to do more physical activity
than they currently did (Institute of European Food Studies 1999). Of great
concern is the largely static percentage of physical inactivity over the past two
decades. Population studies have shown that the level of physical inactivity
remained largely unchanged between 1986 and 2002 (Centers for Disease
Control and Prevention 2003). In summary, the majority of adults in indus-
trialized nations do not engage in sufficient physical activity to gain the health
benefits suggested by research, and levels of physical inactivity have remained
relatively constant, while, paradoxically, the majority of people believe they
do engage in sufficient physical activity.
Given this evidence, researchers have sought to examine the factors that
affect physical activity behaviour. The importance of identifying these factors
is paramount as they can tell on the basis of effective, theory-based interven-
tions the factors that are most subject to change and will have the greatest
influence on physical activity behaviour (Brawley 1993). Indeed, it is these
kinds of social problems that have benefited from research in applied social
psychological theory because these approaches aim to explain the problem,
presenting both an identification of salient influences as well as an explan-
ation of how those influences can affect behaviour. The next section will
outline how research in social psychology can assist in identifying salient fac-
tors and help inform interventions to promote exercise among predominately
sedentary populations.
The role of social psychology
One way to promote exercise participation is to consult theories of social

behaviour that identify the salient antecedent variables and mechanisms
underpinning the motivation of social behaviour, and then test these theories
in the domain of physical activity (Brawley 1993). Theories provide an
explanatory system as to how the antecedents of physical activity behaviour
affect behaviour and provide a general guide as to how physical activity
behaviour can be promoted. For example, if a theory identifies attitudes as an
important determinant of behaviour, then it can be suggested that physical
activity behaviour can be promoted by changing attitudes.
However, it is important to realize that theories from social psychology
can guide promotion of physical activity participation only if they explain
physical activity participation satisfactorily (Brawley 1993). Unfortunately,
empirical evidence suggests that theories from social psychology cannot
explain more than 50 per cent of variance in physical activity behaviour

(Hagger et al. 2002b). Such levels of prediction obtained by social psycho-
logical models are far below the levels of prediction obtained by models in the
pure sciences, and suggest that it would be naïve to expect psychological
interventions to be entirely effective in promoting physical activity participa-
tion. Therefore, from the perspective of social psychology, the promotion of
physical activity participation should be viewed as an ongoing process that
consists of identification of antecedents of physical activity participation
(formative research) and of applied research that evaluates utility of
interventions in promoting physical activity behaviour.
Formative research aims at the identification of the most important psycho-
logical variables that underline exercise participation (Ajzen 1991). Social
psychologists can contribute a great deal to the development of formative
research through a variety of different study designs such as cross-sectional,
longitudinal, panel, and experimental studies. Such studies aim to quantita-
tively evaluate the extended, modified, or refined versions of existing social
psychological models in predicting exercise participation (Hagger et al. 2002b)
through the principle of converging evidence. However, it is important to
distinguish between variables and psychological models that only predict
exercise participation and those that predict and explain exercise participation
(Ajzen and Fishbein 1980). Variables and theories that predict participation in
physical activities are only useful in identifying those who exercise and those
who do not. While such information is important in identifying sub-
populations who are at risk, they may not help explain why some people
exercise and others do not. Explanation of exercise participation is only
achieved when exercise participation is predicted by variables that can be
manipulated. For example, it is widely accepted that attitudes can change
(Eagly and Chaiken 1993), and therefore predictions obtained by attitudes
provide information about behavioural change. Therefore, development of
models of exercise participation should focus on the identification of variables
that both predict and explain exercise participation. In addition to formative
research, promotion of exercise participation can ultimately benefit from
applied social psychological research that evaluates interventions in promot-
ing active lifestyles (Brawley 1993; Hardeman et al. 2002). Intervention studies
have the potential to evaluate whether manipulations of social psychological
variables actually can change exercise habits.
Part I of this book reviews some leading social psychological theories applied
to the explanation of volitional behaviours such as exercise. It also focuses
on the limitations of these theories and the measures taken to address these
limitations. This is important because these potential solutions to resolve
these limitations can produce modified versions of the existing theories to
strengthen the efficacy of the theory to explain and promote exercise
participation.
Suggested reading
Brawley, L.R. (1993) The practicality of using psychological theories for exercise and
health research and intervention, Journal of Applied Sport Psychology, 5, 99–115.
Provides an introductory insight into the utility of social cognitive approaches to
change exercise behaviour.
Cale, L. and Almond, L. (1992) Physical activity levels of school-aged children: a review,
Health Education Journal, 51, 192–7. Provides details on the research suggesting an
epidemic of physical inactivity among young people.
Pate, R.R. (1995) Recent statements and initiatives on physical activity and health,
American Academy of Kinesiology and Physical Education, 47, 304–10. An overview of
the guidelines offered by health campaigns to promote exercise and physical activity
for health.
Summary
• Physical activity refers to all bodily movement that expends energy.

Exercise is formal physical activity expressly to expend energy for health
reasons and sport often involves some form of physical activity, but is
engaged in for a number of different reasons such as competition and
demonstration of competence.
• Physical inactivity is a risk factor for cardiovascular disease and has
numerous psychological and health benefits for adults and young people.
Exercise scientists recommend people engage in at least 30 minutes of
moderate physical activity per day, accompanied by some bouts of vigorous
physical activity.
• Formative research in social psychology can provide useful insight into the
mechanisms behind the antecedents of physical activity and can help
inform interventions applied to physical activity behaviour.
2
Social cognitive theories of exercise

behaviour
Understanding regular participation in exercise requires an understanding of

the psychological constructs that predict and explain exercise behaviour
(Ajzen and Fishbein 1980). This chapter aims to introduce and review the
formative and applied research in social psychology on theories of social cog-
nition and exercise behaviour. Common to these theories is the inclusion of
belief-based constructs such as attitudes and motivational constructs such as
intentions that are learned from previous experience. Another important
feature of these models is that they focus on the formation of motivation and
the processes that lead to intentions. They do not contend with the more
automatic processes that give rise to intentions nor do they aim to explain the
mechanisms by which intentions are converted into behaviour. These will be
dealt with in Chapter 3. Finally, the chapter will outline some of the practical
guidelines that arise from research with these models.
Intentions and social cognitive models of exercise behaviour
Many important social psychological theories of human motivation incorpor-

ate the construct of intention. Such theories propose that people are rational
decision-makers who choose to engage in the target behaviour by processing
the available information relating to the advantages and disadvantages associ-
ated with that behaviour. Theories of intentional behaviour also share the
view that human motivation is unidimensional (Deci and Ryan 1985) and that
the construct of intentions, which represents motivation, is the most immedi-
ate determinant of human action (Ajzen 1991). Importantly, these theories
tend to focus on intention formation and do not offer any explanation for the
processes that underpin the enactment of previously formed intentions (Ajzen
2002b). Volitional theories, which are reviewed in the next chapter, detail the
processes by which intentions are translated into actions (Gollwitzer 1999). In
the next section, theories of intentional behaviour are introduced, and their
16 Social cognitive theories of behaviour
implications for future research and for practice in the exercise domain are
discussed.
The health belief model

The health belief model proposes that an individual’s readiness (intention) to
perform a health behaviour is a function of his or her perceived vulnerability
to a health condition and the probable severity of that condition (Rosenstock
1974). The model posits that readiness is determined by a person’s beliefs
about the benefits to be gained by a particular behaviour such as exercise,
weighted by their perceived barriers to doing that behaviour. Finally, the
model predicts that readiness may not result in overt action unless some insti-
gating event occurs to set the action process in motion. Rosenstock called such
instigating events cues to action. Overall, the health belief model predicts that
if a person feels vulnerable to an illness, and if it is perceived to be severe, and if
he or she believes that a particular health behaviour will reduce the health
threat associated with that illness, then he or she will have a high degree of
readiness to engage in the health behaviour. For example, individuals may feel
susceptible to cardiovascular disease because they have a poor diet and have
been told by their doctor that they have hypertension. They may also believe
that regular exercise will reduce the threat of cardiovascular disease. According
to the model, these perceptions are likely to motivate the individual to partici-
pate in exercise behaviours. The health belief model is summarized in
Figure 2.1.
Figure 2.1 The health belief model

Social cognitive theories of behaviour 17
Formative research has shown that perceived severity and beliefs about the
benefits of a health behaviour exert strong influences on readiness to engage
in that behaviour, while perceived severity and barriers have smaller roles
(Quine et al. 1998; Abraham et al. 1999). In addition, evidence suggests that
the direct effects of perceived vulnerability, severity, susceptibility, benefits,
and barriers on health behaviours are small and are mediated by readiness
(Abraham et al. 1999). There is also evidence to suggest that the health belief
model does not sufficiently capture all the psychological determinants of
social behaviour and that the model might benefit from considering the effects
of other constructs such as self-efficacy on intentions and behaviour.
One limitation of the health belief model is that it does not offer clear
operational definitions of its psychological constructs such as perceived
vulnerability, nor does it specify how different variables can combine in
influencing intentions and behaviour (Quine et al. 1998). For example, it is
unclear which particular behaviours are facilitated by perceived vulnerability
to an illness. In the case of a perceived vulnerability to cardiovascular disease,
the model does not explicitly state whether vulnerability would facilitate
exercise or dieting behaviour, given that both behaviours would be effective
in ameliorating cardiovascular disease risk. As a result, empirical evidence
related to the health belief model varies greatly across studies because differ-
ent studies have used different operational definitions for psychological
constructs (Harrison et al. 1992).
Protection motivation theory

The protection motivation theory (Rogers 1983) is similar to the health belief
model and advocates that the performance of health behaviour is a function of
two distinct appraisals: threat appraisals and coping appraisals with respect to
an illness that poses a health threat (see Figure 2.2). As in the health belief
model, threat appraisals are derived from two sets of beliefs: perceived vulner-
ability and perceived severity. Perceived vulnerability is the person’s belief that
they are vulnerable to the health threat and perceived severity is the person’s
belief that the occurrence of the disease will have severe consequences. Coping
appraisals comprise beliefs that a given behaviour will be effective in reducing
the health threat, known as response efficacy, beliefs that one possesses the
necessary capabilities to perform the health behaviour, called self-efficacy
beliefs, and beliefs regarding the perceived costs associated with performing
the health behaviour (Rogers 1983). Overall, protection motivation theory
predicts that individuals will adopt a health behaviour: (1) if they believe a
disease to be severe and likely to occur; and (2) if they perceive that health
behaviour is effective in reducing the health threat, something that they feel
capable of doing, and if it does not cost a lot.
Formative research has documented the utility of threat appraisals and
coping appraisals such as self-efficacy in predicting health-related intentions
and behaviours (Hodgkins and Orbell 1998; Milne et al. 2000). In addition,
Figure 2.2 The protection motivation theory
there is evidence that corroborates the view that threat appraisals can facilitate
maladaptive coping responses (Hodgkins and Orbell 1998). In the exercise
domain, formative research has also documented the utility of self-efficacy,
response efficacy, and perceived vulnerability in influencing exercise intentions
(Milne et al. 2002). Most important, intervention studies have found effects
from coping appraisals and threat appraisals on intentions (Milne et al. 2002).
However, it is important to note that effects of coping appraisal and threat
appraisal on intentions and behaviour have been modest (Milne et al. 2000).
Designing interventions based on the health belief model and the

protection motivation theory
As previously noted, an important function of formative research is to provide
information about the content of interventions. Generally speaking, the
greater the relative importance of a factor in predicting intentions, the more
likely it is that changing that factor will influence intentions and ultimately
behaviour. Given that studies adopting the health belief model and the protec-
tion motivation theory have shown that appraisals related to the health threat
such as perceived vulnerability and perceived severity, and appraisals related
to coping strategies such as self-efficacy and response efficacy, can influence
intentions to exercise, it is suggested that attempts to change exercise
behaviour should try to manipulate threat appraisals and coping appraisals
(Milne et al. 2000). An important question, therefore, is how health appraisals
and coping appraisals can be influenced.
Threat appraisals can be manipulated through fear-arousing communica-
tions emphasizing (1) the painful and debilitating effects of an illness, i.e. its
perceived severity; and (2) that people who do not exercise regularly are
vulnerable to heart disease, i.e. perceived vulnerability. Coping responses can
be manipulated by providing people with information that explains the
effectiveness of exercise in preventing the disease, that presents an inter-
vention to change response efficacy, and by prompting people to choose to
perform types of exercise that are relatively easy to execute that will alter
self-efficacy. One caveat of interventions based on the health belief model
and the protection motivation theory is that, although they may be successful
in strengthening intentions, they may not always bring substantial changes in
exercise behaviour (Milne et al. 2000). Therefore, it cannot be expected
automatically that application of these theories will have a strong effect on
exercise behaviour. Instead, the effectiveness of the protection motivation
theory in changing exercise behaviour may be enhanced through the imple-
mentation of volitional techniques that can help people translate intentions
into actions. Another limitation of the health belief model and the protec-
tion motivation theory is that the threatening messages can sometimes
undermine rather than enhance intentions (Hodgkins and Orbell 1998).
Therefore, fear-arousing communications should be designed and applied
with caution.
The theory of reasoned action

The theory of reasoned action (Ajzen and Fishbein 1980) is one of the most
influential and oft-cited models of intentional behaviour (Sheppard et al.
1988). According to this theory, the performance of volitional behaviours such
as exercise can be best predicted from a person’s stated intention to participate
in the behaviour. Ajzen and Fishbein (1980) hypothesized that intention indi-
cates the degree of planning a person puts into their future behaviour and
represents how hard people are willing to try and how much effort they expect
to exert in the performance of behaviour. It is therefore a construct that is
motivational in nature and function, much like traditional conceptualizations
of intention (Meiland 1970). Intention is assumed to be the most immediate or
proximal antecedent of behaviour (Ajzen and Fishbein 1980). Intention is in
turn a function of a set of personal and normative perceptions regarding the
performance of the behaviour, the attitudes and the subjective norms, respect-
ively. Attitudes represent an overall positive or negative evaluation towards
the target behaviour. Subjective norms represent perceived influences that sig-
nificant others may exert on the execution of the behaviour. Generally speak-
ing, the theory of reasoned action predicts that the more favourable an indi-
vidual’s attitude and subjective norm, the stronger his or her intentions to
perform the behaviour. Finally, intentions are hypothesized to lead directly to
behavioural engagement and intentions are proposed to mediate the effects of
attitudes and subjective norms on behaviour. This means that intentions
explain the attitude–behaviour and subjective norm–behaviour relationships.
Intentions are therefore necessary to convert attitudes and subjective norms
into behaviour.
The theory also deals with the antecedents of attitudes and subjective
norms. The theory proposes that attitudes arise out of a combination or
composite of the actor’s beliefs that behaviour will lead to certain outcomes,
known as behavioural beliefs, and their evaluation of these outcomes,
termed outcome evaluations (Ajzen and Fishbein 1980). Similarly, the
origins of subjective norms can be traced to corresponding belief-based
judgements that include normative beliefs and motivation to comply.
Normative beliefs refer to behavioural expectations that important referent
individuals or groups will approve or disapprove of the behaviour (Ajzen and
Fishbein 1980) and motivation to comply is the actor’s general tendency to
go along with the wishes of the salient referents. This relationship between
behavioural and normative beliefs and their respective evaluations is
grounded in the expectancy × value model, and the belief-based measures of
attitudes and subjective norms are considered antecedents of the directly
measured attitude and subjective norm constructs. The theory of reasoned
action is shown in Figure 2.3.
The major hypothesises of the theory of reasoned action have been sup-
ported in numerous studies of a number of different behaviours (Sheppard et
al. 1988), including exercise (e.g. Hausenblas et al. 1997; Hagger et al. 2002b)
Figure 2.3 Theories of reasoned action and planned behaviours

Note: Constructs and relationships above the broken line represent the theory of
reasoned action. Constructs above and below the broken line together represent the
theory of planned behaviour.
and sports training (Theodorakis et al. 1991b). In the exercise domain, tests of
the theory of reasoned action have provided strong evidence for the overall
predictive value of intentions and have shown that attitudes have a pervasive
effect on intentions with a lesser role for subjective norms (Hausenblas et al.
1997; Hagger et al. 2002b). Formative research adopting the theory of reasoned
action in the exercise domain has also shown that the typically large effect of
attitudes and small effect of subjective norms on intentions are not influenced
by whether belief-based or direct measures of attitudes and subjective norms
have been used (Theodorakis et al. 1991a). Moreover, panel studies have indi-
cated that the strong effects of attitudes on intentions remain stable over time
(Chatzisarantis et al., in press a; Hagger et al. 2001).
Equally important have been applications of the theory of reasoned action
that reveal a number of salient behavioural and normative beliefs related to
exercise. The beliefs are typically elicited from open-ended questionnaires
administered to a pilot sample prior to the development of standard question-
naires to measure the belief-based constructs of theory of reasoned action
(Ajzen and Fishbein 1980). Behavioural beliefs identified in this research
include: ‘good companionship’, ‘weight control’, ‘benefit my overall health’,
‘take too much time’, ‘fun’, ‘get fit’, ‘stay in shape’, ‘improve skills’, ‘get an
injury’, and ‘makes you hot and sweaty’ (Hagger et al. 2001). Important refer-
ents for the normative beliefs and motivation to comply measures tend to be
family members such as parents, grandparents, and siblings along with friends
and schoolteachers (Hagger et al. 2001). However, these beliefs have not been
shown to unequivocally account for unique variance in the directly measured
attitude and subjective norm constructs and alternative subsets of beliefs may
exist (Hagger et al. 2001).
Boundary conditions of the theory of reasoned action

Although research has consistently supported the validity of the theory of
reasoned action (Sheppard et al. 1988), three conditions, outlined by Ajzen
and Fishbein (1980), limit the utility of the theory of reasoned action in
predicting and explaining behaviour. These are the conditions of correspond-
ence, stability, and volitional control. The boundary condition of correspond-
ence states that the predictive efficacy of the theory of reasoned action
variables such as intention is improved when measures of intentions and
measures of behaviour are phrased in a way that correspond in four key ways:
action (e.g. exercise); target at which action is directed (e.g. four times per
week); time at which the behaviour is performed (e.g. over the next three
months); and context in which the action is performed (e.g. during leisure
time). If the measures of constructs such as intentions do not correspond to
the behaviour in one or more of these elements, then, according to the prin-
ciple of correspondence, the predictive validity of intentions will decrease. The
second boundary condition states that intentions will predict behaviour
provided that intentions have not changed before the behaviour is observed.
However, intentions are more likely to change the longer the time interval
between assessments of intentions and behaviour. This is due to the increased
likelihood that new information will be made available to the individual
which would be likely to change or modify intentions. The third boundary
condition outlined by the theory of reasoned action is also related to pre-
dictive validity of intentions. This condition proposes that intentions will
only effectively predict behaviours that are under the complete volitional
control of the individual (Ajzen and Fishbein 1980). Generally speaking,
behaviours are under complete volitional control when they can be performed
at will, and are not facilitated or impeded at all by personal (e.g. low ability)
and/or environmental (e.g. lack of time) factors (Ajzen and Fishbein 1980;
Ajzen 1985). When other factors influence the performance of a behaviour and
the individual cannot control the impact of those other factors on behaviour,
intentions may be less effective as predictors of behaviour.
The theory of planned behaviour

Although the theory of reasoned action has been successful in predicting and
explaining participation in physical activities, the boundary condition of com-
plete volitional control was considered unrealistic and a major limitation of the
theory (e.g. Ajzen 1985). To counter this criticism, Ajzen proposed a theory of
planned behaviour to account for behaviours that are not under complete
volitional control. As in the original theory of reasoned action, the theory of
planned behaviour proposes that intention is a central determinant of social
behaviour and that intention is a function of attitudes and subjective norms
with corresponding behavioural beliefs and normative beliefs respectively.
Further, as in the reasoned action theory, the theory of planned behaviour
acknowledges that intentions explain behaviour only when it is under

complete volitional control. However, the theory of planned behaviour
also proposes that when perceived control over behaviour is problematic, an
additional factor, termed perceived behavioural control, can influence
performance of behaviour (Ajzen 1985).
For Ajzen (1991), the construct of perceived behavioural control refers to
general perceptions of control. He compared it to Bandura’s (1977) construct
of self-efficacy that captures judgements of how well one can execute vol-
itional behaviours required to produce important outcomes. The construct of
perceived behavioural control is also underpinned by a set of control beliefs
and the perceived power of these beliefs (Ajzen and Fishbein 1980). Control
beliefs refer to the perceived presence of factors that may facilitate or impede
performance of behaviour, and perceived power refers to the perceived impact
that facilitative or inhibiting factors may have on performance of behaviour
(Ajzen 1991). In the same way that an expectancy × value model is used to
form indirect antecedents of attitudes and subjective norm, an indirect meas-
ure of perceived behavioural control can be formed from the composite of
each control belief multiplied by its corresponding perceived power rating
(Ajzen 1991).
The inclusion of perceived behavioural control in the theory of reasoned
action is important because it reveals the personal and environmental factors
that affect performance of behaviour (Ajzen 1985). To the extent that per-
ceived behavioural control influences intentions and behaviour, the
researcher can evaluate which behaviours are under the volitional control of
the individual and the degree to which the behaviour is impeded by personal
and/or environmental factors. Ajzen (1991) hypothesized that when control
over the behaviour was problematic, perceived behavioural control would
exert two types of effects within the theory of planned behaviour. First, per-
ceived behavioural control would influence intentions alongside attitudes and
subjective norms. This additive effect reflects the motivational influence of per-
ceived control on decisions to exercise. For example, an individual expressing
positive attitudes and subjective norms toward exercise but low perceived
behavioural control is unlikely to report stronger intentions to exercise than
an individual reporting the same positive attitudes and subjective norms but
higher levels of perceived behavioural control. Second, perceived behavioural
control may predict behaviour directly especially when perceptions of
behavioural control are realistic. This reflects the effect of actual, real con-
straints or barriers to doing the behaviour. In this case perceived behavioural
control is a proxy measure of actual control over the behaviour (Ajzen 1991).
These relationships are shown in Figure 2.3.
A number of studies have shown the theory of planned behaviour to be
superior to the theory of reasoned action in predicting and explaining vol-
itional behaviour (Armitage and Conner 2001). For example, meta-analytic
studies (Armitage and Conner 2001) and narrative review studies (Conner and
Armitage 1998) have supported the consistent effect of perceived behavioural
control on behavioural intentions. In addition, it has been shown that the

effect of perceived behavioural control on intentions is robust and that it is
not influenced by questionnaire format or social desirability (Armitage and
Conner 1999a). Further, formative research has shown that the effect of per-
ceived behavioural control on intentions is particularly pronounced for
behaviours that are difficult to implement, are not under the control of
volitional processes and for participants who do not believe that they have
the ability to control behaviour and outcomes (e.g. Sheeran 2002). Finally, a
number of researchers have shown that the effects of perceived behavioural
control on intentions remain stable over time (Chatzisarantis et al. in press a;
Hagger et al. 2001). Notwithstanding this corroborating evidence, a number of
studies have demonstrated that, in comparison to the effect of perceived
behavioural control on intentions, the direct effect of perceived behavioural
control on behaviour is small (Armitage and Conner 2001), and statistically
non-significant when belief-based measures of behavioural control are used
and when perceptions of control are not realistic (Notani 1998).
In the exercise domain, studies have supported the hypotheses of the theory
of planned behaviour in a variety of exercise contexts (Hagger et al. 2002b).
Evidence suggests that the theory of planned behaviour has a greater pre-
dictive value than the health belief model and protection motivation theory
(Quine et al. 1998). In addition, a number of studies have shown that per-
ceived behavioural control predicts both intentions and behaviour equally
well (Hagger et al. 2002a). Formative research has demonstrated that perceived
behavioural control moderates the attitude–intention relationship so that atti-
tudes predict intentions only among individuals who report high levels of
perceived behavioural control (Ajzen 1991).
A number of studies have also identified certain control beliefs including
barriers and facilitating factors related to exercise such as ‘bad weather’, ‘age’,
‘heart pain’, ‘costs’, ‘fatigue’, ‘no time’ (e.g. Godin et al. 1991). As with
behavioural and normative beliefs, studies have demonstrated that control
beliefs vary considerably across different populations and behaviours. For
example, studies in the exercise domain have identified ‘age’ and ‘fear of
having a heart attack’ among the control beliefs for older and clinical popula-
tions (Godin et al. 1991), but these beliefs do not feature among the control
beliefs of younger populations.
Designing interventions based on the theories of reasoned action and

planned behaviour
Empirical research findings from studies adopting the theories of reasoned
action and planned behaviour have informed the development of interven-
tions to promote exercise behaviour (Brawley 1993; Hardeman et al. 2002).
Considering that formative research has shown that the theory of planned
behaviour is superior to other models of social cognition such as the health
belief model, protection motivation theory, and the theory of reasoned action
(Hausenblas et al. 1997; Quine et al. 1998; Hagger et al. 2002b), it can be
suggested that exercise interventions should target attitudes and perceived
behavioural control. Interventions should not target subjective norms or
psychological variables espoused by the health belief model and protection
motivation theory because these constructs have minimal impact on
intention and behaviour.
The theories of reasoned action and planned behaviour propose that atti-
tudes and perceived behavioural control can change by modifying their under-
lying belief systems. According to Ajzen (1991), changes in beliefs are most
likely to produce demonstrable changes in attitudes, subjective norms, and/or
perceptions of control only when the modal salient beliefs are targeted. Modal
beliefs are the most frequently cited beliefs regarding the behaviour elicited
from the target population. Modal salient beliefs can be identified by using
open-ended questionnaire techniques that require individuals to recall and list
beliefs about the target behaviour (Ajzen and Fishbein 1980).
Once salient beliefs have been identified, the theories of reasoned action and
planned behaviour propose persuasive communications in the form of
pamphlets, face-to-face discussion, observational modelling, or any other
applicable method that can be gainfully employed in designing interventions
(Ajzen and Fishbein 1980). Persuasive communications are appeals that
involve arguments endorsing the positive aspects and outcomes of the target
behaviour while at the same time downplaying the negative aspects. One of
the key contributions of research adopting the theory of planned behaviour is
concerned with the proposition that arguments are most persuasive when
they are centred on accessible beliefs rather than non-accessible beliefs.1 For
example, a persuasive appeal that aims to change the attitudes of adolescents
towards exercise may take the form of the following text that highlights the
advantages and downplays disadvantages of exercise:
Participating in regular exercise has many benefits. You might learn how
to play a new game or sport as well as improving your general level of
fitness and well-being at the same time. Exercise can be a great fun. It
does not necessarily cause injuries or make you feel uncomfortably hot
and sweaty if you exercise at an intensity you feel comfortable with.
This persuasive appeal should be effective in changing young people’s

attitudes because it targets the accessible behavioural beliefs of young people
identified in formative research. Unfortunately, in the exercise domain, few
studies have examined the utility of the theory of planned behaviour in
promoting participation in physical activities and many have produced incon-
sistent results with significant limitations. For example, Rodgers and Brawley
(1993) administered an intervention to produce changes in the affective
attitudes and self-efficacy and this resulted in concomitant increases in par-
ticipation in an exercise programme. Smith and Biddle (1999), on the other
hand, reported that an intervention targeting salient beliefs was not successful
in changing attitudes, subjective norms, perception of control, and inten-

tions (see also Courneya et al. 2002). Such inconsistent results may be due to
interventions targeting non-accessible behavioural beliefs. Therefore, one
important avenue for future research is to examine the utility of intervention
programmes that target accessible versus non-accessible beliefs in promoting
exercise behaviour.
However, persuasive appeals may not be effective in changing intentions if
mean scores of the predictors such as attitudes and perceived behavioural
control are very high (Ajzen 1991). For example, an intervention that targets
attitudes in a population that already tends to have very high levels of atti-
tudes is unlikely to influence intentions simply because there is not much
room for change. Ideally, therefore, research applications of the planned
behaviour should target important predictor variables that do not display very
high mean scores.
Although the theories of reasoned action and planned behaviour provide
general guidance on how to change intentions and behaviour, they do not
indicate how persuasive messages should be structured so that maximum
effects can be obtained. Practitioners need to consult theories of persuasion in
optimizing the effectiveness of persuasive messages. The elaboration likeli-
hood model is a very popular model of persuasion that has been extensively
used in studies of attitude change (Petty and Cacioppo 1986). The elaboration
likelihood model posits that there are two routes to attitude change. First,
there is a central route in which attitude changes as a result of thoughtful
elaboration processes. Elaboration refers to the extent to which individuals are
motivated to think carefully about the arguments contained in a persuasive
communication. Elaboration can be identified by counting the number of
thought processes that a person generates during information processing and/
or by evaluating the quality of arguments presented in the persuasive com-
munication (Petty and Cacioppo 1986). Second, there is a peripheral route to
attitude change, which occurs as a result of non-thoughtful processes such as
inference, associative, and heuristic approaches. The elaboration likelihood
model predicts that attitude change that arises from the central route demon-
strates greater temporal persistence and resistance to persuasion than attitude
change that arises from the peripheral route.
The elaboration likelihood model has been extensively used in studies of
social attitudes (Petty and Cacioppo 1986) but very few studies have used it in
conjunction with the theory of planned behaviour. Recently, Quine, Rutter,
and Arnold (2001) showed that a series of persuasive messages that motivate
young cyclists to elaborate accessible beliefs was successful in changing
behavioural beliefs and normative beliefs but not control beliefs related to
wearing a helmet. Therefore, an interesting avenue for future research is
to design exercise interventions that take into consideration assumptions
underlying the elaboration likelihood model and assumptions underlying the
theory of planned behaviour. The elaboration likelihood model complements
the theory of planned behaviour: the theory of planned behaviour helps
identify accessible beliefs and the elaboration likelihood model indicates how
arguments that are based on accessible belief can be structured and communi-
cated so that maximum and more lasting effects on attitude change can be
achieved.
Another limitation of interventions that are based on the theories of
reasoned action and planned behaviour is that they can only be directed at
individuals who do not intend to perform a behaviour rather than at those
who have already formed intentions. For example, interventions based on
these theories can produce positive intentions among non-intenders by
changing behavioural beliefs, normative beliefs, and/or control beliefs but
cannot help people convert an already strong intention into behaviour.
This is because the theories are motivational theories that can only facilitate
the formation of intentions. In contrast, volitional theories of intention are
most efficacious in facilitating the enactment of behavioural intentions and
can be gainfully applied alongside interventions based on the theories
of reasoned action and planned behaviour. These theories will be discussed
in Chapter 3.
In summary, the development of interventions based on the theory of
planned behaviour should therefore comprise three interrelated stages. First,
formative research studies should be conducted to identify the most important
determinants of intentions. Second, further research should aim to identify
the salient modal beliefs that underpin these determinants. Third, persuasive
messages that target the salient beliefs should be developed and their efficacy
in changing intentions evaluated.
Sufficiency of the theory of planned behaviour
The theory of planned behaviour focuses mainly on the immediate determin-

ants of intentions and behaviour, but more distal determinants may exist
(Ajzen 1991). Ajzen (1985, 1991) claims that the theory can account for the
most immediate determinants of volitional behaviour, and that more distal
determinants of behaviour are background factors that influence intentions
and behaviour by their impact on behavioural, normative, and control beliefs.
For example, the theory supposes that more generalized constructs such as
personality and other trait-like constructs can influence intentions and
behaviour, but only indirectly via attitudes, subjective norms, and perceived
behavioural control (Ajzen 1991). The immediate, proximal determinants of
intention will therefore mediate the influence of these more distal variables on
intention.
However, formative research has found that these proximal constructs do
not account for all the determinants of intentions and behaviour and
researchers have proposed extensions and modifications to the theory to fur-
ther explain variance in intentions and behaviour (Conner and Armitage
1998; Armitage and Conner 2001). Such suggestions are not new. Ajzen (1991)
insists that the theory is open to additional predictors if it is shown that they
capture a significant portion of unique variance in intentions or behaviour. As a
consequence, a number of constructs have been introduced into the frame-
work of the theory of planned behaviour. These additional constructs and the
modification of the theory will be discussed next.
Frequency of past behaviour and habit

One criticism of the theories of reasoned action and planned behaviour is that
they focus exclusively on deliberative processes and ignore the effect of auto-
matic mental processes on behaviour (Fazio 1990). In general, deliberative
processes are characterized by considerable cognitive work and effort. They
involve the deliberation on the available information and an analysis of costs
and benefits. In contrast, automatic processes facilitate fast decisions and
non-intentional behaviours that are based on simple rules or heuristics that
individuals develop through past experience and observation (Chaiken 1980;
Bargh 1994). Recognizing the possible influence of these non-deliberative
and automatic mental processes on volitional behaviour, researchers have
attempted to control for these effects within the deliberative paradigm of the
theory of planned behaviour. As a consequence, studies have included a habit
measure in the theory of planned behaviour as a means to control for such
influences because habitual behavioural engagement is largely determined by
these automatic mental processes (Ajzen 2002b).
Habit refers to frequently performed acts that are done so often that the
advent of situational cues is sufficient to trigger the behavioural response
automatically without any deliberation or information processing (Bargh
1994). The frequency and consistency of performance of behaviour in the past
are the main ingredients that result in the development of habits. As a con-
sequence, several applications of the theory of planned behaviour have used
self-reported frequency of past behaviour as an indicator of habit (e.g. Godin et
al. 1991; Chatzisarantis et al. 2002; Hagger et al. 2002a, 2002b). These research
applications have shown that past behaviour captures a significant portion of
variance in intentions and behaviour after the theory’s current variables have
been taken into account (for reviews, see Hagger et al. 2002b). This research
evidence therefore corroborates the view that regular exercise is a function of
both deliberative and automatic processes, and that the theory of planned
behaviour is insufficient in capturing these automatic processes. However,
several researchers have questioned the use of past behaviour to control for
these automatic effects as it is very limited as an index of habit (Ajzen 2002b).
Overall, although research evidence has consistently supported strong
effects of frequency of past behaviour on intentions, the nature of this rela-
tionship remains elusive because frequency of past behaviour does not suf-
ficiently represent habit (Bargh 1994; Ajzen 2002b). Some researchers claim
that past behaviour may reflect the influences of other unmeasured variables
on intention (Ajzen 2002b; Hagger et al. 2002b). The effects of past behaviour
on intention may also reflect recent performance of the behaviour. This is

because recent occurrences of an event are more available and accessible in
memory than earlier occurrences (Ajzen 2002b). Past behaviour may also not
be a very precise measure of habit because the consistency with which
behaviour is performed across situations contributes to the development of
habit, and measures of frequency of past behaviour do not take into consider-
ation such cross-situational consistency (Verplanken and Orbell 2003).
Therefore, measures of habit that take into consideration both frequency and
consistency of performance of past behaviour are urgently needed so that
more precise conclusions about the nature of past behaviour effects can be
reached. However, it is important to acknowledge that inclusion of past
behaviour in the theory of planned behaviour is important because it tests the
sufficiency of the theory of planned behaviour. The consistent effect of past
behaviour on exercise intentions and behaviour suggests that the theory of
planned behaviour does not offer a complete explanation of volitional
behaviour (Chatzisarantis and Biddle 1998).
Personality traits
Another key extension of the theory of planned behaviour is the inclusion of
personality traits to predict intentions and behaviour. In general, personality
traits are generalized dimensions of individual differences in tendencies to
show stable patterns of thoughts, feelings, and actions (McCrae and Costa
1996). Recent research in the exercise domain has indicated that personality
traits, such as those identified by McCrae and Costa, can influence exercise
intentions and behaviour (e.g. Conner and Abraham 2001; Courneya et al.
2002; Rhodes et al. 2002a). Although considerable evidence corroborates the
link between personality traits and social behaviour in the general social
psychology literature (e.g. Sherman and Fazio 1983), little is known about the
processes by which personality traits influence behaviour.
Sherman and Fazio proposed that there are at least two processes through
which personality traits can influence social behaviour. The first process is
deliberative (Fazio 1990) or systematic (Chaiken 1980). The deliberative model
proposes that people engage in an analysis of costs and benefits of behaviour
and may base their descision-making on whether the behaviour is congruent
with their global tendencies. For example, the indirect effects of extroversion
on exercise via the mediation of attitudes and intentions in the theory of
planned behaviour implies that personality traits motivate people to compare
various behavioural outcomes in order to decide whether or not a behaviour is
appropriate and suitable (Sherman and Fazio 1983). The second process by
which traits can influence social behaviour is through activation of automatic
pathways (Sherman and Fazio 1983). Research on automaticity in social
psychology literature suggests that traits can be activated spontaneously by
the presence of trait-relevant behavioural cues and results in behaviour con-
sistent with these personality traits (Bargh 1994). Some studies have shown
that personality traits can facilitate non-intentional behaviour in the theory of

planned behaviour. For example, Rhodes and co-workers (Courneya et al.
2002, 2002) illustrated how the activity facet of extroversion predicted
exercise participation over and above intentions. This evidence therefore
supports the view that active individuals engage in unplanned behaviour
perhaps because they are more likely to notice unplanned situations where
the opportunity to be active presents itself. Despite this evidence, it has to
be acknowledged that in comparison to the effects of intentions on exercise
behaviour, the direct effects of traits on exercise behaviour are small (Rhodes et
al. 2002a). However, little is known about how traits influence exercise
behaviour and future research needs to explore the processes through which
values and personality traits influence exercise behaviour.
Attitude-based modifications of the theory of planned behaviour
Attitude strength
The role of attitude strength has been the subject of recent investigations in
the theory of planned behaviour. The general approach suggests that the pre-
dictive efficacy of the attitude construct is limited by the strength of the atti-
tude in the individual’s memory. The strength of an attitude in memory is
determined by several attitudinal properties such as extremity, consistency,
ambivalence, frequency, polarization, and consensus (Eagly and Chaiken
1993). In support of the attitude strength model, several studies have found
that the attitude–behaviour relationship increases in proportion to the
amount of behavioural information available to the individual and the degree
of past experience an individual has with the attitude object or behaviour
(Nederhof 1989). Further, research suggests that the attitude–behaviour rela-
tionship decreases as attitudinal ambivalence increases (Conner et al. 2003).
Nederhof (1989) has demonstrated that attitudes are strong predictors of
intentions and behaviour when people consider the attitude object or
behaviour to be important to them. More recently, studies in the exercise
domain have included an attitude strength construct in the theory of planned
behaviour. Attitude strength was found to mediate the effects of subjective
norms and attitudes on intentions (e.g. Theodorakis 1994). Overall, these data
suggest that the stronger the attitude, the more effective the prediction of
intentions and behaviour.
Affective and cognitive attitudes

Research has examined the importance of the emotional or affective attitudes
and instrumental or cognitive attitudes within the theory of planned
behaviour. Trafimow and Sheeran (1998) have argued in favour of a multi-
dimensional conceptualization of attitudes on the grounds that the unidimen-
sional model ignores the unique effects of the instrumental and emotional
aspects of attitude on behaviour (Eagly and Chaiken 1993). Indeed, studies
have shown that this distinction can be made on the basis of the belief systems
that underpin attitudes (Trafimow and Sheeran 1998). Studies have found that
affective attitudes exert unique effects on intentions for a number of health
behaviours including exercise (Lowe et al. 2002). Interestingly, Lowe et al.
have shown that the affective component exerts a direct effect on exercise
behaviour after controlling for intentions and past behaviour. This evidence
suggests that the affective component of attitudes facilitates spontaneous or
unplanned behaviour (Fazio 1990), and reinforces the necessity of testing both
the affective and instrumental components of attitudes within the theory of
planned behaviour.
Although the conceptual and empirical distinction between affective and
instrumental attitudes appears sound, there has been a recent resurgence of
interest in the conceptualization of the attitude construct within the theory
of planned behaviour, particularly in its constituent components. Hagger
and Chatzisarantis (in press) argue that a major problem is that affective and
instrumental attitudes are not always consistent. When the affective and
instrumental components are highly consistent, such as in exercise behaviour,
they should be typically measured as indicators of a global attitude compon-
ent, rather than distinct constructs. In contrast, when affective and instru-
mental components are inconsistent, they should be considered distinct and
permitted to exert independent effects on intentions. The basic argument then
seems to be that while a model that acknowledges independence between
affective and instrumental attitudes cannot be rejected on statistical or theor-
etical grounds, a second-order model can be favoured in the exercise domain
in that it gives the most parsimonious account of the attitude influences on
intention.
Normative extensions of the theory of planned behaviour
Normatively controlled individuals

Research applications of the theory of planned behaviour have frequently
cited a lesser role for the subjective norms construct in the prediction of
intentions when compared with the variance explained by the attitude and
perceived behavioural control constructs (e.g. Conner and Armitage 1998;
Armitage and Conner 2001). This may be because only a minority of
individuals form intentions on the basis of norms, so-called normatively con-
trolled individuals (Trafimow and Finlay 1996). Studies using within- and
between-participants designs have found that subjective norms are particu-
larly important in the health domain and that normatively controlled people
are likely to perform more health behaviours than individuals who form
intentions on the basis of attitudes (Finlay et al. 2002). However, Finlay et al.
also found that normatively controlled individuals were less likely to enact
their intentions across a number of behaviours within the exercise domain. It
seems that exercise behaviour may be enacted more if a person has a tendency
to be attitudinally biased in their decision-making.
Descriptive norms
One reason why studies have not always observed moderating effects of sub-
jective norms on the attitude–intention relationship may be the fact that
subjective norms insufficiently capture social influences. Several studies have
made the distinction between subjective or injunctive norms as the perceived
pressure from significant others to engage in the target behaviour and descrip-
tive norms, a construct that taps the extent to which significant others engage
in the target behaviour (Rivis and Sheeran 2003). Research has found
injunctive and descriptive norms to be conceptually and empirically distinct
constructs (Hagger and Chatzisarantis, in press), and that descriptive norms
have a unique effect on intentions independent of the traditional subjective
norms (e.g. Rivis and Sheeran 2003). Indeed, Rivis and Sheeran’s recent meta-
analysis across studies adopting the descriptive norms construct within the
framework of the theory of planned behaviour found a medium-to-strong
sample-weighted average correlation between descriptive norms and inten-
tions (rc = 0.44). The analysis also revealed that descriptive norms increased the
variance explained in intention by 5 per cent after controlling for attitudes,
injunctive norms, and perceived behavioural control. The significant effect of
descriptive norms on intentions has also been supported in the exercise
domain (Baker et al. 2003). However, Conner, Sherlock, and Orbell (1998)
have shown that descriptive norms and subjective norms may possess dis-
criminant validity for some behaviours and there is little evidence for the
moderation of the intention–behaviour relationship by descriptive norms.
Indeed, studies often report strong correlations between descriptive norms and
subjective norms that may question their discriminant validity (Hagger
and Chatzisarantis, in press). Therefore, studies have not consistently found
main and moderating effects of descriptive norms on intentions.
Social support
The logic behind incorporating descriptive norms in the theory of planned
behaviour is based on learning theories that posit that modelling or observing
others’ behaviour is an important source of influence (Bandura 1977, 1997).
However, some researchers have argued that descriptive norms may not
sufficiently capture all the types of influence that social factors exert on
intentions. Courneya et al. (2000) have argued that social support, a construct
that indicates the extent to which significant others are perceived to assist
performance of behaviour, can also exert unique effects on intentions (see also
Rhodes et al. 2002b). The logic behind incorporating measures of social sup-
port in the theory of planned behaviour is that for behaviours that are difficult
to execute, assistance provides additional impetus towards forming intentions
beyond beliefs that the behaviour is beneficial, that the actor has the capacity
to do it, and that significant others want them to do it (Courneya et al. 2000).
Several studies have shown that social support and subjective norms tap
conceptually distinct constructs (Rhodes et al. 2002b), and that social support
predicts intentions (Courneya et al. 2000) particularly among individuals who
perceive that behaviour is difficult to execute (Povey et al. 2000). Further, in

accordance with the contingent consistency hypothesis, Povey et al. have
demonstrated that social support moderates effects of attitudes on intentions
so that attitudes predict intentions only when participants perceive that sig-
nificant others support their performance of the behaviour. Nevertheless, it is
important to realize that studies have reported strong correlations between
social support and descriptive norms, and small effects of social support on
intentions relative to the effects that attitudes exert on intentions (Courneya
et al. 2000).
Control-related extensions of the theory of planned behaviour
Self-efficacy
Recent studies have indicated that it is possible to distinguish two sub-
components of Ajzen’s (1991) perceived behavioural control construct: per-
ceived controllability, defined as the extent that an individual has access to
the means to exert control over the target behaviour (Ajzen 2002a), and self-
efficacy, defined as an individual’s estimate of ability and personal capacity to
engage in the behaviour (Terry and O’Leary 1995). Measures of perceived con-
trollability have often focused on statements regarding the degree of subjective
control an individual has over the target behaviour, while self-efficacy is often
tapped using items referring to the perceived abilities and capacities of the actor
in participating in the target behaviour. Studies have made the explicit distinc-
tion between perceived controllability and self-efficacy and provided con-
ceptual and empirical evidence in support of the distinction in a number of
behavioural domains (Armitage and Conner 1999b, 2001), including exercise
(Terry and O’Leary 1995). In addition, strong support for the distinction
between attitudes and perceived behavioural control has been made experi-
mentally on the basis of the belief systems that underlie these constructs
(Trafimow et al. 2002). Notwithstanding the increasing body of evidence in
support of this distinction, there is little conclusive evidence to support a clear,
unambiguous pattern of predictions for perceived behavioural control and
self-efficacy within the theory of planned behaviour (Ajzen 2002a).
Anticipated regret and moral norms2

Although, in principle, all possible outcomes can be reflected by the different
types of beliefs in the theory of planned behaviour, some theorists suggest that
personal beliefs about whether a behaviour is morally right or wrong, termed
moral norms, and anticipated regret for behaving in a certain way may not be
adequately represented in the theory (Conner and Armitage 1998). The con-
cept of anticipated regret is featured in regret theory (Loomes and Sugden
1982) which proposes that people’s decisions in certain situations can be
based on feelings and emotions that they expect to experience from rejecting
alternative behavioural courses, rather than a rational analysis of the costs

and benefits associated with behavioural engagement. Regret theory proposed
that an attractive behavioural option like eating a tasty ice-cream is likely to
be rejected if people believe that they will experience regret and guilt after
choosing that option because they are on a diet. Therefore, people familiar
with post-behavioural feelings of regret and guilt are likely to reject alterna-
tives that will make them feel regretful because they are motivated to avoid
feelings of regret.
The construct of anticipated regret is closely related to moral norms and for
this reason several studies have considered anticipated regret and moral norms
to constitute indicators of a general construct that is termed personal norms.
Moral norms and anticipated regret are interrelated because often feelings of
regret and guilt arise from breaking a moral rule. For example, individuals are
likely to feel guilty if they harm an important other or when they contravene a
social rule that is highly valued. Therefore, anticipated regret and moral norms
are closely linked because anticipated emotions are contingent on the break-
ing of an internalized moral rule. Given this conceptual distinction, however,
there is no conclusive evidence in support of the conceptual distinction
between moral norms and anticipated regret. Therefore, an important avenue
for future research is to examine the conceptual distinction between moral
norms and anticipated affect.
Several studies have supported the inclusion of moral norms and anticipated
regret in the theory of planned behaviour (e.g. Parker et al. 1995; Bozionelos
and Bennett 1999). In the exercise domain, Bozionelos and Bennett revealed
that personal norms did not contribute to the prediction of exercise intentions
or behaviour. This may be because exercise behaviour is unlikely to have moral
or ethical connotations. However, due to the competitive nature of some types
of exercise such as sports, it is not difficult to envisage situations in which
moral norms may influence intentions to behave in a dishonest way. There-
fore, an important avenue for research is to investigate the types of exercise
intentions influenced by moral norms. Research has also demonstrated the
importance of anticipated regret in understanding the intention–behaviour
relationship. Abraham and Sheeran (2003) have shown that anticipated regret
moderates the intention–behaviour relationship so that participants are most
likely to exercise if they both intend to exercise and report high levels of
anticipated regret if they fail to exercise. In addition, Abraham and Sheeran
(2004) also demonstrated that manipulations of anticipated regret that
prompt people to focus on negative emotions that follow decisions not to
exercise can strengthen exercise intentions.
Summary of theoretical extensions and recommendations for practice

One clear conclusion that emerges from the research that has extended and
modified the theory of planned behaviour is that a host of other variables
can be incorporated into the framework of the theory and account for unique
variance in intentions and behaviour. This suggests that the theory does not
sufficiently capture the psychological determinants of volitional behaviour,
but is both versatile and flexible. However, it is important to acknowledge that
the contribution of these additional variables to the prediction of exercise
intentions and behaviour has been relatively modest (Conner and Armitage
1998; Hagger et al. 2002b; Hagger and Chatzisarantis, in press). Further-
more, these modifications must be evaluated alongside the original theory’s
parsimony in accounting for a large amount of variance in intentions and
behaviour.
Another conclusion that emerges from this research is the practical value of
the extensions and modifications of the theory. Formative research that evalu-
ates extended versions of the theory can provide useful information about the
content of interventions. For example, the contribution of personality traits to
the prediction of exercise behaviour suggests that exercise programmes should
take into consideration the personality characteristics of participants. Exercise
programmes that are tailored around personality characteristics are likely to be
more successful than those that focus solely on the proximal predictors of
intention (Courneya et al. 2002). Another example is the distinction made
between affective and instrumental attitudes as this can help identify whether
affective or cognitive means of persuasion are more appropriate to a particular
behavioural domain (Hagger and Chatzisarantis, in press). However, it must be
recognized that not all additional variables can be manipulated, such as past
behaviour. Interventions in the exercise domain have not attempted to
manipulate the additional variables identified in this research. Therefore,
interventions that evaluate the utility of the extended theory in promoting
exercise behaviour are urgently needed.
One issue that deserves future research attention is the conceptual overlap
between additional variables. Obviously, some psychological variables that
have been incorporated into the theory of planned behaviour are defined and
measured in similar, if not identical, ways. For example, the constructs of
anticipated affect and anticipated regret are very similar, given that both
constructs describe post-behavioural affective reactions (Parker et al. 1995).
Conceptual overlap between the theory constructs was also evident in a
recent study conducted by Hagger and Chatzisarantis (in press) that found
higher-order constructs could satisfactorily model the common content
of these additional variables and could successfully represent influences from
these additional variables on intentions. Such commonalities between the
psychological constructs from the theory give rise to an important practical
question: to what extent are the manipulations implied by different psycho-
logical constructs redundant? Research that investigates commonalities and
differences between additional variables is urgently needed. Trafimow et al.’s
(2002) study of perceived behavioural control is an excellent example of a
series of experimental studies that have pointed out conceptual differences
between perceived controllability and perceived difficulty.
Finally, it is important to investigate conceptual similarities and differences
between the psychological constructs in models like the theory of planned

behaviour in order to arrive at realistic and practical frameworks to under-
stand exercise behaviour. Clearly, people cannot process a great deal of
information simultaneously. As a consequence, the practice of adding new
variables within the theory of planned behaviour may provide an unrealistic
picture of the decision-making process because such extended models imply
that people can deal with lots of different information at the same time. In
addition, extended models may not be very helpful in promoting exercise
behaviour because interventions based on such models provide greater
amounts of information that may over-complicate the issue of behaviour
change for the recipients of the message. One way to address the problem of
limited processing capacity is to point out redundancies between psycho-
logical constructs. Redundant variables should be excluded from intervention
programmes because they provide little or no additional information about
behavioural change.
Other leading theories of social behaviour
The model of goal-directed behaviour

One limitation of the social cognitive theories such as the theories of reasoned
action and planned behaviour is that they focus on behavioural appraisals in
lieu of goal appraisals. The model of goal-directed behaviour extends previous
theories of intentional behaviour by explicitly considering the broader goal-
context in which behaviour is enacted (Perugini and Conner 2000; Perugini
and Bagozzi 2001). The model therefore views the fundamental components
of the theory of planned behaviour with respect to goals rather than
behaviours. For example, attitudes are defined as the extent to which people
think that performing a behaviour x in order to achieve a goal y is beneficial
versus harmful. In addition, the model of goal-directed behaviour also
includes the influence of anticipated emotional reactions toward goal
achievement. Again, the model defines positive and negative anticipated emo-
tions with respect to goal achievement and failure. That is, instead of measur-
ing the extent to which performing behaviour x will make people feel excited
or happy, the model refers to how the achievement of goal z will make people
feel. For example, individuals are asked how they would feel if they succeeded
in achieving goal z.
The model of goal-directed behaviour maintains that desires are the
most proximal determinant of intentions. Desires aim to reflect the motiv-
ational aspects of attitude that are not captured by traditional intentions. In
the model, measures of desires are focused on goals rather than behaviours.
Furthermore, desires are hypothesized to mediate the effects of attitudes,
anticipated affect, subjective norms, and perceived behavioural control
on intentions. Another key feature of the model is concerned with the
conceptualization of intentions. Perugini and Conner (2000) have argued that

the conceptualization of intentions in the theory of planned behaviour does
take the goals of the intended behaviour into consideration. The concept of
intentions in the model of goal-directed behaviour resolves this issue by
including three aspects of volition: (1) a direct expression of intentions (e.g. ‘I
will try to perform a behaviour x in order to achieve a goal z’); (2) a statement of
plans to perform acts that are instrumental to the achievement of a goal (e.g. ‘I
intend to perform a behaviour x in order to achieve a goal z’); and (3) an
expression of volition and effort needed to enact the behaviour (e.g. ‘I will
expend effort to perform a behaviour x in order to achieve a goal z’). Finally, the
model of goal-directed behaviour specifies effects from past behaviour on
behaviour, volitions, and desires. Effects of past behaviour are considered to
represent effects from automatic mental processes on motivational and
volitional variables. The model is depicted in Figure 2.4.
Figure 2.4 The model of goal-directed behaviour
Studies of a number of health behaviours including exercise have suggested

that the model of goal-directed behaviour has a greater predictive value than
the theory of planned behaviour (Perugini and Conner 2000; Perugini and
Bagozzi 2001). Empirical studies have also shown that attitudes, subjective
norms, perceived behavioural control, and positive and negative anticipated
emotions predict desires and that desires predict intention. Perugini and
Bagozzi have also shown that desires mediate the effects of attitude, subjective
norms, perceived behavioural control, and anticipated emotions on inten-
tions. However, there are as yet no intervention studies that have tested utility
of the model in promoting exercise behaviour.
The transtheoretical model

In the context of alcohol addiction and smoking behaviour, Prochaska and
DiClemente (1982) developed a model that incorporated behaviour change
along with social cognitive constructs to identify the progression that people
undergo when changing unhealthy behaviours. At the core of the model is the
premise that people pass through stages towards making a behaviour change
and each stage is characterized by a particular pattern of psycho-social and
behavioural variables. These stages of change are the most widely adopted
aspects of the model, although there are also levels of change and processes
of change. The stages are: pre-contemplation, contemplation, preparation,
action, and maintenance. At the pre-contemplation stage people have no con-
ceptualization of the need to change unhealthy behaviours or lifestyle, let
alone enacting any change. Contemplators, however, have realized the need for
change and are thinking about doing so but have no idea or inclination to
make the change. People who are in the preparation stage want to change and
are making the necessary personal and social moves and commitment towards
changing behaviour. In the action stage, people are beginning to make success-
ful and unsuccessful attempts to change their behaviour and their behavioural
change may be inconsistent and could possibly relapse to a previous
behavioural pattern more akin to their former unhealthy lifestyle. Finally, the
maintenance stage is characterized by a continuity of successful and consistent
behavioural change.
In terms of the processes involved with moving people from one stage to
another in the transtheoretical model, Prochaska and DiClemente (1982) sug-
gested that self-efficacy and decisional balance are key intrapersonal variables
that are responsible for such change. It is the incorporation of such social
cognitive constructs and behavioural constructs as well as the dynamic nature
of the model that inspired the name of the theory because it is proposed to
straddle theories of intention and theories of enactment of intentions. Self-
efficacy in the context of the transtheoretical model refers to an individual’s
beliefs in their ability to make desired behavioural changes to gain desirable
outcomes. Research adopting the transtheoretical model has supported differ-
ences in the level of self-efficacy across the stages of change, with the highest
coinciding with people in the maintenance stage and the lowest for those in
the pre-contemplation stage (Armitage and Arden 2002).
Decisional balance reflects people’s overt beliefs about the advantages and
disadvantages, or the pros and cons, of performing the behaviour, with the
hypothesis that the balance of pros and cons determines when a person begins
to make behavioural changes. Therefore, it is hypothesized that when the pros
just begin to outweigh the cons the person will be in the preparation stage.
Research has supported this hypothesis in a number of behavioural contexts
including exercise, such that the decisional balance is largely in favour of pros
in actors and maintainers, roughly equal in people at the preparation stage
(Prochaska et al. 1994). In summary, the research testing the efficacy of self-
efficacy and decisional balance to change stage in the transtheoretical model

has supported theory hypotheses across a number of studies and a recent
meta-analysis has supported this view in the exercise domain (Marshall and
Biddle 2001).
The transtheoretical model has also been investigated within the framework
of other social cognitive models in an attempt to further examine the pro-
cesses that expedite transfer from one stage to another. Indeed, researchers
have drawn some parallels between the core antecedents of intention in the
theory of planned behaviour and the decisional balance and self-efficacy con-
structs in Prochaska and DiClemente’s exposition of the transtheoretical
model (Armitage and Arden 2002; Armitage et al. 2003). For example, the
belief-based measures of attitude are proposed to reflect the advantages and
disadvantages (behavioural beliefs) relating to behavioural engagement,
and could therefore be considered akin to decisional balance. In addition,
the perceived behavioural control variable has been overtly compared with the
self-efficacy construct so that perceived behavioural control is considered, at
least in part, to reflect levels of self-efficacy. A number of studies across differ-
ent health behaviours (e.g. Armitage and Arden 2002; Armitage et al. 2003),
including exercise (Courneya 1995) have found linear increases in these
theory of planned behaviour constructs across the stages of change.
While the transtheoretical model has demonstrated concurrent and predict-
ive validity alongside other intention models of social cognition, it is not
without its critics. Some have questioned its validity because it represents a less
than coherent mix of cognitive and behavioural constructs with no clear
notion of the exact nature of each stage (Davidson 1992). Furthermore, some
have claimed that it is not a genuine stage model, but only a ‘pseudo-stage’
model because the boundaries of the stages are not distinct and cannot be
discriminated empirically (Sutton 2000; Armitage and Arden 2002; Armitage
et al. 2003). This is supported by the linear trends for all of the social cognitive
constructs from other theories across the transtheoretical model stages. In
some respects, the stages of change adhere very closely to measures of inten-
tion and in this respect the stages themselves are merely a reflection of the
different levels of intention toward engaging in the behaviour. Thus, in theory,
any number of stages could be identified across the continuum of low to high
intention (Armitage and Arden 2002).
Sutton (2000) suggests that for the transtheoretical model to exhibit the
characteristics of a true discrete stage model, discontinuity patterns in key
social psychological variables would have to be observed across the stages. In
other words, the differences in key variables thought to be responsible for
moving a person from on stage to another would have to exhibit a non-linear
change. An example of a discontinuity pattern might be that attitudes may
exhibit an increase across the contemplation and preparation stages but no
increase across the subsequent stages of preparation and action or action and
maintenance. In this example, interventions targeting attitudes would be
efficacious in progressing individuals from the contemplation stage to the
Figure 2.5 Discontinuity pattern for intention in the context of the

transtheoretical model
preparation stage, but would not progress people already in the preparation
stage. This is illustrated in Figure 2.5. There have, however, been few empirical
tests to corroborate these discontinuity patterns.
Recently, another feature of attitudes, that of attitudinal ambivalence, has
been viewed as an important process variable in changing transtheoretical
model stages and has been shown to exhibit the discontinuity patterns sug-
gested by Sutton (2000). Ambivalence suggests that individuals may hold both
positive and negative evaluations of a behaviour simultaneously and this may
result in conflict when decisions are made. Ambivalence has been shown to be
a moderator of the attitude–intention relationship, so that high levels of
ambivalence compromise the motivational efficacy of attitudes (Armitage and
Conner 2000). In the context of the transtheoretical model, Armitage et al.
(2003) found that while the theory of planned behaviour variables exhibited
linear relationships across stages of change, the relationship was non-linear
(quadratic) with attitudinal ambivalence. The data showed that the highest
levels of ambivalence were seen in the contemplation, preparation, and action
stages, and the lowest levels in the pre-contemplation and maintenance
stages. The authors suggested that changing ambivalence may encourage

pre-contemplators into contemplation, preparation, or action stage, but not
across these stages while resolving ambivalence may prevent those in the
maintenance stage from relapsing. These findings need further replication in
an exercise context and would be useful in characterizing the transtheoretical
model as a bona fide stage model.
Suggested reading
Conner, M. and Armitage, C.J. (1998) Extending the theory of planned behavior: a
review and avenues for further research. Journal of Applied Social Psychology, 28: 1429–
64. Covers many of the theoretical issues associated with the theory of planned
behaviour and modifications of this theory.
Conner, M. and Norman, P. (1996) Predicting Health Behaviour: Research Practice with
Social Cognition Models. Buckingham: Open University Press. Highly recommended
review of the theoretical and practical issues related to the health belief model, pro-
tection motivation theory, and the theory of planned behaviour.
Hagger, M.S., Chatzisarantis, N. and Biddle, S.J.H. (2002) A meta-analytic review of the
theories of reasoned action and planned behaviour in physical activity: predictive
validity and the contribution of additional variables. Journal of Sport and Exercise
Psychology, 24: 3–32. The most recent and comprehensive review of research on this
key social cognitive model in exercise contexts.
Marshall, S.J. and Biddle, S.J.H. (2001) The transtheoretical model of behavior change: a
meta-analysis of applications to physical activity and exercise. Annals of Behavioral
Medicine, 23: 229–46. State-of-the-art cumulative review of research using the trans-
theoretical model in the exercise domain.
Perugini, M. and Conner, M. (2000) Predicting and understanding behavioral volitions:
the interplay between goals and behaviors. European Journal of Social Psychology, 30:
705–31. Showcase research publication for the model of goal directed behaviour.
Summary
• Social cognitive theories aim to identify the social psychological factors

that explain variance in exercise behaviour. Intention is a central motiv-
ational construct to many of these theories.
• The health belief model and the protection motivation theory examine the
effects of cognitive appraisals of perceived threat (e.g. severity, vulner-
ability) and perceived health benefits of behaviours such as exercise on
intentions and behaviour. Meta-analyses have supported the relationships
among these variables in health behaviours. Interventions based on these
theories using fear-arousing communications have had limited success in
changing intentions but not behaviour.
• The theories of reasoned action and planned behaviour are parsimonious and
effective social cognitive models of intention and behaviour and have
received considerable attention in the exercise literature. The belief-based

constructs of attitudes and subjective norms in the theory of reasoned
action with the perceived behavioural control construct in the theory of
planned behaviour are strong and consistent predictors of exercise inten-
tions, and intentions have been shown to mediate these distal constructs
on exercise behaviour.
• The theory of planned behaviour is a versatile and flexible theory as it
has often been modified to incorporate numerous additional constructs
that explain unique variance in exercise intention and behaviour. Modifi-
cations have been made to include past behaviour, personality, attitude-
based constructs (e.g. attitude strength, affective and cognitive attitudes),
normatively-based constructs (e.g. normatively controlled individuals,
descriptive norms, social support), control-related modifications (e.g.
self-efficacy), anticipated regret, and moral norms.
• The model of goal-directed behaviour is an alternative theory of intentional
behaviour. The theory aims to explain the concept of desires as the most
proximal predictor of behaviour and focuses on behavioural goals rather
than behavioural enactment per se. The model has been shown to have
some utility among health behaviours.
• The transtheoretical model is a stage model of behaviour change. The
theory proposes that people move systematically through five stages when
changing behaviour: pre-contemplation, contemplation, preparation,
action, and maintenance. The theory has intuitive appeal and has been
linked to decisional balance and intention. Recent support has come from
discontinuity patterns with respect to social cognitive constructs like
attitude.
Notes
1 The concept of accessibility was initially introduced by Higgins et al. (1982) and
Haddock and Zanna (1998) whose research showed that models that take accessible
beliefs into consideration provide a more realistic representation of the processes
underlying attitude formation compared with models that do not account for
accessibility.
2 Some authors have used the terms personal normative beliefs and moral obligation
with reference to the constructs of moral norms and personal norms (Sparkes and
Shepherd 1995; Triandis 1980; Godin et al. 1996). However, the items used to measure
all these constructs are virtually identical.
3
From exercise intention to exercise

behaviour and beyond
The social cognitive theories discussed in Chapter 2 are motivational theories

in that they indicate how intentions are formed and change (Gollwitzer 1999).
However, these theories do not provide an explanation of how people carry
out their previously formed intentions. This is highlighted by the boundary
condition of intentional stability in the theory of planned behaviour that
states that intentions predict behaviour only to the extent that intentions
remain stable over time. For example, several studies have documented that
stable intentions are more predictive of future behaviour than unstable inten-
tions (Sheeran et al. 1999; Conner et al. 2000). As a consequence, researchers
have developed alternative theories of volition to understand how people
carry out their previously formed intentions (e.g. Gollwitzer 1999). In add-
ition, the social cognitive theories of intention propose that the key predictor
variables such as attitudes and subjective norms are formed through specific
sets of beliefs and evaluations. However, little has been proposed as to the
origins of these beliefs. Theories of psychological needs such as self-
determination theory may provide a theoretical basis for belief formation. This
chapter will review these theories and examine how volitional theories com-
plement the motivational social cognitive theories covered in Chapter 2 to
explain exercise behaviour. In addition, self-determination theory is intro-
duced and its efficacy in explaining the origins of the social cognitive con-
structs from theories of intention reviewed. Multi-theory approaches that
combine hypotheses from these theories to explain exercise behaviour will
also be proposed.1
44 From exercise intention to behaviour
Limitations of social cognitive theories and introduction to

volitional processes
Intention stability
Ajzen (1985) originally downplayed the theoretical importance of the bound-
ary condition of intention stability by arguing that the issue of stability is
largely a technical problem and does not challenge the assumption that inten-
tions influence behaviour. Reduced predictive accuracy of intentions may
merely reflect reduced accuracy of measures of intentions available to the
investigator (Ajzen 1985). This can occur when there is a large delay between
assessments of intentions and behaviour (Sheeran et al. 1999b). The greater
the delay, the greater the opportunities for intentional change, and hence the
predictive accuracy of intentions is reduced (Ajzen 1985). For this reason,
Ajzen (1991) has proposed that, ideally, intentions should be measured as
close as possible to the incidence of behaviour. Research has corroborated
Ajzen’s proposition by showing that short-range intentions and stable inten-
tions are stronger predictors of exercise behaviour than long-range intentions
and unstable intentions (Sheeran et al. 1999b; Conner et al. 2000).
However, it can be argued that the issues of stability and predictive validity
are not mere technical problems but do pose important theoretical and
practical questions about the utility of the theory of planned behaviour in
predicting and explaining volitional behaviour. First, the boundary condition
of stability can be considered unrealistic because the construct of intention is
state-like and likely to change over time. Therefore, intentional change is
unavoidable because change is integral to the nature of intentions (Hagger et
al. 2001; Chatzisarantis et al., in press a). In addition, stable intentions predict
behaviour more accurately than unstable intentions because theories of inten-
tion have not been designed to explain the psychological processes underlying
intention change. Indeed, research has shown that the theory of planned
behaviour does not predict adherence to, relapse from, and delayed participa-
tion in health behaviours (Orbell and Sheeran 1998). Having described the
factors that may facilitate changes in intentions, theories of intentional
behaviour should predict behaviour even when intentions are unstable (Ajzen
and Fishbein 1980; Ajzen 1991). Further, from an applied perspective, it can be
argued that the relatively modest levels of behavioural prediction obtained by
social cognitive theories of intentional behaviour substantially undermine the
practicality of these theories. Therefore, the boundary condition of stability is
a theoretical and practical limitation of the theory of planned behaviour and
other theories of intention because they do not provide a realistic view of the
way intentions operate.
From exercise intention to behaviour 45
Scale correspondence
One reason why intentions may not predict behaviour very well is that
measures of intentions and behaviour do not correspond in terms of action,
target, context, and time (Ajzen 1985). For example, participation in exercise
behaviour during leisure time is predicted more accurately and precisely from
measures of intentions to exercise during leisure time than from measures of
intentions to engage in another context, such as physical education classes.
Courneya and co-workers (Courneya and McAuley 1994) argue that the rela-
tionship between intentions and exercise behaviour in previous research is
modest and highly variable because studies use different types of scales in
measuring intentions and behaviour. However, although improving scale and
measurement correspondence strengthens the intention–behaviour relation-
ship, research has shown the principle of correspondence does not fully
explain variations in the intention–behaviour relationship (Courneya and
McAuley 1994). Studies in the exercise domain have shown that a great deal of
variance in exercise behaviour remains unexplained even when measures
adhere to the correspondence rule (Hagger et al. 2002b).
Intention formation
Bagozzi and Yi (1989) argued that when an individual has given much con-
sideration to a behaviour and has well-formed intentions, the intention–
behaviour relationship will be strengthened. They suggested that poorly
formed intentions are more vulnerable to change due to differential retrieval
or forgetting of information on which the intentions are based. This suggests
that poorly formed intentions are more unstable and will be less effective in
accounting for variance in behaviour. Bagozzi and Yi demonstrated empiric-
ally that the intention–behaviour relationship is stronger for well-formed
intentions than for ill-formed intentions. Future studies should examine the
issue of intention formation in an exercise context.
Self-schema
Self-schema is another variable that has been shown to affect the intention–
behaviour relationship. Self-schemas are cognitive generalizations about the
self that are derived from past experience. Self-schemas are typically measured
by participants’ self-ratings of the descriptiveness and importance of traits
relevant to a domain (Sheeran and Orbell 2000). For example, a person who
rates domain relevant traits as highly self-descriptive and important to their
self-image are considered to be schematics. Non-schematics are people who rate
such traits as not descriptive of their self but important. Aschematics are people
whose ratings of domain-relevant traits are moderately self-descriptive but
unimportant. The rationale behind the argument that self-schema moderates
the intention–behaviour relationship is that schematics have a greater number
of strategies for acting intentions and will therefore exhibit greater consistency
between intentions and behaviour than non-schematics. In accordance with
this hypothesis, Sheeran and Orbell have shown that schematics display
stronger relationships between exercise intentions and behaviour than
non-schematics. However, once again self-schema does not fully explain why
intentions do not always translate into actions.
Volitional and forced intentions

Building upon the work of Meiland (1970), Chatzisarantis and Biddle (1998)
have argued that one reason why the long-term predictive efficacy of inten-
tions is relatively low is because intentions do not necessarily reflect a
willingness to perform a behaviour as originally assumed by Ajzen (1985).
This assumption may be untenable considering that previous research has
differentiated between volition and intentions, and demonstrated deferential
effects from volitional and forced decisions on behaviour. Specifically, more
than four decades of experimental work on cognitive dissonance theory have
shown that decisions can be either volitional or forced and that volitional
decisions exert greater effects on attitudes and overt behaviour than forced
decisions (Festinger and Carlsmith 1958). Further studies have shown that
volitional decisions are associated with effort and persistence in health-related
behaviours whereas this is not always the case for forced decisions (Deci et al.
1999a). This evidence suggests that making the distinction between volitional-
forced intentions in the theory of planned behaviour may assist in further
explaining how intentions are translated into actions.
Chatzisarantis, Biddle, and Meek (1997) developed a simple measure of vol-
itional and forced intentions in order to examine the contribution of these two
types of intentions to the prediction of exercise behaviour. Participants are
considered to hold volitional intentions if they described their intentions as
something that they want to execute by themselves. In contrast, participants
are considered to hold forced intentions when they describe their intentions
as something that they are obliged to do. Chatzisarantis et al. found that
volitional and forced intentions added to the prediction of exercise behaviour
after controlling for traditional measures of intentions, but the effects of
volitional and forced intentions on exercise behaviour were small. On balance,
the distinction between volition and intentions does not fully explain the
intention–behaviour relationship and has not been shown to have a major
impact on the explanation of exercise behaviour.
Implementation intentions
One reason why theories of intentional behaviour do not fully explain the
processes by which intentions translate into actions is that people often forget
to carry out their intentions (Orbell et al. 1997; Gollwitzer 1999; Sheeran and
Orbell 1999). Alternatively, people’s execution of their intentions may be
interrupted because other competing goal-directed behaviours take priority

over the original intended behaviour (Verplanken and Faes 1999). Motiv-
ational theories do not address these difficulties associated with enactment of
intentions, and for this reason they may not fully explain the intention–
behaviour relationship.
One approach that has been put forward to resolve the inadequacies of the
intention–behaviour relationship in the theory of planned behaviour is
implementation intentions. Implementation intentions are self-regulatory
strategies that involve the formulation of specific plans that specify when,
how, and where performance of behaviour will take place. Experimental para-
digms have been used to initiate implementation intentions and require
research participants to explicitly specify when, where, and how they will
engage in an intended behaviour to achieve their behavioural goals. According
to Gollwitzer (1999), implementation intentions are powerful self-regulatory
strategies that help people move from a motivational phase to a volitional
phase and they ensure that their intentions are converted into action. In
accordance with this hypothesis, research has indicated that forming imple-
mentation intentions decreases the probability of people forgetting to initiate
their goal-directed action at the point of initiation (Orbell et al. 1997; Sheeran
and Orbell 1999). This is because planning when and where to initiate a pro-
spective action strengthens the mental association between representations
of situations and representations of actions. Research has also shown that
increased accessibility of situational representations in memory increases the
probability of action opportunities getting noticed, and of action initiation
occurring, given that the mere perception of action opportunities can auto-
matically trigger a behavioural response (Bargh 1994; Orbell et al. 1997;
Gollwitzer 1999; Sheeran and Orbell 1999; Koestner et al. 2002). Importantly,
implementation intentions increase behavioural engagement through these
post-decisional, automatic mechanisms, and not by concomitant increases in
motivation or intention (Orbell et al. 1997).
Recent research has evaluated the effectiveness of interventions that combine
motivational techniques with volitional techniques such as implementation
intentions in influencing the performance of social behaviour (Koestner et al.
2002; Milne et al. 2002; Prestwich et al. 2003; Sheeran and Silverman 2003).
The rationale behind this combined approach is that motivational strategies
focus on increasing intention levels but do not facilitate the enactment of
intentions, while volitional strategies such as implementation intentions
increase the probability that these strong intentions will be converted into
action but not on changing intentions. Research has corroborated the utility
of these combined techniques in increasing exercise behaviour. For example,
Milne et al. (2002) demonstrated that implementation exercises strengthen
the ability of protection motivation theory constructs. Similarly, Sheeran and
Silverman (2003) found that an intervention that combined an application of
the theory of planned behaviour and implementation intentions is more
effective in promoting attendance at health and training courses than an
intervention that is based on the theory of planned behaviour alone. In add-

ition, Prestwich et al. (2003) have demonstrated that an intervention with a
combination of a rational decision-making strategy, or decisional balance sheet,
and implementation intentions was more effective in promoting exercise
behaviour than either of the strategies alone. Koestner et al. (2002) also found
an interaction between a motivational strategy to enhance self-concordant
goals and implementation intentions in goal attainment and persistence.
These results support the existence of two distinct phases of motivation: a
motivational or pre-decisional phase during which people decide whether or
not to perform a behaviour and a volitional, post-decisional, or implemental
phase during which people plan when and where they will convert their
intentions into behaviour. In addition, research evidence suggests that the
combination of motivational and volitional techniques is most effective in
promoting exercise behaviour.
Continuation intentions
One final reason why intentions change over time and consequently do not
always translate into actions is that people tend to devalue the behaviours and
outcomes as they get closer to achieving them (Chatzisarantis et al., in press c).
Consider for example the relationship between exercise behaviour and body
weight. Initially, people may be highly motivated to participate in exercise
because they expect doing so will help them to lose weight. However, people
may start losing interest in the goal of losing weight once they do start to lose
weight. According to Lewin, this is because outcome evaluations such as the
value attached to losing weight are inversely proportional to the psychological
distance between the self and the outcome like actual weight loss. This means
that the greater the distance between the person’s current state and the
actual outcome, the higher their belief that they need to lose weight. Given
that intentions are associated with outcome evaluations (Ajzen and Fishbein
1980), it is hypothesized that intentions will decline over time and may be less
effective in predicting behaviour in the long run.
Chatzisarantis et al. (in press c) have proposed that the long-term intention–
behaviour relationship can be improved if the theory of planned behaviour
can be modified to account for the decline in interest in the behavioural out-
come as the attainment of that outcome approaches. These authors proposed
that to account for the changes that occur in intentions as the achievement of
a long-term outcome nears, theories need to account for people’s intentions to
continue the performance of the target behaviour after they have achieved or
failed to achieve the outcome. Chatzisarantis et al. (in press c) defined two
constructs: continuation intentions of success and continuation intentions of
failure. Continuation intentions of success were defined as an individual’s
intent to continue engaging in a behaviour when presented with a hypo-
thetical situation in which the individual is successful in attaining their
behavioural outcomes. Continuation intentions of failure refer to intentions
to continue the performance of the behaviour when presented with a hypo-

thetical situation that signifies unsuccessful progress toward the behavioural
outcome. Their study showed that these continuation intentions contributed
to the prediction of exercise behaviour after traditional components of the
theory of planned behaviour were taken into consideration.
Recommendations for future research and practice in volitional

strategies
A trend clearly evident in the literature is that motivational techniques do not
necessarily facilitate successful enactment of behavioural intention (Ajzen
1991) and volitional techniques do not effect changes in intentions (Orbell
2003). The theoretical and practical implication of these findings is that inter-
ventions that combine motivational techniques with volitional strategies are
more effective in promoting exercise behaviour than motivational techniques
alone (Koestner et al. 2002; Sheeran and Silverman 2003). These results
provide strong evidence that human motivation consists of two phases: a
motivational or pre-decisional phase and a volitional, implemental, or post-
decisional phase (Gollwitzer 1999). In the motivational phase people deliber-
ate over advantages and disadvantages of behaviour and this culminates in the
formation of intentions. The theories of reasoned action and planned
behaviour (Ajzen and Fishbein 1980; Ajzen 1985) and protection motivation
theory (Rogers 1975) are examples of motivational theories that explain inten-
tion formation. In the volitional phase, people decide when, where, and how
to implement their intention (Gollwitzer 1999).
One caveat is that it is not yet known whether interventions that facilitate
continuation intentions and implementation intentions improve the utility
of motivational interventions in promoting exercise behaviour. There is
some preliminary evidence to support the supplementation of the motiv-
ational strategies based on the theory of planned behaviour with volitional
strategies such as implementation intentions (Koestner et al. 2002; Prestwich
et al. 2003). Further, the different volitional techniques have not been com-
bined or compared in terms of their effectiveness in promoting exercise
behaviour. It is possible that while implementation intentions promote ini-
tiation of exercise behaviour by enhancing memory of action initiation,
continuation intentions may promote a more sustained participation in
exercise behaviour by strengthening motivation to persist under adversity.
Therefore, one important avenue for future research is to examine the differ-
ential processes by which different volitional techniques such as implemen-
tation intentions and continuation intentions facilitate participation in
exercise behaviour.
Self-determination theory
Self-determination theory is a theory of human motivation that has been

applied extensively to the understanding of exercise behaviour (Deci and Ryan
1985, 2000). It comprises three sub-theories: cognitive evaluation theory,
causality orientation theory, and organismic integration theory.1 In general,
self-determination theory argues that theories of intentional behaviour are too
simplistic and mechanistic to provide a complete picture of human motiv-
ation (Deci and Ryan 1985). For example, the two-phase conceptualization of
motivation advocated by social cognitive theories of intentional behaviour
does not account for the motivational factors that give rise to these social
cognitive predictors in intentional theories. Self-determination theory pro-
poses that the three essential and innate psychological needs for self-
determination, competence, and relatedness form the basis of motivation.
Self-determination refers to the need to experience oneself as initiator and
regulator of one’s actions. Competence refers to the need to be able to produce
behavioural outcomes. Relatedness refers to the need to experience satisfac-
tory relationships with significant others (Deci and Ryan 2000).
These three psychological needs are considered to be essential for optimal
psychological development, well-being, and growth (Deci and Ryan 2000).
Just as humans have biological needs such as thirst and hunger, people also
have the universal need to seek out and obtain experiences of self-
determination, competence, and relatedness (Deci and Ryan 2000; Sheldon et
al. 2001). In addition to addressing issues related to the energization of human
motivation by psychological needs, self-determination theory also places
importance on the social contexts that can affect motivated behaviour (Deci
and Ryan 2000). According to Deci and Ryan, psychological need satisfaction
is a necessary but not sufficient condition for optimal psychological develop-
ment, well-being, and growth. Psychological needs must also be supported by
the social context because otherwise individuals will feel alienated and will
not experience need satisfaction (Deci and Ryan 1985). It is evident therefore
that self-determination theory is a dialectic theory that views the environment
as nurturing need-satisfaction and motivation.
Another important difference between theories of intentional behaviour
and self-determination theory is concerned with form or quality of motivation.
Unlike theories of intentional behaviour that view motivation as varying in
strength alone, self-determination theory proposes the interplay between the
environment and the individual gives rise to different types of motivation
(Deci and Ryan 1985). Self-determination theory distinguishes between two
forms of motivated behaviour. Intrinsic motivation refers to doing an activity
for its inherent satisfaction rather than for some tangible or extrinsic outcome,
while extrinsic motivation refers to doing an activity for external outcomes that
are separable from the activity itself (Deci and Ryan 1985, 2000). In addition,
organismic integration theory, a sub-theory of self-determination theory,
distinguishes between five different forms of extrinsic motivation, external
regulation, introjected regulation, identified regulation, integrated regulation, and

intrinsic motivation.
The specification of these different types of motivation permits the study
of individual differences in the quality of motivation toward volitional
behaviours. Deci and Ryan (2000) argue that the intensity or level of motiv-
ation of two people may be equivalent and can lead to behavioural per-
formance that is quantitatively comparable, but if the form of motivation is
different, the quality of performance may differ. For example, self-determined
or intrinsic motivation may lead to better performance on creative tasks but
controlling or extrinsic motivation may facilitate performance on more
mundane, mechanistic tasks. Furthermore, a person may be either intrinsically
or extrinsically motivated towards certain behaviours, but, from the point of
view of an external observer, their level of behavioural engagement is the
same. However, when it comes to persistence with that behaviour, the
observed behaviour may differ depending on the type or form of the motiv-
ation, suggesting that the quality of motivation may explain variance in
behaviour across different situations. Therefore, acknowledging different
forms of motivation is important because it provides a means of explaining
behavioural persistence. Finally, unlike theories of intentional behaviour that
consider many different short- and long-term consequences of behaviour,
self-determination theory addresses the ultimate or universal reasons for
motivation and behaviour (Sheldon et al. 2001). These reasons or motives are
more distal than the social cognitive predictors in theory of intention such as
attitudes and intentions in the theory of planned behaviour.
However, it is important to realize that self-determination theory does share
some similarities with social cognitive theories of intentional behaviour. For
example, it considers both internal events such as attitudes and intentions and
social environments to be important determinants of behaviour. Therefore,
self-determination theory and theories of intentional behaviour are not
antithetical but may be able to complement each other. The differences lie in
the proximity and origin of the constructs that influence behaviour. While
theories of intention deal with more specific cognitive and affective deter-
minants of action, self-determination theory focuses on the more general,
organismic conditions that give rise to motivation and behaviour. Deci and
Ryan (1985) recognize the potential for such a theoretical union: ‘Cognitive
theories begin their analysis with . . . a cognitive representation of some future
desired state. What is missing, of course, is a consideration of the conditions of
the organism that make these future states desired’ (1985: 228). Self-
determination theory therefore offers reasons why such social cognitive vari-
ables exist. In the next sections, cognitive evaluation theory and organismic
integration theory will be introduced and their application to the explanation
of exercise behaviour discussed.
Cognitive evaluation theory

Cognitive evaluation theory is a sub-theory of self-determination theory that
was developed to explain the effects of intrinsic motivation on behaviours and
how social contexts affect intrinsic motivation (Deci and Ryan 1985). Intrinsic
motivation can be measured or indicated in two ways. First, time spent
engaged in a target task during a free choice period in which several attractive
alternatives other than the target task are available constitutes a behavioural
measure of intrinsic motivation. To the extent that people freely choose and
persist with the target task, the researcher has a strong indication of the degree
on intrinsic motivation towards that task. Therefore, the longer the time spent
in the task, the more the person is assumed to enjoy the task and the higher
their inferred level of intrinsic motivation. Second, self-reports of the degree of
interest a person derives from the target task during the free choice period is an
alternative measure of intrinsic motivation.
Cognitive evaluation theory aims to explain variance in intrinsic motiv-
ation. It proposes that intrinsic motivation is engendered when external and
internal events relevant to the initiation and regulation of behaviour support a
person’s innate psychological needs for self-determination and competence
(Deci and Ryan 1985). In addition, the theory proposes that there are three
general processes by which external events such as tangible rewards and
internal events such as goals can influence intrinsic motivation. The first pro-
cess is through perceived locus of causality. The construct of perceived locus of
causality indicates whether people perceive their behaviour as emanating from
their self, and is therefore volitional and self-determined, or whether their
behaviour is regulated by some external force such as a significant other. With
regards to perceived locus of causality, Deci and Ryan have predicted that
events relevant to initiation and regulation of behaviour that promote a more
internal perceived locus of causality will promote intrinsic motivation,
whereas events that promote a more external perceived locus of causality are
likely to undermine intrinsic motivation.
The second process by which external events influence intrinsic motivation
is via perceived competence. Events that affirm personal competence are likely
to maintain intrinsic motivation, while events that diminish perceived com-
petence undermine intrinsic motivation. The third process of influencing
intrinsic motivation relates to how people construe events relevant to initi-
ation and regulation of behaviour. Deci and Ryan suggest that external events
can be viewed as having informational, controlling, or amotivating aspects with
respect to motivation. A social context can be considered as supporting one of
these aspects and this is likely to affect an individual’s perceived competence
and intrinsic motivation. The informational aspect provides competence-
relevant information in the context of choice. When this aspect is salient, it is
likely to promote intrinsic motivation. The controlling aspect undermines
feelings of self-determination by pressuring people to behave, think, and feel
in particular ways. When the controlling aspect is salient, intrinsic motivation
is undermined. The amotivating aspect signifies that a sense of competence

cannot be attained by participating in the behaviour. When the amotivating
aspect of an external event is salient, people feel amotivated and they do not
intend to engage in the behaviour (Wild and Enzle 2002).
Applications of cognitive evaluation theory

Cognitive evaluation theory has important ramifications for practice because
predictions can be made about the impact that various environmental and
internal events exert on intrinsic motivation (Deci and Ryan 1985, 2000). For
example, cognitive evaluation theory predicts that external events such as
rewards and feedback that do not support competence and promote an
external perceived locus of causality will undermine intrinsic motivation,
whereas events that affirm competence and support an internal perceived
locus of causality will enhance intrinsic motivation. A number of experi-
mental studies have examined effects of various external events such as tan-
gible rewards, verbal rewards, and interpersonal context on intrinsic motiv-
ation. A meta-analysis of experimental studies adopting cognitive evaluation
theory found strong effects of tangible rewards on intrinsic motivation and
behaviour (Deci et al. 1999a). Specifically, the analysis showed that the
expectation of tangible rewards such as money, trophies, and scholarships
resulted in a decrease in intrinsic motivation. In addition, the analysis found
that the effects of expected rewards on intrinsic motivation varied as a func-
tion of how rewards are introduced or communicated (Deci et al. 1999a).
When rewards are expected and given for being present at an experimental
session, called task-noncontingent rewards, then intrinsic motivation will not be
undermined. This is because expected rewards that are not contingent on
doing the task do not pressure people to behave in particular ways, and there-
fore are less likely to influence intrinsic motivation. Task-noncontingent
rewards are likened to real-life situations in which people are paid for their
occupation.
Research has also shown that task completion rewards that are given for com-
pleting a task undermine intrinsic motivation because people feel obliged to
complete the task to gain the reward. In such cases the reward is experienced as
controlling and results in a shift in the perceived locus of causality from
internal to external (Deci et al. 1999a). In addition, engagement contingent
rewards, which are given for engaging in a task but not completing it, are also
experienced as controlling, and therefore undermine intrinsic motivation.
This is because people have to work on the task to get the reward (Deci et al.
1999a; Deci and Ryan 2000). Further, performance contingent rewards, which are
given for performing an activity well by matching some standard of excellence
or surpassing some criterion, undermine intrinsic motivation because people
have to meet some standard in order to receive the reward. However, when
performance contingent rewards convey competence-affirming information
such as when people meet the standard implied by the reward, they may have
a diminished negative effect on intrinsic motivation. This is because com-

petence-affirming information offsets some of the negative controlling effects
that performance contingent rewards exert on intrinsic motivation (Deci and
Ryan 1985, 2000).
In addition, research has shown that competitively contingent rewards, which
are given to people competing directly with others for a limited number of
rewards undermine intrinsic motivation (Reeve 2002). Competitively contin-
gent rewards and face-to-face competitive situations are highly controlling
because winning is instrumental to attaining the reward (Deci and Ryan 1985;
Reeve 2002). In the domain of exercise, the effects of reward contingencies on
intrinsic motivation have not been examined thoroughly. However, there
have been a number of studies examining the effects of rewards on intrinsic
motivation in sports competitors and these will be reviewed in Chapter 5
(McAuley and Tammen 1989; Vansteenkiste and Deci 2003).
In addition to examining effects of tangible rewards on intrinsic motivation,
several studies have examined the effects of verbal rewards on intrinsic motiv-
ation. Positive feedback that is administered immediately after completion of a
task is an example of a verbal reward. Research has shown that verbal rewards
usually enhance intrinsic motivation because they tend to affirm personal
competence (Deci et al. 1999a). For example, positive feedback that indicates a
person has done well on a task is likely to enhance intrinsic motivation by
increasing their sense of competence. This is in accordance with cognitive
evaluation theory that hypothesizes that perceived competence enhances
intrinsic motivation in the context of self-determination (Deci and Ryan
1985). Therefore, it is important to realize that the effect of a verbal reward on
intrinsic motivation can vary as a function of the context in which the reward
is communicated. If rewards are communicated in a controlling way or imply
evaluation then they are likely to undermine intrinsic motivation (Deci and
Ryan 1985, 2000). For example, if positive feedback is reported in a controlling
manner (e.g. ‘you have done as well as you should’), then intrinsic motivation
is likely to be undermined because the modal operator of ‘should’ facilitates a
sense of being controlled (Deci et al. 1994).
In addition, verbal rewards will only enhance intrinsic motivation if they are
communicated in the context of self-determination (Deci and Ryan 1985).
This can be achieved by simply informing people about their performance
avoiding the use of pressuring language, by instructing people how to self-
administer informational feedback, and by structuring feedback in way that
does not imply evaluation (Deci and Ryan 1985, 2000). In the exercise
domain, Whitehead and Corbin’s (1991) experimental study found that posi-
tive feedback enhanced intrinsic motivation in a physical task. Further, longi-
tudinal studies have supported the effect of perceived positive feedback
on intrinsic motivation (Koka and Hein, in press).
Although cognitive evaluation theory and the concept of intrinsic motiv-
ation have attracted a great deal of scientific interest and debate (Deci et al.
1999a), it is important to realize that cognitive evaluation theory applies only
to behaviours and tasks that are highly interesting, such as puzzles and prob-
lems. The theory does not apply to mundane tasks that are monotonous and
boring and unlikely to be intrinsically motivated such as brushing one’s teeth
or wearing a seat belt. As a consequence, cognitive evaluation theory does not
explain how interest develops or how a boring task can be transformed into an
interesting task. Another limitation of cognitive evaluation theory is that it
does not account for the psychological need for relatedness, despite evidence
that relatedness is an essential and fundamental psychological need (Sheldon
et al. 2001). In fact, Deci and Ryan (2000) have flagged the psychological need
for relatedness as important because it explains why people perform mundane
tasks that are important for personal development and growth but are unlikely
to be intrinsically motivated.
Organismic integration theory

Organismic integration theory is a second sub-theory of self-determination
theory that explains the motivation of non-intrinsically motivating
behaviours on the basis of all three of the psychological needs identified by
Deci and Ryan (2000). The theory proposes that people engage in behaviours
that are unlikely to be intrinsically motivated because of the need for related-
ness. That is, because people are intrinsically motivated to experience satisfac-
tory relationships with others, they engage in non-intrinsically motivated
behaviour in order to function effectively in the social world (Deci and Ryan
1985, 2000). Central to organismic integration theory is the hypothesis that
people engage in non-intrinsically motivated behaviours for the attainment of
extrinsic outcomes such as praise, but it also recognizes that people can
internalize such extrinsically motivated behaviours and eventually accept
them as emanating from the self. Thus, the theory proposes that some
extrinsically motivated behaviours can eventually become self-determined
through the process of internalization (Deci et al. 1994; Deci and Ryan 2000).
A model describing internalization and human motivation from an organ-
ismic integration theory perspective is presented in Figure 3.1. On the right-
hand side of the model is amotivation, referring to a person’s lack of inten-
tionality and sense of personal causation (Ntoumanis et al., in press). Lack of
perceived competence and beliefs that behaviour cannot reliably lead to
desired outcomes can precipitate amotivation. Adjacent to amotivation on the
continuum are the four different forms of extrinsic motivation, namely,
external regulation, introjected regulation, identified regulation, and inte-
grated regulation. Each of these forms of regulation reflects the pursuit of
behaviours to attain external outcomes that are separate from doing the
behaviour for its own sake. However, they do reflect varying degrees of
internalization and self-determination. Internalization of an action is depend-
ent upon receiving praise or positive feedback from significant others to whom
individuals feel attached or related (Deci and Ryan 2000). Behaviours that are
externally reinforced are characterized as externally regulated. Introjected
regulation lies alongside external regulation and it refers to behaviours that are
performed to avoid the pressuring emotions of guilt or shame. External regula-
tion and introjected regulation, therefore, describe less internalized and more
controlling forms of behaviour because they refer to behaviours that are
performed under some form of internal or external pressure.
A less controlling and more self-determined form of external regulation is
identified regulation, which refers to a behaviour that is performed to achieve
personally relevant and valued outcomes. Identified behaviours are therefore
characterized as being more internalized and self-determined. The most
autonomous and least controlling form of externally regulated behaviour is
integrated regulation which refers to identified behaviours that are brought
into congruence with other behaviours and roles that are enacted in life.
Finally, adjacent to integrated regulation lies intrinsic motivation, which
refers to behaviours that are performed for their own sake and not for the
attainment of external outcomes that are independent of the activity itself.
Intrinsically motivated behaviours display many similarities with identified
regulation and integrated regulation. For example, intrinsically motivated,
integrated, and identified behaviours all are characterized by high levels of
enjoyment. However, integrated and identified behaviours imply that people
derive enjoyment from the attainment of the valued outcome rather than the
activity itself. In contrast, the construct of intrinsic motivation implies that
people derive enjoyment from the activity itself and not from the attainment
of valued outcomes.
Organismic integration theory also places great emphasis on the effect of
context on the types of motivation. Contexts that support psychological needs
are hypothesized to promote more self-determined forms of extrinsic motiv-
ation such as identified regulation and integrated regulation, while contexts
that frustrate psychological needs promote less self-determined forms of
extrinsic motivation such as external regulation and introjected regulation
(Deci and Ryan 2000). In general, the theory differentiates between two types
of interpersonal context that can either support or frustrate psychological
needs. Interpersonal contexts, in which significant others encourage choice
and participation in decision-making, provide a meaningful rationale for
doing the behaviour, use neutral language (e.g. use of ‘may’, ‘could’ and not
‘should’, ‘must’) during interpersonal communication, and acknowledge
people’s feelings and perspectives support self determined forms of motivation
(Deci et al. 1994). Conversely, interpersonal contexts in which significant
others do not explain why performance of the behaviour may be important,
use pressuring language during interpersonal communication (e.g. use of
‘should’ and ‘must’), and do not acknowledge difficulties associated with
performance of behaviour tend to frustrate psychological needs.
Figure 3.1 The perceived locus of causality
Notes: ++ = Strong positive correlations expected between these scales
+ = Weak positive correlations expected between these scales
− − = Strong negative correlations expected between these scales
− = Weak negative correlations expected between these scales
0 = No correlation expected between these scales
Applications of organismic integration theory

A measure of the different forms of motivation in organismic integration the-
ory was developed by Ryan and Connell (1989), termed the perceived locus of
causality (PLOC) scale.2 The PLOC measures external regulation, introjected
regulation, identified regulation, and intrinsic motivation. In a preliminary
study, the intercorrelations among these scales were found to conform to a
simplex-ordered structure. A simplex-ordered structure is evident when the cor-
relation between scales measuring adjacent types of motivation such as
external regulation and introjected regulation is higher than the correlation
between dimensions that lie further apart such as external regulation and
identified regulation (see Figure 3.1). Such a pattern indicates the presence of a
continuum which Deci and Ryan (2000) describe as a developmental con-
tinuum of self-determination. However, Deci and Ryan also suggest that sup-
port of a continuum does not preclude the possibility for individuals to
internalize a new behaviour at any point along this continuum depending on
prior experience and current situational factors.
Several studies in a number of contexts have corroborated the importance of
the distinction between self-determined and controlling forms of extrinsic
motivation in understanding behaviour (e.g. Williams et al. 1998; Williams et
al. 2002). In particular, intervention, prospective, and longitudinal panel
design studies have shown that forms of extrinsic motivation such as external
regulation and introjected regulation influence health behaviours and out-
comes directly or indirectly via perceived competence (Williams et al. 1998;
Williams et al. 2002). In addition, perceived autonomy support has been
shown to influence behaviour mediated by self-determined forms of extrinsic
motivation such as identified regulation and perceived competence (Guay et
al. 2001; Hagger et al., 2003b; Hagger et al., in press b).
A number of valid and reliable measures of intrinsic motivation and the
perceived locus of causality have also been developed in the exercise domain.
Most prominent of these is the behavioural regulation in exercise question-
naire (BREQ) developed by Markland and colleagues (2004). Other non-
generic measures have been adopted such as Goudas, Biddle, and Fox’s (1994)
and Hagger et al.’s (2003b) adaptation of items from Ryan and Connell’s
(1989) self-regulation questionnaire for use in physical education contexts.
These instruments have measured external regulation, introjected regulation,
identified regulation, and intrinsic motivation but not integrated regulation.
Despite marginal differences in the items used by these inventories, these
scales have been shown to conform to the simplex-ordered structure as
hypothesized by organismic integration theory (e.g. Chatzisarantis et al.
2003).
Research in an exercise context has also examined utility of self-determined
forms of motivation in the prediction of exercise behaviour. Several studies
have shown that the types of motivation from the PLOC influence exercise
behaviour directly (Ntoumanis 2001; Chatzisarantis et al. 2002; Chatzisarantis
et al. 2003) and indirectly via the mediation of intentions (Hagger et al.
2003b) and attitudes (Hagger et al. 2002a). However, the direct effects of self-
determined forms of motivation on intentions and behaviours are generally
small (Chatzisarantis et al. 1997; Chatzisarantis et al. 2002; Chatzisarantis et
al., in press d; Hagger et al. 2002a).
In exercise and physical education contexts, research has investigated the
influence of self-determined forms of extrinsic motivation and intrinsic
motivation on some key motivational determinants of exercise behaviour.
Goudas et al. (1994) found that perceived competence mediated the effects of
self-determined forms of extrinsic motivation on intentions. These results are
consistent with studies conducted in health psychology that found perceived
competence to partially mediate the effects of self-determined forms of
extrinsic motivation on health outcomes (Williams et al. 2002). Conversely,
several other studies have found that self-determined forms of extrinsic
motivation mediate the effects of perceived competence on exercise intentions
and behaviour (Ntoumanis, 2001; Sarrazin et al. 2002; Hagger et al. 2003b;
Standage et al. 2003). Hence, the role of self-determined forms of extrinsic
motivation in predicting and explaining exercise intentions and behaviour is
complex and there has not been complete convergence in the exact pattern of
influence of these variables on intentions and behaviour.
The trans-contextual model

The trans-contextual model builds upon previous applications of self-
determination theory and aims to explain how motivation in one context,
namely physical education, facilitates motivation, intentions, and behaviour
in a different but related context, namely exercise behaviour during leisure
time. From its outset, research on intrinsic motivation and self-determination
theory has demonstrated the importance of context in influencing perceived
locus of causality (Deci and Ryan 1985). Contextual factors such as presenta-
tion of tasks and support of choice with a clear rationale, acknowledgement of
conflict, and informational feedback have been shown to enhance intrinsic
motivation (Deci et al. 1994), while rewards, threats, evaluation, and deadlines
have been shown to undermine intrinsic motivation (Deci and Ryan 1985).
The results of this research have led to recommendations suggesting that qual-
ity and quantity of motivation can be enhanced by environments that support
psychological needs and an internal perceived locus of causality (Deci and
Ryan 2000). Following this, research has supported these premises in natural-
istic settings, demonstrating that perceived autonomy support influences
behaviour in a motivational sequence (Chatzisarantis et al., in press b). The
sequence dictates that perceived autonomy support affects perceived locus
of causality, which in turn influences motivation and behaviour. The
trans-contextual model builds upon this evidence and postulates that the
teaching styles of physical educators influence the forms of motivation that
pupils endorse in physical education classes. Specifically, it is suggested
that autonomous supportive physical education teachers facilitate intrinsic

motivation or self-determined forms of extrinsic motivation whereas teachers
whose teaching style is not autonomy-supportive facilitate controlling forms
of extrinsic motivation (Figure 3.2).
The trans-contextual model also proposes a cross-contextual interplay
between motives, such that intrinsic motivation in one context such as phys-
ical education can lead to intrinsic motivation in another, such as leisure time.
In turn, this increased motivation can lead to participation in related activities
in a different context such as leisure-time exercise behaviour (Vallerand 1997).
The trans-contextual model therefore proposes that intrinsic motivation and
self-determined forms of extrinsic motivation in physical education classes is
more likely to facilitate intrinsic motivation with regard to leisure-time exer-
cise behaviour than controlling forms of extrinsic motivation. This is because
people tend to search for opportunities to re-enact intrinsically motivating
behaviours in relevant contexts (Chaiken 1980).
Recent research on intentional behaviour has also indicated that the inte-
gration of constructs from self-determination theory and the theory of
planned behaviour provide complementary explanations of the unexplained
processes within each theory. The inclusion of self-determination theory in
social cognitive models has helped researchers to explain the quality of the
relationships in the theory of planned behaviour (Chatzisarantis et al. 1997;
Chatzisarantis and Biddle 1998; Sheeran et al. 1999a) and the antecedents of
the theory of planned behaviour variables (Chatzisarantis et al. 2002; Hagger
et al. 2002a). For example, Chatzisarantis et al. (1998, 1997) have demon-
strated that the predictive utility of the theory of planned behaviour in an
exercise context can be increased by the inclusion of volitional intentions.
Similarly, Hagger et al. (2002a) have shown that an internal perceived locus of
causality has a pervasive impact on attitudes and perceived behavioural con-
trol with respect to exercise behaviour. However, the strong correlation
between intrinsic motivation and intentions to participate in exercise
behaviour was completely mediated by attitude and, to a lesser extent, by
perceived behavioural control. These results were replicated by Chatzisarantis
and co-workers (2002), who also demonstrated that an internal perceived
locus of causality determined the effort people invested in pursuing exercise
behaviour. Such research acknowledges the role of different types of motiv-
ation in explaining the bases for intentions, attitudes, subjective norms, and
perceived behavioural control. It also recognizes that the theory of planned
behaviour has utility in providing a basis for the translation of general motives
from the perceived locus of causality into intentional action (Chatzisarantis et
al. 2002; Hagger et al. 2002a).
In congruence with the findings of these studies examining the impact of
perceived locus of causality on intentions and behaviour within the theory of
planned behaviour (Chatzisarantis et al. 1997; Chatzisarantis and Biddle
1998; Sheeran et al. 1999a; Chatzisarantis et al. 2002; Hagger et al. 2002a), the
trans-contextual model predicts that motives of perceived locus of causality in
Figure 3.2 The trans-contextual model
a leisure-time context will influence exercise intentions and behaviour

only via the mediation of the belief-based constructs of attitude, subjective
norms, and perceived behavioural control. Thus the perceived locus of
causality is presumed to act as a source of information for the formation of
attitudes, subjective norms and perceived behavioural control, and, indirectly,
intentions.
Thus far, research has shown that the autonomy-supportive teaching styles
of physical educators influence participation in exercise behaviour during
leisure time via the mediating role of intrinsic motivation, attitudes, perceived
behavioural control, and intentions. In addition, evidence has shown that
perceived autonomy support in physical education influences exercise
behaviour during leisure time even after effects of past behaviour are taken
into consideration (Hagger et al. 2003b). Further, preliminary results suggest
that the effects of the trans-contextual model have been replicated across four
cultures in Europe and Asia (Hagger et al., in press b). However, it is important
to note that there are no intervention studies or large-scale studies that exam-
ine the impact that teaching styles have on motivation for exercise behaviour.
Further, little is known about the influence that peer groups and parents
exert on exercise behaviour. Therefore, one direction for future research is to
evaluate effectiveness interventions based on the trans-contextual model in
promoting exercise behaviour.
Practical recommendations based on self-determination theory
Perceived autonomy support

In terms of practical recommendations, Reeve (2002) identifies the specific
behaviours exhibited by those involved in the promotion of exercise
behaviour that are likely to support autonomy in individuals seeking to
increase their exercise levels. One way is to avoid the use of external incentives
and controlling, non-informational forms of feedback as the only contingency
for doing the activity (Deci and Ryan 1985). Clearly, however, this is only a
small aspect of the provision of an autonomy supportive environment.
Behavioural strategies exist that enhance motivation such as providing the
individual with opportunities to choose and express themselves in exercise
contexts (Chatzisarantis et al., in press a), fostering an environment focusing
on activities involving task or mastery rather than the ego (Ames 1992), and
avoiding competition and external rationales for participation (Ryan et al.
1984). Furthermore, providing appropriate feedback that is informational and
competence-related is important in terms of monitoring progress (Deci et al.
1994).
Motivational interviewing
One technique that has been shown to be efficacious in changing people’s
exercise motivation and behaviour is motivational interviewing (Rollnick
and Miller 1995; Miller 1999), and this has recently been linked with
self-determination theory approaches to motivation (Markland 2004). Motiv-
ational interviewing is a clinically developed intervention that aims to alter
problem behaviours by investigating the nature of the problem with the
‘client’ and assisting them in self-directed behavioural change. Rollnick and
Miller, the central proponents of motivational interviewing, define it as,
a directive, client-centered counselling style for eliciting behavior change

by helping clients to explore and resolve ambivalence. Compared with
nondirective counselling, it is more focused and goal-directed. The
examination and resolution of ambivalence are its central purpose, and
the counselor is intentionally directive in pursuing this goal. (1995: 325)
Originally developed in the context of alcoholics and problem drinkers,

motivational interviewing should not, according to Rollnick and Miller, be
confused with a set of techniques that are applied to the client or person with a
problem, but an ‘interpersonal style’ that is adopted by a clinician or inter-
viewer in the pursuit of directing a client-centred programme of cognitive and
behavioural change. Importantly, this set of styles should not be considered as
confined to clinical settings, but can be applied to interventions and contexts
wherever behaviour change is salient.
According to Rollnick and Miller, motivational interviewing contrasts with
traditional approaches to clinical interventions for behaviour change because
it removes the conflict caused by the confrontational styles, argumentation,
and direct persuasion often seen in other clinical approaches. Rather, it adopts
an approach that is directed by the interviewer, but the focus is primarily on
conflict resolution and behaviour change that is based upon the client or indi-
vidual’s own personal suggestions. As such, the traditional resistance and scep-
ticism often expressed by individuals undergoing traditional confrontational
means of behaviour change are allayed and the client is much more
empowered to pursue changes to his/her own problem behaviour that are
suggested by they themselves. Furthermore, there are no direct suggestions
made by the clinician or interviewer, and thus no direct persuasion ‘tech-
niques’ are used as with interventions and campaigns based on the theory of
planned behaviour covered in the previous chapter. Motivational interviewing
does not therefore involve dictating changes to individuals or involve point-
ing a finger at their unhealthy lifestyle, instead it should be viewed as a part-
nership between the clinician or interviewer and patient or person undergoing
change.
The starting point of motivational interviewing is investigating and resolv-
ing the ambivalence that surrounds the problem behaviour. The ambivalence
is characterized by conflicts between the problem behaviour and other, usually

healthier, courses of action. The client’s task in motivational interviewing is
therefore to elicit their own reasons for change, not those suggested by the
interviewer. It is therefore the role of the client to express personally relevant
benefits (pros) and costs (cons) of changing the difficult behaviour. For
example, a client may make statements like: ‘I actually do like exercising, it
makes me feel good, and I know it’s good for me losing weight, but it’s far too
much effort when you have kids and I just can’t seem to fit it in.’ It is the role of
the interviewer to help the client investigate both sides of this and direct them
toward a resolution that may lead to motivation and sustained action toward
changing behaviour. Therefore, ambivalence is viewed by the motivational
interviewing approach as the primary obstacle toward changing people’s
behaviour, and once this is overcome, behaviour change will follow, provided
the information and skills are given to the individual undergoing change.
In addition, there are four other core principles that must be adhered to at all
times by the interviewer when conducting a motivational interview: (1)
express empathy; (2) support self-efficacy; (3) roll with resistance; and (4)
develop discrepancy. The client expects to be told in the interview that their
behaviour is unhealthy and needs to change, instead principles of motivating
interviewing compel the interviewer to acknowledge their understanding of or
empathy with the difficulties that the interviewee encounters with the problem
behaviour and trying to change. There is also no single manner to change
behaviour and the interviewer should provide support to the client in terms of
supporting their confidence or self-efficacy towards the strategies that they
adopt. In addition, the interviewer should not counter any resistance with
arguments or by challenging the views of the client, this may only create more
resistance, instead the interviewer ‘rolls’ with the resistance by acknowledging
the conflict in the individual and using the momentum to investigate the
conflict and possible means for resolution. Finally, motivational interviewers
aim to direct the client into highlighting discrepancy between their current
behavioural status and their ideal situation. Together these principles underlie
the motivational interviewing approach and assist in preventing relapse or
resistance often associated with other approaches.
Motivational interviewing is conducted via a ‘menu of strategies’ that is
adopted by the interviewer to direct the client towards their investigation of
ambivalence. The menu is varied according to the state of readiness expressed
by the client and verified by the interviewer prior to the interview. Import-
antly, in keeping with the ethos of the approach, any resistance, defensive
manoeuvres, or scepticism expressed by the client is not viewed as a failure by
the client to comply with the wishes of the interviewer, but is viewed as a
failure of the interviewer in estimating a state of readiness to change that is
more advanced than the interviewee’s current status. An example of the menu
of strategies, a summary of the general approach and aims of each strategy,
and example statements likely to be evoked from the client for each strategy
is given in Table 3.1. The menu of strategies is adopted alongside the core
Table 3.1 Menu of strategies for motivational interviewing
Strategy Purpose Type of comments from

interviewer
Reviewing a typical day Builds rapport and focuses ‘Can we spend the next 5–10
on entire lifestyle not minutes going through a day
specific problem from beginning to end. What
behaviours happened, and how did you
feel? Is there any time you could
have fitted in exercise? Let’s
start at the beginning.’
Looking back Explores what life was like ‘So things have really changed.
before they experienced Tell me a little more about what
health problems like life was like back then.’
obesity, low activity
Good things and less Explores the pros and the ‘You said that your weight has
good things cons of the problem affected your self-confidence.
behaviour i.e. lack of Tell me about a time when that
activity happened.’
Discussing the stages of Interviewer introduces ‘Can you think of any ways
change stages of change and enters that you could change from
into a discussion with being someone who is aware of
client, client encouraged to the need for change (contempla-
provide ways of moving/ tor) and someone who is getting
changing stage ready to make a change
(preparer)?’
Assessment feedback Interviewer provides client ‘Would it be okay if I offered a
with a summary of their little information to you based
achievements so far on what we’ve talked about so
far? Correct me if I’m wrong
about anything . . .’
Values exploration Clients encouraged to ‘Think about the amount of
explore their ‘ideal self’ and exercise you do now, how much
compare it to their ‘current would you have to change so
self’ to reveal their values that you were doing the amount
of exercise that is ideal for you?’
Looking forward Asks the client to explore ‘What do you think would hap-
two different futures: one pen in future if you made the
in which they made changes you have said to your
changes to their unhealthy exercise? Compare that to what
behaviour and the other if would happen if you didn’t
they did not. make those changes.’
Exploring importance Asks clients to explore the ‘Thinking about the changes
and confidence importance of each change you have said to your exercise
they have proposed and habits, how confident do you
how confident they are at think you are in making those
achieving the change changes?’
Table 3.1 (cont’d)
Strategy Purpose Type of comments from

interviewer
Decisional balance Client is asked to list the ‘Can you list all of the pros and
pros and cons of making cons of doing exercise behaviour
behavioural change – in future?’
similar to good things and
less good things
Change planning Used if a client is at ‘What are the possible things
advanced stage of readiness, you could do in the next month
interviewer explores to increase your current level of
possible actions that could exercise? Try to come up with as
change behaviour in future many as you can.’
principles in the pursuit of client-centred change. Importantly, the change

does not have to be actual behaviour change. Instead, successful change is
often established on the basis of movement from one stage of change to
another. Thus increases on readiness to change is considered a successful out-
come rather than actual behaviour change. This can often be reflected using
assessments such as a decisional balance sheet. These changes are expected to
be progressive over a series of interview sessions in which readiness for change
is evaluated. Motivational interview sessions are expected to last about 40
minutes with several sessions (up to 6), but recent research has resulted in the
development of brief motivational interviewing that last approximately 20
minutes and with single sessions (Rollnick and Miller 1995).
Research with motivational interviewing has generally supported its
effectiveness in a number of behavioural domains, including exercise.
Research in an exercise context has reported motivational interviewing as hav-
ing applicability to this behaviour in primary care settings, however, few ran-
domized control studies have used motivational interviewing as an interven-
tion technique to promote exercise behaviour. For example, Harland et al.
(1999) conducted a randomized control trial that examined the effectiveness
of an intensive motivational interviewing intervention (6 interviews over 12
weeks), a brief interviewing session (1 interview), and a financial incentive
(vouchers) on participants’ exercise behaviour. The intensive interview inter-
vention group exhibited 55 per cent more exercise behaviour than control after
6 weeks but the vouchers and brief interviewing patients did not exhibit a
significant increase, suggesting that intensive motivational interviewing is the
most effective in sustained exercise change.
While motivational interviewing seems to be efficacious, it seems to be
based on limited theoretical evidence and testable hypotheses and evidence
for the mechanisms for its effectiveness are limited. Miller admitted not
devoting sufficient attention to the provision of a comprehensive theoretical

background to motivational interviewing and agrees that ‘we do not have, in
my view, a satisfactory explanation of why and how motivational interview-
ing works’ (1999: 2). Motivational interviewing has been only loosely allied
with the trans-theoretical model, particularly the stages of change, self-efficacy
theory, cognitive dissonance theory, and client-centred therapy (Miller 1999).
Recently, Markland (2004) has drawn parallels between the processes adopted
by motivational interviewing and self-determination theory. His suggestion is
that motivational interviewing may alter motivation and behaviour by chan-
ging the key constructs from the theory. Markland suggests that motivational
interviewing creates an interpersonal environment likely to foster a sense of
self-determination and support the autonomy of individuals trying to change
their behaviour. It does this by supporting autonomy through means such as
providing positive feedback that is competence related, by the provision
of an appropriate structure to enhance intrinsic motivation through the
identified regulation of personally relevant goals and the presentation of clear
contingencies between behaviour and outcome, and by the involvement of
the individual in the process of determining goals, courses of action, and
planning for change. Markland summarizes these parallels in Figure 3.3.
Further parallels can be drawn between self-determination theory and
items from the guiding principles and menu of strategies in motivational
interviewing. These parallels, in particular, arise from the types of feedback
put forward by Deci et al. (1994) likely to facilitate the internalization of
externally regulated behaviours in the repertoire of behaviours that would be
important to personal functioning and an autonomous existence. These types
of feedback were provision of a clear rationale, acknowledgement of conflict,
and the provision of informational feedback. In the core principles of motiv-
ational interviewing, the development of discrepancy provides informational
feedback that is self-directed regarding the relationship between personal
behaviour (e.g. doing exercise) and personally relevant goals (e.g. losing
weight). In the menu of strategies, the examination of the ‘good things and
the bad things’ and the exploration of value items provide a self-directed
rationale for doing the behaviour. Finally, the setting of personally relevant
goals that address these values may shift the locus of causality for the
behaviour towards identified regulation – because the behaviour is related to
a specific value. Furthermore, expressing empathy is akin to acknowledging
conflict because the interviewer provides some understanding of the aspects
of the behaviour that the interviewee does not want to do and the conflict
that leads to ambivalence. In summary, it seems that motivational interview-
ing is an effective strategy to change exercise behaviour, provided it is
sustained, and it may be that motivational interviewing changes self-
determination theory constructs that may be responsible for sustained
behavioural change.
Figure 3.3 Relations between psychological needs and motivational styles

from self-determination theory and the core principles from motivational
interviewing
Source: Markland (2004)
Suggested reading
Deci, E.L. and Ryan, R.M. (2002) Handbook of Self-Determination Research. Rochester, NY:
University of Rochester Press. The most comprehensive reader of self-determination
theory to date with contributions from many of the influential authors in the area.
Gollwitzer, P.M. (1999) Implementation intentions: strong effects of simple plans,
American Psychologist, 54: 493–503. Gollwitzer’s lucid review of his own work on
implementation intentions.
Hagger, M.S., Chatzisarantis, N., Culverhouse, T. and Biddle, S.J.H. (2003) The processes
by which perceived autonomy support in physical education promotes leisure-time
physical activity intentions and behavior: a trans-contextual model, Journal of
Educational Psychology, 95: 784–95. A review of our own work on self-determination
theory and the theory of planned behaviour and introduction to the trans-contextual
model.
Summary
• Social cognitive theories of intention explain how intentions to exercise are

formed, but are inadequate in explaining how exercise intentions are con-
verted into behaviour.
• Several properties of intention likely to enhance the relationship between
exercise intentions and behaviour have been identified including intention
stability, scale correspondence, intention formation, self-schema, and
volitional versus forced intentions.
• The formation of an implementation intention in which an actor specifies
when, where, and how they are going to exercise has been shown to
enhance the intention–behaviour relationship. This is because the state-
ment of such cues serve to enhance the link between intention and
behaviour and reduce the likelihood of an exerciser forgetting to carry out
their intentions.
• Self-determination theory is a theory of motivation that aims to explain
motivation on the basis of psychological need satisfaction. The theory has
two influential sub-theories: cognitive evaluation theory and organismic
integration theory.
• Cognitive evaluation theory hypothesizes that context will affect the qual-
ity of motivation experienced by exercisers. The quality of motivation is
characterized by a perceived locus of causality, a continuum of different
forms of motivation ranging from intrinsic to extrinsic. Informational
contexts tend to enhance intrinsic motivation, a sense that exercise is per-
formed through a sense of choice, enjoyment, and interest. Controlling
contexts tend to undermine intrinsic motivation and behaviours are
viewed to be controlled by others and performed out of a sense of
obligation.
• Organismic integration theory explains how intrinsic motivation is
developed and how extrinsically motivated behaviours can become
internalized or integrated and viewed as emanating from the self with a
locus of causality closer to intrinsic forms of motivation.
• Perceived autonomy support is a means to enhance intrinsic motivation
through key autonomy-supportive behaviours of teachers and instructors
such as non-contingent positive feedback.
• Motivational interviewing is a client-centred counselling approach to
changing exercise behaviour and adopts the basic principles of expressing
empathy, supporting self-efficacy, rolling with resistance, developing
discrepancy, and a menu of strategies to change interviewee’s readiness
to change. Parallels with this approach have been drawn with self-
determination theory in that it supports autonomy, provides an appropriate
structure for self-determined change in exercise behaviour, and involves the

client in the process of change.
Notes
1 This chapter will focus on cognitive evaluation theory and organismic integration
theory because these have been applied most frequently to explaining behaviour in a
physical activity context.
2 Cognitive evaluation theory has also used the construct of perceived locus of causality
to explain effects of external events on intrinsic motivation. The difference between
perceived locus of causality as conceptualized by cognitive evaluation theory and
organismic integration theory is that organismic integration theory considers per-
ceived locus of causality to comprise four dimensions (external regulation, introjec-
tion, identification and intrinsic motivation). In contrast, cognitive evaluation theory
considers perceived locus of causality to comprise two dimensions (extrinsic versus
intrinsic).
4
Exercise and the physical self
The self is one of the most widely researched and focal concepts in social,
educational, and personality psychology. According to Oyserman, ‘Self-
concept and identity provide answers to the basic questions “Who am I?”,
“Where do I belong”, and “How do I fit (or fit in)” ’ (2004: 5). The self is
therefore considered to be of high importance as it is implicated in a person’s
decision-making, motivation, and behaviour. It is also intrinsically implicated
in our choices, such as the behaviours we choose to pursue and the groups we
choose to affiliate with (Deci and Ryan 2000). As a consequence, the self is an
integral part of volition. This considered, it is not surprising that the self is
prominent in many theories that aim to explain variance in human behaviour
and motivation, such as self-efficacy theory (Bandura 1997), and self-
determination theory (Deci and Ryan 1985). Such theories suggest that social
cognitive variables related to the self are integral to self-regulatory behaviour.
Researchers have identified both personal and social identities, and these
strongly affect people’s behaviour, particularly in private and public domains.
Indeed, theories aimed at explaining group dynamics and inter-group
behaviour such as social identity theory, view the self and the self-esteem
paramount. This is because the way in which we view ourselves is also tied in
with the groups to which we belong. Self-categorization theory, a sub-theory
of social-identity theory, suggests that people are motivated to preserve in-
group consistency and be less tolerant or favourable towards outgroup mem-
bers when the self is vulnerable. Suffice to say, the self, like attitude, is an
essential and important construct in social psychology.
Similarly, the self is an essential construct in theories of motivation and
behaviour in exercise and sport contexts. Along with dieting, exercise is the
only means available other than invasive surgery, for an individual to modify
their physical appearance in terms of body fat content and muscle tone. Phys-
ical appearance is an important aspect of the self and plays an important role
in determining self-perception. As a result, aspects of the self associated with
physical appearance are likely to be implicated in the decision-making
72 Exercise and the physical self
processes related to engage in exercise behaviour. In addition, self-esteem may

be an important psychological outcome of exercise behaviour. We have already
noted that participation in regular exercise is a determinant of a number of
psychological variables, such as positive affect and well-being. Self-esteem or a
positive self-regard is also likely to be affected by regular exercise participation,
particularly if that participation is accompanied by the achievement of positive
outcomes such as the attainment of personal goals and positive affect. Self-
esteem may also be part of the process by which exercise behaviour determines
positive outcomes. For example, psychological well-being may arise from
exercise behaviour providing increased positive self-regard, in which case self-
esteem may mediate the influence of exercise behaviour on psychological well-
being. This chapter will review and evaluate the contribution that the physical
self has made to social psychological explanations of exercise behaviour. The
theoretical structure and effects of self-esteem in the physical domain will be
covered, as well as behavioural phenomena related to the self and exercise such
as eating disorders and exercise addiction.
Defining self-esteem
What is self-esteem?
A perusal of the social psychology literature will reveal that self-related ter-
minology is both multitudinous, varied, and, at times, inconsistent. A variety
of terms have been introduced that refer to the same or similar constructs, and
tend to be used interchangeably: self-esteem, self-worth, self-concept, self-
description, self-regard, self-perceptions, and self-image. All these terms have
been used at one time or another to refer to the construct of self-esteem.
Early researchers adopted the term self-concept and defined the construct as
statements about the self such as ‘I have brown hair’ or ‘I am an athlete’
(Rosenberg 1979). However, recent theories have made the distinction
between purely descriptive statements such as those listed previously and
more evaluative statements about the self ‘I like Italian food’ or ‘I am good at
tennis’ (Harter 1996). The latter statements therefore not only describe attrib-
utes of the self, as the individual sees himself or herself, but qualify it with
opinions of ‘worthiness’. Often this distinction is encapsulated in the adop-
tion of the term ‘self-description’ to refer to the purely descriptive, non-
evaluative statements about the self, and ‘self-esteem’ or ‘self-concept’ to refer
to statements that contain evaluative information, although this distinction is
not always clear. Among contemporary theorists there is a general consensus
that self-esteem comprises the perceptions that individuals have regarding
themselves, incorporating both descriptive and evaluative content (Harter
1996). This has not, however, stemmed the tide of indiscriminate adoption
and use of the various self-related terms, and has introduced some degree of
ambiguity in the self-esteem literature. Suffice to say, the use of self-esteem
Exercise and the physical self 73
terms in the design, investigation, and discussion of research on the self

should be treated with caution and with close attention to the meaning of the
terminology adopted. We will therefore operationally define self-esteem as
both descriptive and evaluative self-related statements.
As a social psychological construct, self-esteem is attractive because
researchers have conceptualized it as an influential predictor of pertinent out-
comes, such as academic achievement (Marsh 1990) or exercise behaviour
(Hagger et al. 1998). In addition, self-esteem has also been treated as an
important outcome in itself due to its close ties with psychological well-being
(Marsh 1989), and self-esteem may also predict motivational tendencies as
people seek behaviours in areas of competence in order to maintain or
enhance self-perceptions. Indeed, one key assumption in research on self-
esteem is that it represents the virtuous, ideal, and generally positive elements
of an individual’s personality, and by implication this means that the
enhancement of self-esteem is a desirable, adaptive outcome. Therefore, self-
esteem has evaluative and descriptive components, and has efficacy both as an
independent and dependent variable in social psychological models that aim
to explain behaviour. Further, it has been implicated in the mechanisms of
such models in the prediction of behaviour. Accordingly, self-esteem has been
‘variously conceptualized as a dependent, independent, mediator, moderator
construct’ (Oyserman 2004: 5).
However, two notes of caution should be heeded at this juncture. First, we
have generally focused on self-esteem as an adaptive, or positive outcome or
influential construct. While this may be true, there is evidence to suggest that
high levels of self-esteem could be accompanied by maladaptive outcomes
and behaviours (Sedikides and Gregg 2003). Theorists suggest that it is non-
contingent self-esteem, i.e. high self-regard that is not dependent upon
clearly defined goals or outcomes and self-esteem striving without intrinsic
goals that may be implicated in such social ills (Crocker and Luhtanen 2003).
This issue will be visited in greater detail in the next section. Second, self-
esteem has thus far been proposed to be implicated in a number of social
psychological processes that underlie motivation. However, some researchers
have argued that the influence of self-esteem on behavioural and psycho-
logical outcomes may be spurious and its effect may fall away with the
inclusion of other motivation-related constructs (Miller and Downey 1999).
Furthermore, if the observed effects are present between self-esteem and
psychological and behavioural outcomes, the direction of causality cannot
be inferred, so studies with sophisticated longitudinal, panel designs may be
necessary to establish the true role of self-esteem in an exercise context
(Sedikides and Gregg 2003).
Models of self-esteem
When self-esteem research was in its infancy, self-esteem was largely regarded
as a dispositional, unitary construct, much like the way in which global self-
esteem is viewed in more contemporary models of self-esteem (Rosenberg
1979). Self-esteem was therefore viewed as a global, all-encompassing con-
struct that comprised all self-related statements made by an individual. This
conceptualization was limited because it was global; self-esteem constructs
were too general and distal from the specific behaviours they were proposed to
predict to account for a large amount of variance. However, these early models
represented pioneering research into the self and represented a ground-
breaking attempt to define, conceptualize, and quantify self-esteem.
Another major criticism of the unidimensional conceptualization of self-
esteem was that it did not account for multiple dimensions of self-esteem
(Marsh and Shavelson 1985). Self-esteem researchers therefore proposed a
global or overarching conceptualization of self-esteem comprising a number of
self-esteem evaluations made by individuals based on their experiences and
reinforcements in a variety of behavioural domains and contexts (Shavelson
et al. 1976; Marsh and Shavelson 1985). This instigated the possibility that self-
esteem could be relatively high in one context or domain, but low or com-
promised in another. This made conceptual, empirical, and intuitive sense,
individuals demonstrate competency or lack of competence in a variety of
domains, and a person’s sense of self draws on experiences and engagement in
behaviours in each of these domains. The multidimensional model proposed
that overall or global self-esteem comprised self-esteem in areas like academic,
social, physical, and occupational domains (Marsh and Shavelson 1985). It
was an important advancement because it permitted the estimation of the
relative contribution of self-esteem statements to global or overall self-
esteem. In addition, it also permitted the examination of each of the domain-
levels self-esteem statements with respect to each other i.e. inter-domain
relationships.
Shavelson, Hubner, and Stanton (1976) adapted the multidimensional
model to account for the issue of generality, proposing that self-esteem was
organized hierarchically with global self-esteem sitting at the apex of a hier-
archy governing domain-level constructs that, in turn, determined self-esteem
in the different contexts. A schematic representation of the model is given in
Figure 4.1. The self-esteem domains are one level removed from the global self-
esteem construct in terms of generality, and are therefore more specific, less
enduring, less stable, and more subject to change from external constructs. It
was proposed, however, that the hierarchy further operated within the
domains, so that the domains could be further subdivided into more sub-
ordinate constructs that reflected more specific self-esteem evaluations within
each context. For example, academic self-esteem would comprise self-esteem
evaluations in the subdomain areas of mathematics, English, science, and so
on. In addition, within each subdomain, further subordinate facets could be
Figure 4.1 A multidimensional and hierarchical model of self-esteem
Source: Shavelson et al. (1976: 408)
identified until, theoretically, the level of specificity could be traced to indi-

vidual perceptions of competence at the situational level, that are constantly
subject to change, relatively unstable, and highly dependent on context.
Therefore, the different levels of the hierarchy of self-esteem are distinct on the
basis of generality and stability. Global self-esteem is regarded as relatively
stable and enduring compared with domain and subdomain-level self-
evaluations. Subdomain level and further sub-divisions reflect situational,
more transient, and less stable evaluations of the self.
Nearly two decades of research into the structure of self-esteem have pro-
vided considerable support for the multi-faceted and hierarchically organized
model proposed by Shavelson, Hubner, and Stanton (1976). This model has
considerable advantages over other unidimensional approaches (Rosenberg
1979) because it recognizes that self-esteem arises from multiple sources and
operates in a variety of contexts. The multidimensional, hierarchical model is
advocated by leading researchers and has received support in many domains,
especially educational settings (e.g. Marsh and Shavelson 1985; Harter 1988;
Marsh 1989). Marsh and co-workers have provided strong evidence for the
validity of a broadly stated self-esteem model that incorporates many domains
(e.g. Marsh and Shavelson 1985; Marsh 1990). This research has led to the
development of a series of rigorously tested psychometric instruments to test
this model, known as the Self-Description Questionnaire (SDQ) series (Marsh
and O’Niell 1984). The latest version, SDQ-III, measures self-esteem in 13 dif-
ferent areas: physical abilities, physical appearance, opposite-sex relation-
ships, same-sex relationships, parent relationships, honesty/trustworthiness,
spiritual values/religion, emotional stability, verbal/reading, mathematics,
problem solving, general school, and general self-esteem. The SDQ series has
demonstrated construct, discriminant, and cross-cultural validity (Marsh and
O’Niell 1984; Marsh 1990). The multidimensional, hierarchical model has
also been adopted by researchers using a divergent approach in their research
to examine the effects of self-esteem on numerous dependent variables such as
academic performance, academic achievement, psychological well-being
(Marsh 1990), and perceived competence (Harter 1996). Such studies reinforce
the predictive validity of a multidimensional, hierarchically organized model
of self-esteem. In addition, researchers adopting a convergent style of research
have shown enhanced self-esteem to be the outcome of a number of psycho-
logical processes such as competence (Harter 1996), perceived ability (Marsh
1990), and perceived autonomy support (Reeve 2002). In sum, the proposed
model of self-esteem has been supported and rigorous testing of self-
description questionnaires has provided self-esteem measurement instru-
ments yielding valid and reliable scores.
One key advantage of the Shavelson et al. (1976) model is that it permits the
detailed study of self-esteem in a single domain while simultaneously
maintaining the relevance of the domain to global self-esteem. Adopting the
model for this purpose enables the study of the organization and predictive
validity of domain-relevant self-esteem statements but does not isolate the
domain-level self-perception from global self-esteem. Instead, the relative con-

tribution of the domain to global self-esteem and the mediation of the sub-
domain facets by the domain-level construct as implied by the hierarchy are
explicitly modelled. This makes the multidimensional and hierarchical model
of self-esteem a versatile and adaptable model for the examination of the struc-
ture of self-esteem in the physical domain. Once validated, it also permits the
study of the effects of a number of motivational process variables on self-
esteem as an outcome and the effect of self-esteem, as an independent vari-
able, on a number of salient outcome variables in the physical domain. The
next section will outline the application of multidimensional, hierarchical
models of self-esteem in the physical domain.
Multidimensional and hierarchical models of self-esteem in exercise

and physical activity
While the SDQs measured physical self-esteem as part of the elaborate multi-
dimensional and hierarchical model of self-esteem proposed by Marsh et al.
(1985), specific sub-facets of the physical domain were not examined. This is in
contrast to the inclusion of specific mathematics and English sub-facets iden-
tified as part of the general school or general academic domain in the SDQ-III
(e.g. Marsh and O’Niell 1984). As a result, researchers interested in self-esteem
in the physical domain have adopted the Marsh et al. model to study the
structure of sub-facets specific to this domain (Fox and Corbin 1989; Marsh
et al. 1994) and the impact of self-esteem on health-related behaviours such as
exercise behaviour (Sonstroem et al. 1994) and physical fitness components
(Marsh and Redmayne 1994). Fox and Corbin (1989) were the first to propose
an elaborated model in this regard, and introduced a multidimensional,
hierarchical model of physical self-esteem. They adopted the proposed
structure of Shavelson et al. (1976) and Marsh et al. (1985) and the profile
approach of Harter (1988) to develop the accompanying measure, the Physical
Self-Perception Profile (PSPP).
Fox and Corbin (1989) proposed that a general physical self-esteem was
superordinate to four subdomain factors: sports competence, physical condition-
ing, body attractiveness, and physical strength. In keeping with the Shavelson
model, general physical self-esteem mediated relations between the sub-
domains and global self-esteem at the apex of the hierarchy (see Figure 4.2).
Conceptually, global self-esteem is the most general, enduring, and stable con-
struct and is least likely to change; it is, in effect, trait-like in nature. The
subdomains are more changeable, less stable, and less enduring, they are there-
fore considered more state-like in their outlook. This is congruent with the
original Shavelson et al. (1976) model that proposed top-down and bottom-up
effects in the hierarchy, so that specific situational experiences of competence
in the difference subdomains areas effect change in the upper levels, while the
upper levels are used as a source of information for motivational decisions
in specific exercise experiences (see also Sonstroem and Morgan 1989). A
Figure 4.2 Fox and Corbin’s (1989) multidimensional and hierarchical

model of physical self-esteem
Source: Fox and Corbin (1989: 14)
schematic diagram of these effects is given in Figure 4.3 with examples from
the sports competence and body attractiveness PSPP domains. The factor
structure of the proposed model was supported in a number of studies using
both exploratory and confirmatory factor analytic methods (Fox and Corbin
1989; Marsh et al. 1994; Sonstroem et al. 1994), and has been shown to be
cross-culturally invariant (Asçi et al. 1999). In addition, the predictive validity
of the model has been supported (Kowalski et al. 2001) and components of the
model have been shown to be important outcomes of exercise behaviour (Fox
2000).
Marsh and Redmayne (1994) elaborated on Fox and Corbin’s initial model
and rigorously tested an alternative instrument that adopted the same multi-
dimensional, hierarchical approach. The Physical Self-Description Question-
naire (PSDQ, Marsh and Redmayne 1994; Marsh et al. 1994) comprised nine
subdomain scales, four of which were proposed to be equivalent to the PSPP
subdomains. The strength, appearance, condition/endurance, and sport scales
of the PSDQ were equivalent to the physical strength, body attractiveness,
physical condition, and sports competence scales from the PSPP, respectively.
Additional subdomain scales in the PSDQ generally referred to aspects of phys-
ical fitness: flexibility, health, coordination, activity, and body fat (see Figure 4.4).
The scale demonstrated excellent factorial validity on the basis of several
confirmatory factor analyses (Marsh and Redmayne 1994; Marsh et al.
1994). Furthermore, concurrent and predictive validity of the additional scales
was confirmed alongside objective measures of children’s physical fitness
Figure 4.3 Examples of specific sub-facets of two subdomains of Fox and

Corbin’s model of self-esteem
Source: Adapted from Fox (1990: 4)
components (Marsh and Redmayne 1994). Marsh et al. (1994) also provided
some concurrent validity for the PSDQ in relation to competing instruments
adopting the same theoretical model, namely the PSPP and the physical self-
esteem scale proposed by Richards (1988) using multi-trait, multi-method
(MTMM) analyses.
While Fox and Corbin’s (1989) and Marsh and Redmayne’s (1994) models
have exhibited satisfactory validity in terms of both the multidimensional
nature and hierarchical arrangement of the factors, recent research has ques-
tioned the hierarchical nature of the model (Marsh and Yeung 1998; Kowalski
et al. 2003). While cross-sectional data supports the existence of the hier-
archical model, and has even rejected competing models that break the
hierarchy with direct effects from the subdomain level to global self-esteem
(Hagger et al., in press a; Sonstroem et al. 1994), when tested longitudinally
there is only very weak support for the hierarchy. Using the physical and
academic domains from the SDQ, Marsh and Yeung (1998) adopted a one-year
cross-lagged panel design and found little support for top-down (global
self-esteem predicting domain and subdomain-level self-esteem over time) or
bottom-up (subdomain-level self-esteem predicting domain-level and global
Figure 4.4 Marsh and Redmayne’s (1994) multidimensional and hierarchical

model of physical self-esteem, including additional scales introduced in
subsequent revisions of the model by Marsh et al. (1994)
self-esteem over time) effects. This supported the notion of multidimensional-

ity, but did not support the hierarchy. Adopting a similar design, Kowalski et
al. (2003) examined top-down, bottom-up, and longitudinal effects using the
PSPP physical self-esteem constructs over a one-year period, and their results
again supported a horizontal rather than hierarchical model. Future studies
would examine these effects in the hierarchical physical self-esteem constructs
over variable time periods.
The effect of physical self-esteem on exercise behaviour

Both Fox and Corbin’s (1989) and Marsh and Redmayne’s (1994) models of
physical self-esteem have been used to predict exercise behaviour. Theoretic-
ally, physical self-esteem may act as an antecedent of exercise behaviour
because perceptions of competence in a given domain, or how confident they
feel that exercise will enhance their sense of self, will compel a person to
perpetuate those sensations of competence. A number of studies have found
that the PSPP subscales could adequately distinguish between active and non-
active individuals, with physical conditioning and sports competence
accounting for the most variance (Hagger et al. 1998). Strong relationships
have been found between the PSPP subdomains and prospective exercise
behaviour. For example, Marsh and Redmayne (1994) found significant correl-
ations between the activity PSDQ subdomain and self-reported exercise
behaviour, but self-reported exercise behaviour was also significantly related to
the endurance, strength, flexibility, coordination, sport ability, and general
physical self-esteem. It is evident that these results suggest that a multitude of
physical self-esteem components may be implicated in participation in phys-
ical exercise. However, the processes behind the relationship are not easily
understood without motivational models to explain the pattern of influence
(Biddle 1997), and a process model of self-esteem and exercise behaviour will
be presented in subsequent sections (Sonstroem and Morgan 1989; Sonstroem
et al. 1994).
Studies of self-esteem in an exercise context
Physical self-esteem in children and young people

Recently, there has been increased interest in the importance of physical self-
esteem in young people, particularly in the light of guidelines recommending
the promotion of exercise behaviour in young people (Sallis and Patrick 1994).
The Fox and Corbin model of physical self-perceptions has demonstrated
adequate validity in young people and shown considerable utility in the
prediction of exercise behaviour. Whitehead (1995) introduced a children’s
version of Fox and Corbin’s physical self-perception profile (C-PSPP) and sub-
sequent validation studies have supported the proposed structure in young
people (Hagger, Biddle et al., in press a). Furthermore, physical self-esteem

components have been positively related to exercise behaviour. For example,
Hagger et al. (1998) and Raudsepp, Liblik, and Hannus (2002) showed the
subscales of body attractiveness, physical strength, and sports competence to
be positively related to exercise behaviour in children. These data support the
structure of the Fox and Corbin model of physical self-esteem in young people
and indicate that physical self-perceptions are an important influence on
exercise participation in young people in leisure-time and physical education
contexts. There may also be important developmental changes in physical
self-esteem components, and perhaps an effect for Shavelson et al.’s (1976)
proposed differentiation, or increased diversity in self-esteem with increasing
age and experience, within the physical domain. These issues will be visited in
the next section.
Age and gender differences in physical self-esteem

Examining the effects of age and gender in physical self-esteem, Whitehead
and Corbin (1997) found that C-PSPP scale scores were typically one-half point
lower in female adolescents than males in research with Fox and Corbin’s
(1989) model. Hagger, Biddle, and Wang (in press a) conducted a meta-analysis
of the original data from these studies and found that these differences
equated to medium effect sizes across subdomains in the studies by Whitehead
(1995) (median Cohen’s d = 0.34) and Marsh et al. (1994) (median Cohen’s
d = 0.38). These findings support the notion that adolescent males tend to
have higher physical self-esteem ratings than their female counterparts. One
problem with these studies is that they do not support the multidimensional,
hierarchical structure of the self-esteem constructs in the samples they were
testing. Marsh and co-workers (1989) provided support for the structure of
self-esteem across age and gender groups before examining the main and
interactive effects of age and gender on self-esteem levels, and cited this as a
necessary condition to confirm that the differences were true differences
and not just artefacts of structural change such as increased self-esteem
differentiation (Shavelson et al. 1976; Marsh 1989).
In the physical domain, a recent study adopting a representative sample of
2949 school children aged 12 to 14 years supported the structure of Fox and
Corbin’s model (1989) within the age and gender groups (Hagger, Biddle et al.,
in press a). Tests of difference in the levels of the constructs revealed significant
gender differences, with boys scoring significantly higher than girls in all of
the physical self-esteem subdomains, and in global self-esteem. There were
also significant age effects, with eighth-grade children scoring higher on the
body attractiveness, physical strength, and general physical self-esteem C-
PSPP scales. In summary, research trends suggest that there are significant
gender and grade differences in physical self-esteem with boys and younger
children tending to score higher in the PSPP subdomains. Gender differences
may be explained by the fact that competence in the physical domain is
stereotypically masculine in nature, while age differences may reflect self-

esteem differentiation (Shavelson et al. 1976) but more likely indicate more
realistic impressions of self-esteem with increasing age (Marsh 1990).
Elite athletes and self-esteem

Elite and high-level athletes represent a specific sub-population whose pursuit
of achievement in sport may well result in this being manifested in signifi-
cantly higher perceptions of physical ability or sports competence than those
in the normal population. Marsh et al. (1995) examined SDQ-III scores in
representative samples of mixed-gender and female athletes and non-athletes.
They found that athletes scored significantly higher on a number of the PSDQ
subdomains, but most markedly for the physical ability scale. Other sub-
domains that exhibited significantly higher levels in athletes included same
sex, opposite sex, and parent relationships and global (Marsh et al. 1995).
Given these differences, one can speculate that high-level athletic participa-
tion has a positive effect on physical ability scores. Future research is needed to
identify the level of sports participation required to produce this effect or
whether the effect is a linear one. In addition, future research will differentiate
between sport type to examine whether these demographic variables affect the
perception of physical ability. For example, one hypothesis might be that per-
ceived physical ability may be different across sports with gross-muscle
movements compared with those requiring more fine motor skills.
Exercise and physical self-esteem model

Sonstroem and Morgan (1989) proposed a conceptual model aimed at explain-
ing the process by which exercise experience affects physical competence and
self-esteem. Sonstroem proposed that the effects of situational experiences
of competence in sport and exercise settings affected global self-esteem in a
bottom-up fashion. The proposed exercise and self-esteem model hypothesized
that situation-specific estimates of competence, or self-efficacy (Bandura 1977,
1997), influenced self-esteem in the physical domain mediated by physical
competence. A schematic diagram of the model is given in Figure 4.5. In the
model, physical competence is closely related to self-esteem in the physical
domain, while global self-esteem is included because it acts as an indirect
measure of psychological well-being. In addition, the model also includes
physical self-acceptance, a variable proposed to influence physical self-esteem
along with self-competence (Sonstroem and Morgan 1989). Physical accept-
ance is an individual’s subjective personal regard for themselves in a given
domain, regardless of levels of physical competence. In Figure 4.5, the pre-
intervention state of the model reflects baseline correlations among the study
constructs, and the directional relationships in the post-intervention section
of the model reflects the pattern of influence among the study variables after
experiences of competence in an activity in the physical domain. In this
Figure 4.5 Sonstroem and Morgan’s (1989) exercise and self-esteem model
Source: Sonstroem and Morgan (1989: 333)
respect the model is dynamic and implies continued modification of self-

esteem, the outcome variable. Finally, an additional premise of the model is
that global self-esteem affects exercise behaviour in a top-down fashion, such
that the impact of global self-esteem on exercise behaviour is mediated by the
competence and self-efficacy paths.
The model has been tested in a number of studies and results have generally
supported its major propositions. Sonstroem et al. (1991) used confirmatory
factor analysis and structural equation modelling to test the proposed model
on cross-sectional data. The factor analysis supported the construct and fac-
torial validity of the measures and factors adopted, and significant structural
relations were found between self-efficacy and physical competence con-
structs, and between physical competence and global self-esteem as measured
by Rosenberg’s (1979) self-esteem scale. There was no direct effect of self-
efficacy on self-esteem, as predicted. In addition, the dynamic premises of
the model were tested longitudinally across a competitive swim season
(Sonstroem et al. 1993). Results indicated significant increases in self-esteem,
perceived competence, and skill level over the season. However, an autoregres-
sive path analysis indicated that changes in skill level were not caused by
self-esteem or perceived competence. Additional tests of the model have been
proposed to examine its ability to account for variance in exercise behaviour. It
was expected that the expanded exercise and self-esteem model would operate
in a top-down fashion in the prediction of exercise behaviour (Sonstroem et al.
1994). Results indicated that self-efficacy was the most proximal predictor of
exercise behaviour, and mediated the impact of the physical self-esteem sub-
domains of physical condition and body attractiveness on exercise behaviour.
Cross-cultural developments in physical self-esteem

Many of the studies conducted on physical self-esteem are undertaken in
Western European contexts with high individualist and moderate-to-low col-
lectivist cultural orientations. In cultures with a predominantly collectivist
cultural norm, individuals have a greater tendency to view their self as inter-
dependent with other members of society, while in cultures with a largely
individualist cultural norm individuals tend to regard their self as independent
of others (Triandis 1995). This is particularly relevant for self-esteem research,
given findings that suggest that people’s self-esteem may be organized into
private and collective views of the self (Trafimow et al. 1991). Furthermore,
some researchers have argued that self-esteem is a largely individualist phe-
nomenon, and may not be relevant to cultures that have a collectivist notion
of the self (see Triandis 1995).
Researchers have therefore sought to examine the cross-cultural validity of
the physical self-esteem constructs, particularly in nations that generally
endorse a high collectivist cultural norm. While some studies have replicated
the multidimensional and hierarchical model of physical self-perception
in individual samples from different cultures (e.g. Asçi et al. 1999), true
cross-cultural evaluation should involve comparisons across at least three cul-
tures (Marsh et al. 2002). Marsh and co-workers (2002) and Hagger, Lindwall,
and Asçi (2004) have examined the cross-cultural validity of the two most
prevalent measures of a multidimensional and hierarchical model of physical
self-esteem in the literature, the PSDQ and the PSPP respectively. Marsh et al.
(2002) examined the feasibility of the factor structure of the PSDQ in two
different European cultures, Australia and Spain, and a Middle Eastern culture,
Turkey. Results indicated that the factor structure, i.e. the same number and
arrangement of factors and questionnaire items, was invariant across cultures,
supporting the generalizability of the model in these cultures. Further, in a
recent study Hagger et al. (2004) examined both the factor structure and mean
differences in the levels of the PSPP constructs across three diverse cultures in
Northern Europe (Sweden), the Middle East (Turkey), and Western Europe
(Great Britain). Results supported the replicability and generalizability of the
factor structure in all three cultures, but, most importantly the PSPP scales
tended to be rated more highly in the British sample. Cross-cultural studies
have also evaluated the cross-cultural validity of Fox and Corbin’s (1989)
model of physical self-esteem in children and results have corroborated find-
ings in adult studies (Raudsepp et al. 2002; Hagger et al. 2003a; Hagger, Biddle
et al., in press a) and supporting the hypothesis that individualist cultures to
endorse the physical self more than collectivist cultures.
Eating disorders in exercise and sport and exercise dependency
While regular exercise yields positive health benefits, particularly with respect
to healthy weight management, individuals who participate in regular exercise
for the control of body weight may be vulnerable to psychological disorders
that are manifested in compulsive dietary and exercise patterns. Among these
disorders are anorexia nervosa and bulimic nervosa. In addition, disorders
relating to excessive and compulsive exercise participation may also arise
among exercisers. Given that diet and exercise are the primary means available
to individuals to modify their weight and body image, it is not surprising that
some have argued that both dieting and activity-related disorders arise, some-
times concurrently, in exercisers, and have similar aetiology. Furthermore,
there is evidence to suggest that a compromised body image or physical self-
esteem may be a risk factor for these conditions to arise. This section will
define these conditions, outline their prevalence, identify their causes, and
propose potential interventions to ameliorate their incidence.
Eating disorders and exercise

Anorexia nervosa and bulimia nervosa are considered psychological disorders
classified by the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV,
American Psychiatric Association 1994) and occur in a relatively small number
of people among both athletic and non-athletic populations. The vast major-
ity of eating disordered people are female; only an estimated 5 to 15 per cent of
eating disordered people are male (Andersen 1995). Anorexia nervosa occurs
in between 0.50 and 1.50 per cent of the population in the United States
(American Psychiatric Association Work Group on Eating Disorders 2000) and
between 0.00 and 0.90 per cent in the countries of Europe (Institute of Psych-
iatry 2004). Clinical definitions of anorexia in the DSM-IV make reference to a
refusal or inability to maintain a body weight that is within 85 per cent of the
expected weight of the person for a given height (American Psychiatric Associ-
ation 1994). In young people, anorexics are classed as unable to reach an
expected increase in weight during a period of growth, which is particularly
relevant to athletic at-risk groups such as gymnasts. It is also characterized as
an intense fear of becoming fat or overweight, even when clearly underweight,
and extreme distortions in perceived body image and low physical self-esteem.
Anorexics develop unusual eating habits, such as avoiding food and eating
small quantities, and avoiding social situations centred about food like
restaurants. Anorexic people may repeatedly check their body weight and
engage in other behaviours to control their body weight and composition,
such as intense and compulsive exercise, or purging by means of vomiting and
abuse of laxatives, enemas, and diuretics. Girls with anorexia often experience
a delayed onset of their first menstrual period, which is often the case of young
gymnasts.
Bulimia nervosa occurs in between 1.10 and 4.20 per cent of the population
in the United States (American Psychiatric Association Work Group on Eating
Disorders 2000) and in between 0.00 and 3.90 per cent of European popula-
tions (Institute of Psychiatry 2004). Bulimia is defined by the DSM-IV as a
condition in which an individual consumes excessive amounts of food in a
short space of time with a concomitant sensation of loss of control and follows
this by compensatory behaviour to prevent weight gain such as purging
(vomiting), misuse of laxatives and diuretics, enemas, excessive exercise, and
fasting. The frequency of the binge–purge cycle should be of the order of at
least twice a week for three months for a clinical diagnosis. In some bulimics
no purging occurs, but other compensatory behaviours are adopted, known as
the non-purging type. Similar to anorexics, body image tends to be distorted
and mainly focused about body shape and weight. Often anorexia and bulimia
tend to coincide and up to 50 per cent of anorexics develop a binge–purge
cycle similar to that in bulimics (National Institute of Mental Health 1993).
Bulimics, in terms of diagnostic criteria, do not have the clinically defined low
body weight exhibited by anorexics.
While the incidence of these eating disorders in the general population is
relatively small, their profile in the media is exceptionally high, and this has
been attributed to links made between the portrayal of the body, particularly
the female body, in magazines, television, film, and other media, especially in
advertising (Kalodner and DeLucia-Waack 2003). One reason for this might be
the exceptionally young average age for the onset of eating disorders, between
16 and 18 years. Ironically, those portrayed in the media have an increased
tendency to succumb to eating disorders in order to perpetuate the prevalent
waif-like images among fashion models in the media and advertising. How-
ever, these eating disorders have a complex aetiology, and their development
may be the result of the prevalence of a number of risk factors. Further, there
may be factors unique to sport and exercise that result in the development of
eating disorders among athletes and exercisers.
The potential health risks of anorexia and bulimia are serious and have the
potential to result in mortality. In anorexics, the health risks are consistent
with severe malnutrition such as extreme muscle wastage, not only in the
skeletal muscle but also in respiratory and cardiac muscle, amenorrhea in
females, anaemia, and loss of bone mineral leading to conditions such as
osteoarthritis (Becker et al. 1999). The physical symptoms of anorexia nervosa
are excessive weight loss and skeletal muscle wastage, loss of hair, fatigue, and
low energy. Psychologically, eating disorders can lead to depressed mood and
depression. Mortality in anorexic people and, in fewer cases, bulimic people is
often due to complications from the condition such as cardiac arrest or electro-
lyte imbalance. In athletes, there is also enhanced potential for injury due to
fatigue and overuse injury as well as increased potential for bone fractures if
bone mineralization is limited (Yates 1999; Golden 2002). In addition, female
athletes with low body fat tend to have intermittent periods of amenorrhoea
which makes it difficult to detect the amenorrhoea that results from an eating
disorder (Golden 2002).
Numerous factors, biological, demographic, cultural, and social psycho-
logical have been identified as contributing to the development of eating dis-
orders. Research suggests that the development of eating disorders may have a
genetic component and that a complex combination of inherited traits and
environmental situations may give rise to the development of eating disorders
(Walters and Kendler 1995). As introduced earlier, such environmental factors
may be the culture of ‘thinness’ observed in media images in the Western
world. Globalization, the increase of media access in developed nations, and
the trend toward thin, waif-like physiques for females has been attributed to
the increased incidence of eating disorders in the past century (Toro et al.
1994). In terms of psychological factors, low self-esteem and body image
(Ackard et al. 2002), high trait anxiety (Lehoux et al. 2000), elevated anger
and depression (Breaux and Moreno 1994), high levels of social anxiety
(Hinrichsen et al. 2003), and a compulsive/dependency personality type
(Bornstein 2001) may also contribute to the development of an eating
disorder.
In athletes, there may be additional factors that lead to the development of
eating disorders. There is evidence to suggest that athletes with eating dis-
orders may differ from eating disordered people from a normal population. For
example, Madison and Ruma (2003) found that athletes have a lower correl-
ation between the severity of their eating disorder and psychopathy with level
of exercise, suggesting that activity level is less associated with a disordered
psychological profile in athletes. In addition, Martin and Hausenblas (1998)
indicated that among aerobic instructors, levels of eating disordered symp-
toms were low, and much lower than other athletic populations, suggesting
that it may be competitive aspects of sport that could lead to eating disorders
in athletes. It may therefore be that exercise per se is not an essential aspect of
the eating disorder for athletes, possibly because they were exercisers before
they developed their condition. Furthermore, the focus of athletes’ eating
behaviour modification may not be linked to appearance per se. Instead, an
emphasis on a lean physique and the association with such a physique to
performance may result in an athlete developing eating and associated
behavioural patterns that, over time, become disordered. Halley and Hill
(2001), for example, found that eating-disordered elite runners had much
lower self-esteem, mental health, and placed considerable emphasis on a lean
physique compared with runners who did not have an eating disorder.
Furthermore, social factors may be important influential factors on an
athlete’s development of disordered eating patterns. Although perhaps
unintentional, coaches, managers and parents, particularly those of young

athletes, may convey an impression to the athletes that their performance is
contingent on appearance (Griffin and Harris 1996). In some cases, this mes-
sage is very overt and coaches have been known to be very critical of ath-
letes’ weight and appearance, which can at best undermine the confidence of
the performer and at worst can contribute to the development of disordered
patterns of eating and exercise. Furthermore, in sports such as gymnastics,
figure-skating, ice-dance, and diving there is external judging criteria which
may provide additional external pressures on the need for the athlete’s
appearance to conform to an ideal-performer stereotype. In addition, like
non-athletic populations of eating disordered individuals, low self-esteem
has been implicated as a risk factor for the development of eating disorders in
athletes, as at-risk groups tend to have low self-esteem as eating disordered
athletes (Yates 1999). All these criteria are external to the athlete and,
coupled with compulsive personality traits, may result in the development of
eating disorders.
Treatment for eating disorders in athletes often has additional complica-
tions because often weight loss per se is not the sole reason for the manipula-
tion of the diet. Furthermore, the detection of eating disorders in athletes may
be difficult given that athletes that do not have eating disorders share the same
personality traits as those that do (Thompson and Sherman 1999). One of the
main challenges is raising the awareness of the problem with the athlete, with
the secondary problem of getting them to curtail their training regime or give
it up altogether (Yates 1999). Since sport and competition are central to the
athlete’s self-esteem and may also be the focus of their disordered eating
behaviour, it is not easy to convince them to stop and show that their activity
is compromising their health, a likely paradox in the mind of an athlete. Regu-
lar screening of athletes, early detection, and counselling have been shown to
be efficacious in raising awareness and curtailing the disorder (Golden 2002).
Once detected and awareness raised, the treatment methods adopted are often
similar to those in non-athletes. Athletes are often referred to counselling and
clinicians, depending on the severity of the disorder, particularly if weight loss
is substantial and life-threatening. Much emphasis is placed on support from
coaches and family members (Ryan 1992). Furthermore, there are education
programmes to raise coaches’ awareness of potential problems and to educate
them to adopt appropriate language and behaviours that do not give messages
to athletes that could be misinterpreted and lead to psychological profiles
matching those of eating disordered individuals and result in persistent diet-
ary modification (Ryan 1992). Specific recommendations for coaches are to
seek specialist help if a case of an eating disorder in an athlete is suspected,
avoiding comments about physique and weight, make referrals, and to provide
information about nutrition, diet, and eating disorders. Importantly, coaches
are advised to be open in their approach to eating disorders, but to encourage
athletes to talk confidentially about their weight and diet concerns if they feel
the need.
Exercise addiction, exercise dependence, and activity disorders

In the first four chapters of this book, the weight of research evidence seems to
favour the view that regular exercise is beneficial to health with a vast array of
positive physiological and psychological outcomes. This is clearly not the case
for the individuals who exercise to the extreme, which can lead to damaging
outcomes such as injury, maladaptive motivational profiles, and negative
affective states. This section will outline the issues relating to excessive exer-
cise, establish whether excessive and persistent exercising is an ‘addiction’,
outline the clinical basis of such disorders, highlight the effects and psycho-
logical antecedents of these disorders, and introduce strategies that have been
proposed to treat the disordered individual.
The term ‘exercise addiction’ refers to the condition in which individuals
develop disordered patterns of exercise characterized by extreme levels of exer-
cise and this has largely been abandoned in favour of terms like activity dis-
orders (Yates 1991) or exercise dependence (Cockerill and Riddington 1996).
Exercise dependence is defined as ‘a craving for leisure-time physical activity
that results in uncontrollable excessive exercise behaviour and that manifests
in physiological and/or psychological symptoms’ (Hausenblas and Symons-
Downs 2002a: 90). Such a broad definition of activity disorders based on
clinical observation and psychological constructs has been favoured because
evidence to support a biological basis for such disorders and align them with
other clinical addictions has been limited (Cockerill and Riddington 1996). For
example, evidence that exercise dependence is based on an ‘addiction’ to
endogenous opiates like β-endorphins has not conclusively been linked to
exercise dependency (Pierce et al. 1993). Instead, the condition has been more
associated with personality and psychological constructs that are thought to
influence this pattern of exercise behaviour (Cockerill and Riddington 1996).
It is difficult to reconcile exactly the level or extent of participation in exercise
with the existence of an activity disorder. In other words, exercise dependence
makes reference to the experience of the excessive patterns of exercise rather
than the extent of the exercise per se. Therefore, an excessive level of exercise is
a necessary but not sufficient condition for exercise dependence to arise. The
example given previously regarding the high-level athlete illustrates this
point. Often athletes who compete at a high level may exhibit levels of activity
that are indicative of an activity disorder, but exhibit none of the characteristic
patterns of psychological disturbance that are associated with such a disorder.
It is therefore the manner in which the exercise is experienced by an individual
and how it impacts on their lifestyle that are indicative of an activity disorder.
Some researchers have made the distinction between a primary form of
exercise dependence that is independent of dietary manipulation and second-
ary exercise dependence that arises concurrent with an eating disorder (Veale
1995), although research findings have subsequently questioned this distinc-
tion. However, Yates (1991) claims that the development of eating disorders
shares a similar aetiology with the development of compulsive and disordered
patterns of exercise and physical activity. Yates identified a group that had
developed these unusual patterns of exercise behaviour among male runners.
Yates et al. interviewed a number of male runners who covered more than 50
miles per week in training. The majority of those interviewed were high
achievers who participated in running because it gave them a sense of
achievement and felt it contributed to their psychological well-being and their
emotional stability. For these runners, their exercise patterns were considered
adaptive and contributing to overall psychological functioning. However, a
small percentage of the runners for whom their sport ‘instead of . . . contribut-
ing to their adaptation, running had become their adaptation. They seemed
locked into and controlled by the activity. These men were dubbed the
“obligatory” runners’ (1991: 26). The obligatory runners were characterized by
very high running mileage (up to 70 or 80 miles per week), but most import-
antly, their patterns of behaviour were consistent with the obsessive and com-
pulsive behavioural patterns of anorexic or bulimic patients. Their lifestyle
revolved about their running and was characterized by its disciplined, well-
ordered, and inflexible routine. The obligatory runners maintained a strict,
regimented running programme and often subordinated other aspects of their
life such as diet, occupation, and socializing to their running. Indeed, dieting
and exercise, Yates argues, are ‘interrelated methods of managing the body’
(1991: 49), and therefore obligatory runners are often subject to obligatory
dieting, particularly among females.
The detrimental effects of obligatory or compulsive exercise are numerous.
There is risk of over-use injury, potential for excessive weight loss which
results in similar health difficulties as anorexics, psychological burnout, mood
disturbance, depression, eating disorders, and low level of psychological well-
being (Hausenblas and Symons-Downs 2002a, 2002b). As a group, compulsive
exercisers like Yates’ obligatory runners tend to claim that they are content and
happy with their life when they are running, and become depressed and frus-
trated when they become injured and cannot run. However, a different picture
emerges when obligatory runners are interviewed several months after an
injury that has forced them to stop running. They often cite that their running
was ‘a chore they forced themselves to engage for reasons they did not com-
pletely understand’ (Yates 1991: 59). Thus, an obligatory runner exhibits a
maladaptive pattern of motivation, consistent with what Ryan and Deci
(1989) call an introjected locus of causality. Such a motivational orientation
suggests that although the activity has been partially integrated into the per-
son’s repertoire of behaviours, it still remains outside a sense of true self.
Importantly, individuals with these conditions participate for largely extrinsic
reasons and tend to be afflicted by guilt and shame when they do not complete
their daily routine. Therefore, individuals with ‘activity disorders’ tend to
exhibit low levels of psychological well-being and a maladaptive pattern of
motivation. Furthermore, they have the potential to result in serious harm to
themselves due to injury and excessive weight loss.
The obligatory runners are exemplars of activity-disordered individuals
whose exercise and dieting behaviours are defined as clinically compulsive.

Yates (1991) argues that activity disordered individuals, like anorexics and
bulimics, are compulsive rather than ‘addictive’ disorders like drug and alco-
hol addictions. Activity disorders may appear on the surface to be ‘addictive’
because activity disordered individuals obsessively engage in exercise and
seem unable to moderate their participation, just as a drug addict or alcoholic
finds themselves unable to moderate their intake of their drug (Griffiths 1999).
Furthermore, some theorists argue that activity disordered individuals do gain
a physiological ‘reward’ through their disordered behaviour such as increased
levels of β-endorphins, linked to the blood–brain reward system (Yates 1991),
although there is little evidence that levels of β-endorphins are related to
degree of exercise dependence (Pierce et al. 1993). Moreover, this is an over-
simplification of the disorder, as addictive or obsessive behaviours are funda-
mentally different to compulsive behaviours. Drug addicts and alcoholics
enjoy their behaviours and the euphoric state they achieve as the result of
their behaviour, their behaviour is termed ego-syntonic, and therapy seeks to
motivate individuals to cease their damaging impulsive behaviour. Addicts are
content to relinquish self-control over their addiction. However, as noted earl-
ier, activity disordered individuals do not enjoy their exercise or sport, are
introjected-oriented towards their actions, but engage in their behaviour
because they feel that they should. Their compulsive behaviour is driven by an
intense desire to exert self-control over their life and behaviour, and their
behaviour is called ego-dystonic. Their behaviour is motivated by negative
affective states such as fear of losing control, guilt, and shame, even though
they may wish to stop their pattern of behaviour. Therapy therefore focuses on
helping compulsive individuals to moderate their behaviour and demonstrate
that they can exert control over their lives without the need for compulsive
behaviours. Yates claims that activity disorders share their compulsions with
anorexics and bulimics, in that these individuals exert similar control over
their behaviour and physical appearance, and feel compelled to diet, binge,
and purge in order to maintain their control.
The complex pattern of behaviour observed in exercise-dependent indi-
viduals develops as a result of multiple risk factors and antecedents. These are
related to theoretical explanations of the disorders. As stated previously, bio-
logical bases of activity disorders such as β-endorphin and arousal from the
sympathetic nervous system have not been supported (Pierce et al. 1993), and
current research suggests that personality characteristics along with situational
factors may predispose individuals to becoming activity disordered (Hausen-
blas and Giacobbi 2004). Furthermore, the pathology of the development of
activity disorders is similar, Yates argues, to those that result in conditions like
anorexia and bulimia. This is corroborated by research that has associated the
diagnosis of an activity disorder with measures of disordered eating patterns
(Keski-Rahkonen 2001). A number of personality factors have been linked to
the development of activity disorders, such as an obsessive-compulsive ten-
dency (Davis 1999), extroversion (Yates 1991), trait anxiety (Spano 2001), and
perfectionism (Hausenblas and Symons-Downs 2002b). Research has also

shown that parents of anorexics and bulimics emphasize independent
achievement and outcome-oriented success in school and occupation (Yates
1991). Some researchers suggest that failure to succeed under these pressures
can result in the development of disordered eating patterns like anorexia to
exert control over themselves and ‘achieve’ in the outcome of losing weight,
and a high achievement or outcome orientation may be implicated in the
development of activity disorders, as well (Yates 1991). In addition, family or
parents who attach a great deal of value to the control of weight, appearance,
and exercise may also serve as risk factors for the development of an activity
disorder. Finally, compromised self-image and appearance self-esteem are
implicated in the development of exercise dependence (Hausenblas and
Symons-Downs 2002a).
Detection of exercise dependence or activity disorder can be achieved using
validated questionnaire methods and also by adopting an appropriate diag-
nostic framework from clinical guidelines (Ogden et al. 1997; Hausenblas and
Symons-Downs 2002b). Adams, Miller and Kraus (2003) have proposed the
criteria for inclusion of exercise dependence in accepted clinical diagnostic
manuals such as the DSM-IV (American Psychiatric Association 1994). A rec-
ognized pattern of observed and self-reported symptoms of exercise depend-
ence has been put forward, supported by evidence from the development of
exercise dependence diagnostic tools. Converging evidence has identified the
following characteristic symptoms of exercise dependence: tolerance or
increased requirement of exercise to gain desired effect, withdrawal symp-
toms, an intention effect or taking more exercise than intended, lack of control,
time spent in activity, reduction of other activities, activity viewed as an end
rather than a means to an end, and continuance even in the face of persistent
injury (Ogden et al. 1997; Hausenblas and Symons-Downs 2002a, 2002b). The
exercise dependence questionnaire (EDQ, Ogden et al. 1997) and the exercise
dependence scales (EDS, Hausenblas and Symons-Downs 2002b) are psycho-
metrically sound inventories that adhere to the DSM-IV criteria for the classifi-
cation of a clinical disorder and have been validated in clinically diagnosed
populations.
Interventions and therapy to address exercise dependence and activity dis-
orders are drawn from both the psychological and clinical literature. Yates
(1991) advocates the use of cognitive behavioural therapy which is a combin-
ation of the two most effective clinical techniques for the treatment of psycho-
logical disorders: behaviour therapy and cognitive therapy. Behaviour therapy
aims to intervene and break the cycle of situational cues to action and the
conditioned maladaptive behavioural responses to those cues. Examples of
this are social situations such as drinking in a bar or public house coupled with
the habitual response of lighting a cigarette. Cognitive therapy aims to iden-
tify and change the thought patterns or beliefs that underpin the maladaptive
behaviour. This is usually achieved through the investigation of the thought
processes that often accompany situations that evoke the undesired
behavioural response and produce rational alternatives. Yates (1991) suggests

that the substitution of alternatives activities for the disordered behaviours
during therapy is a useful means of correcting the behavioural response to the
situational cues that give rise to the maladaptive behavioural pattern. Therap-
ists adopting cognitive behavioural therapy initially focus on behavioural
means to control the disordered behaviour, and this is then followed by the
investigation and substitution techniques to modify the dysfunctional beliefs,
and thereafter focus on maintaining both the behavioural and cognitive
changes.
Other intervention techniques include information giving and educational
programmes and are suitable for mild activity-disordered individuals as their
behaviour is often perpetuated by misconceptions and fallacies regarding
exercise and its physiological and psychological effects (Sundgot-Borgen
2000). Educational sessions with an activity-disordered individual should be
followed up with goal setting with reasonable targets focused on optimal
health. For example, activity-disordered individuals are likely to have persist-
ent injuries and a goal could be to reduce the incidence of these injuries
(Carnes and Sachs 2002). Goals could also involve activity substitution such
that the activity disordered individual is encouraged to substitute exercise with
other engaging and rewarding activities such as socializing or going for a walk.
Role models or self-help groups may be of assistance in this regard. Although
exercise dependence is a relatively rare condition, the inclusion of self-help
groups where activity-disordered individuals can discuss and share their
experiences may have positive effects on the success of changes made towards
reducing dependence upon exercise (Carnes and Sachs 2002).
Suggested reading
Fox, K.R. (1997) The Physical Self. Champaign, IL: Human Kinetics. A recent com-
pendium of research on self-esteem and exercise with a heavy focus on hierarchical
models.
Yates, A.B. (1991) Compulsive Exercise and the Eating Disorders. New York: Brunner/Mazel.
A readable clinical text on social psychological aspects of exercise and eating
disorders.
Summary
• Physical self-esteem is the descriptive and evaluative perceptions about the

self in physical contexts.
• Physical self-esteem is often viewed in the context of multidimensional
and hierarchical models of physical self-esteem, with general or global
self-esteem at the apex, physical self-esteem at the domain level and more
specific facets of self-esteem such as physical appearance, physical strength,
sports competence, coordination, flexibility, and physical conditioning at

the subdomain level.
• Physical self-esteem is a modest predictor of exercise behaviour and
research has suggested that physical self-esteem tends to be higher in males
and elite-athletes.
• Anorexia nervosa and bulimia nervosa are clinical psychological disorders
that result in severely altered eating patterns and pose a risk to health. These
occasionally occur in athletes and exercisers with risk factors including por-
trayal of body images by the media, association of performance with lean
physique, and judging criteria.
• Interventions to assist eating-disordered individuals include hospitaliza-
tion, cognitive behavioural therapy, counselling and psychotherapy, and
drug therapy, and coaches of at-risk athletes are required to refer athletes
with suspected eating disorders for professional assistance.
• A tendency to engage in exercise and physical activity in an obsessive man-
ner with a lack of control, increased tolerance for exercise, withdrawal
symptoms, a tendency to take more exercise than intended, a reduction of
other activities, and persistence when injured are indications of an activity
disorder.
• Intervention strategies to help activity-disordered individuals include
cognitive behavioural therapy and information giving/educational
strategies.
Part II
The social psychology of sport

5
Social psychology and motivation in sport
Motivation is central to many theories that aim to explain the complex

processes involved in people’s goal striving and attainment in achievement
contexts such as school or university (e.g. Reeve 2002), the workplace (e.g.
Stajkovic and Luthans 1998), and sport (e.g. Duda 1993). In sport, motivation
is recognized as an important and necessary ingredient for success, and has
been a focal construct since the outset of social psychological investigations
into sport behaviour. As is often the case in applied social psychology, every-
day use of the word ‘motivation’ is usually unstructured with little consensus
or consistency in its meaning. For example, when a person describes another
as ‘highly motivated’ it could mean that through continuous observation that
person has displayed behaviour in a number of different contexts that is goal-
directed and persistent. Conversely, it could be that in a given situation a
person exhibits high persistence and puts in a lot of effort to their behaviour to
achieve their goals and outcomes in that situation. The study of motivation in
sport contexts must therefore begin with a clear definition of the construct.
The assumptions and thought processes that underpin the previous example
reveal some interesting properties of motivation and the way information in
the social context such as a behavioural observation reveals these properties
in others. First, motivation is a social psychological construct, therefore it
operates in social contexts, determines behaviour, and, in the social cognitive
tradition, is determined by the processing of contextual and intrapersonal
information regarding behavioural contingencies and outcomes. Second,
motivation can be global or contextual, generalized or specific, and stable or
unstable – motivation can therefore be trait-like or state-like in its character-
istics depending on the situation and context. Third, motivation is influenced
by a number of internal (e.g. needs, drives, effort, goals) and external (e.g.
relationships, rewards, social influence) sources of information. Formal def-
initions of motivation will clarify and expand on the above properties. These
definitions are a prologue to studying the main theoretical explanations of
motivation in social psychology in sport contexts covered in this chapter.
100 Social psychology and motivation
Defining motivation and motivational theories
At the most basic level, motivation is defined as an internal state that activates,
energizes, or drives action or behaviour and determines its intensity and direc-
tion. In addition, motivation is also thought to encompass the arousal and
persistence of action or behaviour. This definition tends to account for both
biologically based motives that serve to satiate physiological needs such as
hunger, thirst, and sex, as well as psychological or higher-order needs such
as fulfilment and self-actualization. Biological and emotion-based theories
view motivation as arising from an organism’s need to satisfy biological
requirements in a drive reduction hypothesis, derived from behaviourist prin-
ciples of arousal and reinforcement. Approaches that have focused on so-
called higher needs tend to view humans as having innate psychological
requirements such as the need for achievement or to demonstrate competence
(White 1959; Deci and Ryan 1985). The latter are derived from humanistic
approaches to motivation such as those put forward by Maslow (1943) in
which biological need satisfaction precede the satisfaction of higher-order
needs towards a complete satisfaction of a person’s needs repertoire in a
hierarchical fashion.
In social psychology, recent approaches to the study of motivation adopt a
social cognitive perspective, stressing informational processing, volition, and
explicit motivational constructs as the important processes that give rise to
goal-directed behaviour. Theories of intention (e.g. Meiland 1970; Ajzen and
Fishbein 1980) and self-efficacy theory (Bandura 1997) are influential social
cognitive theories to motivation (cf. Chapter 2). In such theories motivation
is viewed in terms of the pursuit of target behaviours to achieve a particular
goal. In sport and other achievement situations like academic and occu-
pational performance, goals are often essential to functioning and instru-
mental in fostering motivation. For example, attribution theory (Heider
1958) and achievement goal theory (Nicholls 1989) are two goal-oriented
social cognitive approaches to motivation and behaviour in sport. Such social
cognitive models of motivation are often criticized for not being explicit
enough in explaining the origins or formative antecedents of motivation.
Theories based on psychological needs such as self-determination theory (Deci
and Ryan 1985, 2000) are attractive in this regard as they provide the basis on
which motives toward specific behaviours arise. Need theories therefore
address the ‘why’ questions related to motivation. As Deci and Ryan state:
‘cognitive theories begin their analysis with a cognitive representation of some
future desired state (outcomes). What is missing, of course, is the consider-
ation of the conditions of the organism that makes these states desired’ (1985:
228). Therefore, need theories complement the social cognitive approaches
because they suggest that human behaviour in a multitude of contexts is
underpinned by psychological needs.
In summary, motivation is the driving force behind behaviour and is
characterized by its focus (selection and direction) and valence (intensity and
Social psychology and motivation 101
duration). Social psychological theories of motivation in sport have typically

adopted a social-cognitive perspective and assume that motivation is based on
the processing of social information. Some models have adopted a needs-based
approach in which psychological needs explain the origin or formation of the
social cognitive antecedents of situated action. The next sections will critically
evaluate the different social psychological theories of motivation applied
to the sport context, with a focus on social cognitive and needs-based
approaches.
Attribution theory
Sports participants and athletes are always striving to make sense of the causes
of good and bad performances. Sometimes they make excuses for not making
the right decisions or not living up to expectations. Other times they are bash-
ful or magnanimous in the face of success or failure, citing good luck rather
than good judgement as reasons for their accomplishment. Some blame them-
selves or others for the misfortune of losing. Others brashly state that they
knew that they would be successful and expect to be successful again. In all
these scenarios, athletes are making attributions for their successes and
failures. They are assigning perceived causes to the outcomes they have
experienced as a result of their performance or behaviour in sport. Attribution
theory is the study of the processes that underlie these attributions and how
they affect subsequent motivation and behaviour.
Weiner’s (Weiner et al. 1972) theory of attribution focused on understand-
ing the nature, causes, and consequences of attributions of success and failure
in achievement situations. He adopted Heider’s (1958) proposal that attribu-
tions could be internal or external and extended them to account for the range
of attributions possible in achievement situations. Weiner hypothesized that
people’s attributions for success or failure could be characterized by three
bipolar dimensions: locus of causality (internal or external), the stability of the
cause (stable or unstable), and whether the cause is controllable by the indi-
vidual making the attribution (controllable or uncontrollable). In terms of
mechanisms, it was theorized that after an achievement event, an individual
assesses whether they have succeeded or failed and they experience a general
positive or negative emotional response, respectively. The attribution is then
ascribed following the experience of more specific affective responses based on
specific performance outcomes (e.g. pride if success is attributed to ability, or
hopelessness if failure is attributed to luck). On the basis of these attributions,
expectations of future success regarding prospective behavioural engagement
are formed. This model resulted in the development of a three-dimensional
taxonomy of attributions in achievement situations, illustrated in Figure 5.1.
Weiner initially focused on the locus of causality and stability dimensions
and illustrated how attributions of success or failure could be internal
and stable (ability), internal and unstable (effort), external and stable (task
Figure 5.1 Weiner’s (1972) three dimensions of attribution
difficulty), or external and unstable (luck). It was hypothesized that most

attributions would be characterized by this 2 × 2 conceptualization (see front
face of Figure 5.1). The model was viewed as dynamic and changing such that
attributions and expectancies of success varied according to the available
information regarding performance. It was also expected that attributions
would influence motivation and persistence (approach) or desistence (avoid-
ance) with respect to future participation in the behaviour, depending on the
type of attribution generated by past experience. An adaptive attribution pat-
tern likely to contribute to heightened motivation and persistence would be to
attribute success to internal factors (e.g. ability or effort) and failure to external
factors (e.g. task difficulty or luck). For example, if an athlete is successful in
achieving his or her goal when performing a sports skill, he or she would feel
confident and efficacious towards that task if they attributed it to their own
skill or effort, and this would result in a positive expectation of future success.
Analogously, the athlete’s confidence would not be undermined if they
attributed an unsuccessful performance to external factors. However, this
attribution profile may be problematic if the attribution of success to internal
and stable (i.e. ability) causes is followed by a failure on the next attempt. This
may lead to an attribution of the failure to ability as well, which will result in a
maladaptive attributional profile. Beliefs about control over future task per-
formance are important in this regard. Ability tends to be relatively fixed and
uncontrollable, and therefore an attribution of success to a more controllable
form of cause (i.e. effort) may result in a more adaptive motivational pattern in
subsequent performance.
In sport, a large body of research has examined the efficacy of attribution
theory to explain the attributions that athletes make regarding their success
and failure, and also whether those attributions have implications for future
motivation (Biddle et al. 2001; McAuley and Blissmer 2002). Early research in
attribution theory examined the prevalence and nature of attributions accord-
ing to Weiner’s (Weiner et al. 1972) conceptualization. These studies demon-
strated that the attributions did not fall into the neat taxonomy offered by
Weiner and found that certain attributions were interpreted differently under
success and failure. For example, Iso-Ahola (1977) found that effort was related
to ability in a win situation but to luck and task difficulty in a lose situation in
young baseball players. This suggested a flexibility in the children’s attribu-
tions such that effort was treated as an internal attribution in successful
circumstances, and as external in conditions of failure. Further, different types
of attribution were found in sports participants, for example, lack of practice
or training was an important external attribution under failure conditions
(Iso-Ahola 1977). Similarly, Roberts and Pascuzzi (1979) found that the
attributions made by undergraduate sports students regarding sport situations
could not be adequately categorized as ability, effort, luck, and task difficulty.
However, the attributions could be adequately categorized into Weiner’s 2 × 2
(locus of causality × stability) contingency of attributions. These findings sug-
gest that while Weiner’s dimensions were applicable to attributions in
sport, the rigid classification of the type of attributions, i.e. ability, effort, task
difficulty, and luck did not hold.
Studies have also investigated whether a self-serving bias exists when indi-
viduals make attributions for their success and failure in sport contexts. Self-
serving biases are observed in many areas in social psychology, and reflect an
individual actively trying to maintain their self-esteem or a positive and con-
sistent view of themselves. In attribution research, individuals have been
shown to generally attribute success to internal causes and failure to external
causes (Weiner et al. 1972). This self-serving bias was supported by a meta-
analysis of 91 studies conducted by Mullen and Riordan (1988). They found
moderate effect sizes for a general internal/external and ability dimensions
and low effect sizes for effort, luck, and task difficulty. This seemed to point to
a deliberate bias towards internal attributions, however, group size also
affected the size of the bias leading the authors to conclude that the attribu-
tion profile was more a function of information processing than intentional
bias towards the self. Similarly, research has suggested that athletes may have
dispositional tendencies to report certain types of attributions, known as
attributional style, although little evidence of a clear pattern in attributional

style is evident (Biddle et al. 2001). Taken together, this evidence does not
unequivocally support a self-serving bias in attributions in sport.
In terms of outcomes, attributions have been shown to be related to motiv-
ational and emotional outcomes, as predicted by Weiner’s theory. Weiner
anticipated that individual expectancy of success in subsequent behavioural
engagement would be influenced by an attribution to stable causes. This pro-
vides greater certainty in the expectation that future outcomes could be
repeated in future, an unstable attribution would suggest that future expected
outcomes were uncertain. Research examining attributions in high-school
athletes indicated that positive expectancies of success occurred more fre-
quently after success than failure and due to stable attributions (Singer and
McCaughan 1978). Further, Rudisill (1989) showed that future expectancies of
success and adherence to a balancing task with a stabliometer were predicted
by controllable and internal attributions and with high competence, but,
against predictions revealed that unstable attributions were more influential
than stable ones. Finally, Grove and Pargman (1986) found that effort rather
than ability attributions were associated with positive future expectancies of
success, and, given that effort is not a stable construct, this indicates that
unstable but controllable attributions are implicated in future expectancies of
success. In addition, internal attributional profiles have been associated with
positive emotional profiles in sport (Biddle et al. 2001). Research has indicated
that attributions tend to account for significant variance in emotional
outcomes after controlling for appraisals in models examining cognitive-
motivational-relational models of emotion (Vlachopoulos et al. 1997). These
findings suggest that encouraging athletes to make controllable and internal
attributions for success are likely to give rise to positive future expectancies of
behaviour.
Biddle outlines the problems with attribution theory in sport and exercise.
He notes a decline in the productivity of researchers in this area suggesting
that interest in attribution theory as an explanation of motivation in sport
may be waning. He suggests that a possible reason for this is that attribution
research has focused on too narrow a set of constructs and further research
incorporating dispositional constructs from other motivational theories is
necessary for research in the area to progress. Specifically, he stresses that the
incorporation of constructs relating to perceived control, competence, and
achievement goals (see section on achievement goals in sport, this chapter) in
models of sport motivation will answer important attribution-related ques-
tions an athlete might ask themselves in sport, such as ‘Why did I fail at this
task?’ (Biddle et al. 2001).
Social cognitive theory and self-efficacy
Self-efficacy is an important social psychological construct derived from

Bandura’s (1977, 1997) influential social cognitive theory. Self-efficacy has
established itself as a key variable in many social psychological theories and
models of motivation and in a number of applied contexts such as edu-
cational, health, occupational, and sport settings. Bandura’s work has been
extremely influential and enduring. To appreciate the measure of the import-
ance of Bandura’s theories to the study of motivation, one need only consider
his place in the psychology canon. For example, Hagbloom et al.’s (2002)
research on the citation records of psychologists over the past century ranks
Bandura fifth on the list of the most frequently cited psychologists in the
professional literature and on the American Psychological Association’s
Members’ list of the most eminent. Further, nearly 30 years of research in sport
psychology has established self-efficacy as one of the most potent and consist-
ent constructs relevant to performance and motivation. Literally hundreds of
sport-related studies have included self-efficacy as an antecedent, outcome, or
process (mediator or moderator) variable. The main reason for its longevity is
that it is a powerful predictor of behaviour and unequivocally implicates the
role of perceived rather than actual ability in sport behaviour and perform-
ance. This section will define self-efficacy, outline the important tenets of
social cognitive theory, and examine the important effects and processes
related to self-efficacy and sport performance.
Definition of self-efficacy
Self-efficacy is an individual’s personal estimate of confidence in his or her
capability to accomplish a certain level of performance. According to Bandura,
‘Perceived self-efficacy is defined as people’s beliefs about their capabilities to
produce designated levels of performance that exercise influence over events
that affect their lives’ (1994: 71). Unpacking this rather dense definition, self-
efficacy is a set of beliefs held by an individual regarding his or her capacity or
ability regarding their performance of an upcoming behaviour or action. It
also comprises beliefs related to their ability that doing the behaviour will
result in salient outcomes. Self-efficacy beliefs are a function of a number of
different experiential and situational factors as determined by social cognitive
theory. An important property of self-efficacy is that it is situation-specific and
reflects judgements or expectations in a given context directed towards per-
sonally-defined outcomes. It has therefore been defined by some authors as
‘situation-specific self-confidence’ (Feltz and Chase 1998: 60). Indeed, as with
many other theories that adopt a correspondence rule or boundary condition
(e.g. Ajzen 1985), self-efficacy generally has most potency in predicting out-
comes such as performance and behaviour when it is specific to that
behaviour. In addition, due to its situated specificity, self-efficacy is also modi-
fiable. As will be seen later, an individual’s level of self-efficacy is dependent on
a number of interpersonal and external variables that serve as sources of

information for the formation of self-efficacy judgements. Self-efficacy can
therefore be considered a state-like construct that can be manipulated and
changed.
Self-efficacy has been shown to have a profound influence on self-
regulation. It affects cognition (e.g. interpretation of information), motivation
(e.g. intentions to perform a behaviour), affect (e.g. satisfaction), and
behaviour (e.g. performing a motor task) (Biddle et al. 2001). Self-efficacy is
implicated in human achievement motivation and subjective well-being pro-
cesses (Weiner et al. 1972). A person reporting high self-efficacy towards a task
in a given situation is able to have influence over their environment, persist in
the face of difficulty and failure, put a great deal of effort into the task to attain
desirable goals, solve problems more effectively, be more interested in the task
and immerse themselves in it more deeply, have a conceptualization of ability
as changeable, attain satisfaction from their endeavours toward the task, and
have an approach rather than avoidant orientation towards the task. Given
the pervasive effects of high levels of self-efficacy it is easy to see the import-
ance of the construct and why it has a role in a number of theories
on motivation, in both exercise (see Chapter 2) and sport (see sections on
self-determination theory and achievement goal theory, this chapter).
Social cognitive theory

Social cognitive theory was developed as an offshoot of Bandura’s (1977,
1997) social learning theory and the effects of observational learning in the
treatment of anxiety, phobias, and other behaviours. Social learning theory
examines the effect of vicarious experience and familiarization with a task on
subsequent approach motivation and confidence towards that behaviour.
Bandura was a profound believer in the powerful effects that models, or people
demonstrating a task to be learned, had on behaviour and supported this
through behaviourist theories of reinforcement but also through obser-
vational learning. The importance of observational learning and vicarious
experience has been supported in numerous experiments (see Bandura
1977). However, it was the presence of situation-specific self-confidence or
‘self-efficacy’ that was proposed to have the most pervasive effect in these
experiments, such as snake phobics overcoming their phobias.
In social cognitive theory, self-efficacy was deemed to arise from several
sources of information about a person’s confidence or subjective beliefs in
their skills or ability to affect outcomes by performing a given behaviour or
action. These sources of confidence information are previous performance
accomplishments, vicarious experience, verbal persuasion, and physiological
states. The relations between self-efficacy beliefs and these sources of informa-
tion are outlined in Figure 5.2. There are two broad categories of information
sources, previous experience and current influences. Performance accomplish-
ment reflects previous successful personal experiences with the action or the
behaviour. Importantly, it is not necessarily past experience per se but the

mastery of the performance that is important in influencing self-efficacy
beliefs. Performance accomplishments are viewed as the single most import-
ant influence on self-efficacy beliefs in social cognitive theory (Bandura 1977).
As another source of confidence information, vicarious experiences have strong,
albeit less potent, effects on self-efficacy beliefs. It is assumed in social cogni-
tive theory that the correct execution of a behaviour is reinforced when an
individual observes another performing the behaviour. Observational learning
is assumed to be a potent source of confidence. For example, an intermediate-
level athlete may watch a skilled professional performing a sports skill and
their confidence may be positively enhanced because their own execution of
the skill is reinforced by the act of observation. The features of the model and
the matching of the characteristics of the model to the observer are import-
ant contributory factors in the effectiveness of vicarious experiences in
influencing efficacy beliefs.
Self-efficacy beliefs are also affected by sources of information from the
current behavioural context. Verbal persuasion represents task-related informa-
tion available from coaches and other team members or personal. This can be
in form of positive feedback and technical commentary or person-centred
Figure 5.2 Relationships between information sources, self-efficacy, and

psychological and behavioural outcomes in Bandura’s (1977) social cognitive
theory
information such as the use of imagery or self-talk. These sources of informa-

tion reinforce expectancies of success. Finally, physiological states such as
autonomic arousal, negative emotional states associated with threat in
achievement contexts such as anxiety and worry, or positive emotional states
associated with approach motivation such as state self-confidence also affect
self-efficacy beliefs. These can also refer to other physiological states such as
fatigue and fitness level. In sport, these physiological states are likely to have a
profound effect on self-efficacy beliefs because of the elevated physiological
arousal associated with the stress of exercise and the inherent involvement
of emotional states in sport performance, such as competitive state anxiety
(see Chapter 6). The effects of persuasion, vicarious learning, and physio-
logical states on self-efficacy are hypothesized to be less salient than perform-
ance accomplishments. Together these sources of information contribute
synergistically to efficacy expectations.
Just as self-efficacy has several antecedent variables, it also acts as predictor
of a number of salient psychological and behavioural outcomes (Figure 5.2). In
terms of psychological outcomes, self-efficacy is deemed to act on social cogni-
tive thought processes and affective outcomes such as the choice of goals set,
psychological well-being or worry, and the attribution of behaviour to a given
cause. The higher an individual’s self-efficacy towards a given behavioural
pattern, the more likely they are to set challenging goals, cite positive well-
being, and attribute causality to internal, stable sources (Bandura 1977). In
addition to psychological outcomes, self-efficacy is also hypothesized to influ-
ence a person’s choice, effort, and persistence towards a given behaviour.
Choice is likely to reflect whether a person decides to approach or avoid a
behaviour, in other words, the direction of the behaviour, while effort and
persistence indicate the valence and prolonged nature of their behavioural
application.
Self-efficacy and outcome expectations

Bandura (1977, 1997) also proposed that people not only hold beliefs about
their ability to perform a given behaviour, but also that the behaviour will
result in certain outcomes. Feltz and Chase make the distinction between the
two constructs in terms of the nature of their beliefs: ‘In essence, outcome
expectations are concerned with beliefs about one’s environment, and efficacy
expectations are concerned with beliefs about one’s competence’ (1998: 66).
Bandura (1977, 1997: 511) argues that both sets of beliefs interact in determin-
ing a behavioural response, so that high levels of self-efficacy towards
engaging in a particular task or action to produce a desirable outcome will
predict behaviour when the person also has high expectations that the
behaviour will result in desired outcomes. These action-outcome links are
therefore implicated in a synergistic process that results in approach
behaviours, effort, and persistence. Pragmatically, in achievement contexts, it
is often assumed that an individual holds positive outcome expectations with
regard to propensity of the behaviour to produce outcomes. Further, it is also

assumed that the outcome is deemed to be salient to the individual. Further,
self-efficacy is generally assumed to have a more pervasive effect on behaviour
than outcome expectations (Feltz and Chase 1998). However, when it comes
to goal setting, it is important to consider the role of outcome expectancies
because the outcome or goal needs to have personal value to the individual.
Self-efficacy in sport
Self-efficacy and social cognitive theory have been the subject of intensive and
prolonged research in a number of applied social psychology areas, including
educational psychology, health psychology, occupational psychology, and
sport and exercise psychology. Meta-analyses of research with self-efficacy
have indicated a significant relationship of moderate effect size between self-
efficacy and behaviour such as work-related performance (e.g. Stajkovic and
Luthans 1998). To date, there is no meta-analysis of the effects of self-efficacy
on the salient outcome variables from Bandura’s social cognitive theory in
sport, which is surprising considering the number of studies that have been
conducted in the area. There are, however, a number of narrative reviews of
the effect of self-efficacy in sport that have supported the important tenets of
social cognitive theory (e.g. Feltz and Chase 1998; McAuley and Blissmer
2002). These reviews unanimously acknowledge the pervasive effect of self-
efficacy beliefs on sport performance. Feltz and Chase, for example, cited 11
studies with 29 tests of the explicit relationship between level of self-efficacy
and overt performance and revealed that 25 of these tests were significant.
Reviews have also noted trends in relationships between self-efficacy and other
salient relationships from social cognitive theory, such as the large effect that
performance accomplishment has on self-efficacy above the other antecedent
variables and the effect of self-efficacy on intention and motivation (Feltz and
Chase 1998). The consensus from these reviews is that self-efficacy is an
important and influential construct in sport psychology and motivation, mir-
roring the support for self-efficacy in meta-analyses in other domains. Future
researchers in the field should seek to conduct the first meta-analysis of self-
efficacy in the sport domain to quantify the relationships identified in the
narrative reviews.
Self-efficacy and sport performance

There is a large body of experimental and cross-sectional research that sup-
ports the effect of self-efficacy on sport performance. In experimental studies,
the classic technique to manipulate self-efficacy beliefs involves using mislead-
ing or ‘bogus’ feedback about performance in a pre-trial task. In these experi-
ments, high and low self-efficacy groups of participants are created by having
participants perform the desired task in an ostensible familiarization test
before the ‘main’ trial. The performer is then given feedback that they have
performed better or worse than expected compared to normative results or a

confederate (e.g. Weinberg et al. 1980). Self-reported self-efficacy beliefs and
subsequent performance of the task in a main trial then form the core
independent and dependent variables of the experiment respectively. Using
this paradigm, studies have shown that high self-efficacy participants per-
formed better than low self-efficacy performers (Weinberg et al. 1980), and did
so under conditions of failure alone (Weinberg et al. 1980) and under failure
when a cognitive strategy to enhance self-efficacy was used (Weinberg 1986).
The experimental manipulation of self-efficacy has also been shown to result
in superior sport performance compared to other manipulations designed
to enhance intrinsic motivation and attentional focus (Lohasz and Leith
1997). These studies support hypotheses relating to self-efficacy and sport
performance in social cognitive theory.
Several studies have adopted cross-sectional and longitudinal designs to
examine the effects of self-efficacy on performance. Self-efficacy has been
shown to be strong predictor of objectively measured performance in basket-
ball free throw shooting performance (Kavussanu et al. 1998), a tennis service
task (Theodorakis 1995), gymnastics performance (Weiss et al. 1989), and pla-
cing in distance running races (Martin and Gill 1995). Studies have also exam-
ined the effect of competition on the self-efficacy-performance relationship.
For example, Kane et al. (1996) found that self-efficacy was related to perform-
ance in high-school wrestlers, especially in competitive conditions. Similarly,
studies have shown that self-efficacy predicted performance in high-avoidance
tasks such as difficult routines in platform diving (for review, see Feltz and
Chase 1998). This is consistent with Bandura’s (1977) hypothesis that situ-
ational self-confidence towards performing a task that is feared is highly
predictive of engagement. Overall, these studies support the self-efficacy–sport
performance relationship in keeping with social cognitive theory, the
narrative reviews of self-efficacy in sport, and the meta-analytic reviews of
self-efficacy in other domains.
Outcome expectancies
Another salient variable in social cognitive theory is outcome expectancies.
Bandura (1977, 1997) viewed these as beliefs regarding the potency of the
behaviour rather than beliefs about personal skill and ability. It was proposed
that both sets of beliefs will affect behavioural and psychological outcomes,
but in achievement situations, self-efficacy will have a stronger influence on
sport performance and persistence. Lee provided some evidence to support the
higher validity of personal evaluations of outcome compared with objective
evaluations based on previous performance in gymnasts. It was found that
outcome expectancies were a more accurate predictor of gymnasts’ scores than
previous scores, supporting the importance of subjective judgements regard-
ing the behaviour. Interestingly, Martin and Gill (1991) found a stronger influ-
ence of what they called outcome self-efficacy on the performance of a fine
motor task compared with performance self-efficacy. However, the definition

of outcome self-efficacy was not in terms of outcome expectancies, but rather a
different focus of the self-efficacy variable. Other studies have indicated that
outcome expectancies moderate the influence of self-efficacy beliefs on
behaviour and outcome, but only in the exercise domain and not in sport
(Williams and Bond 2002), and this seems to be an area where research is
lacking.
Antecedents of self-efficacy
Research in sport and exercise has examined the effect of each of the four key
sources of information: performance accomplishments, vicarious experience,
verbal persuasion, and physiological states (see Figure 5.2) on self-efficacy
beliefs. Previous experience, although a slightly crude proxy for performance
accomplishment, has been shown to predict self-efficacy in a number of stud-
ies (e.g. George 1994; Feltz and Chase 1998). In terms of vicarious experience,
studies have indicated that models have a pervasive effect on self-efficacy
(Gould and Weiss 1981; McAuley 1985). The effects of self-modelling on self-
efficacy have been inconclusive, although Starek and McCullagh (1999) found
no differences in novice swimmers’ self-efficacy levels when either self or other
models were used. Carnahan, Shea, and Davis (1990) demonstrated that parti-
cipants completing a bench-press exercise reported higher self-efficacy when
their ‘spotter’ provided visual and verbal cues compared with no cues, indicat-
ing that verbal persuasion can affect self-efficacy. Finally, physiological states
such as cognitive and somatic anxiety have been implicated in predicting self-
efficacy (e.g. Martin and Gill 1991), although some studies have examined
anxiety from a trait perspective (e.g. George 1994; Martin and Gill 1995)
which does influence self-efficacy but is not strictly a physiological state but a
tendency to be in that state in threatening situations.
A limited number of studies have investigated the impact of multiple sources
of information on self-efficacy simultaneously. Gould and Weiss (1981) exam-
ined the effects of similar and dissimilar models that gave varying levels of
verbal persuasion (they called it ‘self-efficacy talk’) on participants’ perform-
ance of a muscular endurance task. Model similarity had the strongest effect
on performance, with participants given a model similar to themselves per-
forming better on the motor task than participants given a dissimilar model.
Participants receiving positive ‘self-efficacy talk’ and no-talk conditions per-
formed better than those receiving negative talk and ‘irrelevant’ talk models.
Wise and Trunnell (2001) studied the effects of three types of information on
bench-press exercise performance: performance accomplishment, use of a
model, and verbal message. Results demonstrated that performance
accomplishment resulted in the highest levels of self-efficacy, followed by
modelling and use of a verbal message. This is supported by Feltz and Chase
(1998) who report that previous performance is a better predictor of self-
efficacy than autonomic (physiological) perception. Previous performance
accomplishment seems to be a major source of information from which

athletes draw their self-efficacy beliefs.
Self-efficacy and outcomes

Self-efficacy is also hypothesized to influence psychological and behavioural
outcomes, according to social cognitive theory (see Figure 5.2). Other than
performance of the task, these key outcomes are grouped into psychological
processes, e.g. goals, affective states, and attributions, and behavioural out-
comes, e.g. task choice, effort, and persistence. In terms of psychological
processes, Duncan and McAuley (1987) found that self-efficacy did affect the
pattern of post-performance attribution of outcome in basketball players
performing a free-throw task, but Gernigon and Delloye (2003) found that self-
efficacy mediated the influence of performance-related feedback on causal
attributions and some research has suggested that attributions may be a source
of information for self-efficacy (Bond et al. 2001). Turning now to affective
outcomes, Hall and Kerr (1998) and George (1994) found that high levels
of self-efficacy coincided with low levels of cognitive and somatic anxiety
in competitive fencers and baseball players respectively. With respect to
behavioural outcomes, a vast number of studies have supported the prediction
of sport behaviour in terms of task performance, but few have examined effort.
George (1994) showed that self-efficacy beliefs were associated with effort
across a six-game period in a longitudinal study of baseball games across a
season. These results lend some support to the proposed effects of self-efficacy
on outcomes.
Self-efficacy, goal setting, and imagery

In sport, goal setting is often mooted as an important means to enhance the
motivation of athletes. One of the mechanisms by which goal setting might
affect performance, adherence, and persistence in sport is through enhanced
self-efficacy. This is because goal setting is likely to provide salient information
about the task, particularly when such a task has been performed successfully
in the past, is valued by the performer, and is likely to demonstrate com-
petence. Furthermore, goals must be of the optimal level of difficulty to permit
the performer to be confident in their ability to achieve it, but not so difficult
as to be unreachable or so easy as to be unchallenging. For example, Miller and
McAuley (1987) showed that basketballers who received training to set
appropriate goals and targets had a stronger relationship between self-efficacy
and basketball free-throw performance compared with those that did not
receive goal training. In addition, studies have demonstrated that goal setting
mediates the relationship between self-efficacy and performance (Theodorakis
1995). These results suggest that variations in performance due to self-efficacy
are accounted for by changes in personal goal setting. The role of goal setting
may determine the self-efficacy–performance relationship (a moderation
effect) but it may also explain why self-efficacy influences performance

(a mediation effect).
Given the effectiveness of modelling on self-efficacy, sports psychologists
have adopted interventions to replicate the modelling effect through the use
of imagery and examine whether it alters self-efficacy levels. Many of these
interventions use multiple strategies so it is difficult to disentangle the effects
of imagery from other elements of the intervention. Studies using these com-
posite intervention techniques have shown significant increases in self-
efficacy levels, but they have used other techniques alongside imagery such as
relaxation and coping skills (e.g. Meyers et al. 1982). However, other than
training and conditioning drills, Gould et al. (1990) found that athletes rated
imagery among the most effective strategies to enhance self-efficacy. Several
intervention studies examining imagery alone have indicated that it positively
affects self-efficacy levels in rock climbers (Jones et al. 2002) and participants
performing a laboratory muscular endurance task (Feltz and Chase 1998).
Recently, Short et al. (2002) have examined the effect of imagery type and
direction (facilitative or debilitative) of imagery on self-efficacy and perform-
ance of a golf-putting task. Results indicated that imagery that had a motiv-
ational rather than cognitive function and was facilitative in direction
enhanced self-efficacy, suggesting that interventions should consider both
the function and interpretation of the imagery when attempting to enhance
self-efficacy through these means.
Achievement goal theory
Achievement goal theory is a prominent approach that aims to explain motiv-

ation and behaviour in sports settings (Duda 2001). Theories of achievement
goals have been put forward by a number of theorists and have common
elements and some subtle differences (Nicholls 1989; Ames 1992; Dweck
1992). Common to these theories is the hypothesis that individuals focus on
demonstrating competence in achievement situations. Thus, the way an indi-
vidual views their ability or success in an achievement context will result in
their persistence or desistance in competence-demonstrating behaviours.
These theories suggest that two distinct motivational tendencies or achieve-
ment goal perspectives prevail in achievement situations and influence the
way competence information is interpreted: (1) a task or mastery orientation;
and (2) an ego or performance orientation. The theories propose that these goal
orientations define the way people tend to view their ability in a given con-
text. If that person’s view tends to be more personal or self-referent, then he or
she will be motivated to persist, even under failure, while if his or her view
leans towards being normative or other-referenced, he or she may persist but
generally only under conditions of success.
In terms of motivation in competitive situations, an individual who adopts a
predominantly task-oriented approach, i.e. a high-level of task orientation, is
likely to view success as demonstrating competence by achieving task-related

successes. For example, a soccer player may view their success in terms of the
number of passes they complete relative to their own personal target or a
basketball player may view success in the number of rebounds they take rela-
tive to their previous match. An individual who adopts a predominantly ego
orientation will view success in competitive situations by demonstrating com-
petence through beating others, winning, and achieving a high rank com-
pared to others. Thus, finishing first in a 100m race, topping the leaderboard in
a golf tournament, or being on the winning side in a hockey match are all
examples of ego-oriented demonstrations of competence. It is important to
note, of course, that although task and ego orientations are considered
orthogonal (independent) constructs, it is possible to hold high levels of
both task and ego orientations simultaneously so that competence can be
demonstrated from both a self-referenced and other-referenced perspective.
However, it is under conditions of failure or losing where the theory makes
its most influential predictions with respect to motivation. An individual in an
achievement situation who has a predominantly ego orientation and has a low
task orientation is likely to view success only in terms of demonstrating com-
petence through winning. In this case, the individual will be unable to dem-
onstrate competence if they lose or experience failure which is likely to
undermine motivation to engage in that behaviour again. If, however, they are
task-oriented as well (or task-oriented alone, with low ego orientation), then
they will be able to demonstrate competence through the achievement of task-
related goals that are personally relevant and have meaning to themselves,
regardless of success or failure. They may have lost the match, for example, but
at least they played well and were able to demonstrate competence though
their own goals independent of the win–lose outcome. In this respect, a task
orientation alone or in conjunction with an ego orientation is motivationally
adaptive in terms of persistence and effort on subsequent occasions (Goudas et
al. 1994). Importantly, it can be seen that goal orientations may determine a
number of psychological processes that lead to persistence in motivation. For
example, task orientation reflects the demonstration of personally relevant
goals internal to the individual and is likely to predict the type of motivation
associated with persistence and satisfaction, namely intrinsic motivation (But-
ler 1987; Ryan and Deci 1989). Further, a person with a high-task orientation is
likely to make an attribution of success to high ability and internal reasons
regardless of objective outcome. Thus, achievement goal theory may influence
motivational constructs from self-determination theory and attribution
theory respectively.
Research adopting achievement goal theories in sport have been dominated
by Nicholls’ (1989) approach to achievement goals, mainly because early
research and development of valid instrumentation have focused on this ver-
sion of the theory. Two instruments have been developed for the measure-
ment of achievement goal orientations in sport settings. The task and ego
orientation in sport questionnaire (TEOSQ, Duda and Nicholls 1992) and the
perceptions of success in questionnaire (POSQ, Roberts et al. 1998) contain

items measuring task orientation (e.g. ‘I feel successful in sport when I learn
new skills’) and ego orientation (e.g. ‘I feel most successful in sport when
others can’t do as well as me’). Exploratory and confirmatory factor analytic
studies have supported the factor structure, internal consistency, and test–
retest reliability of these instruments (e.g. Duda and Nicholls 1992; Roberts et
al. 1998). Studies have also confirmed the criterion validity of these instru-
ments with high and significant correlations between the two traits across the
different methods (Fonseca and Balague 1996), although no multi-trait, multi-
method study has been conducted to compare these instruments. Together
these instruments have been adopted for almost all of the research using
achievement goal theory in sport contexts.
To date there is no meta-analysis that has tested the hypothesized relation-
ships between achievement goal orientations and motivational constructs.
However, there are a number of good narrative reviews of the area (e.g. Biddle
1999; Ntoumanis and Biddle 1999; Duda 2001). Research has indicated that
achievement goals predict a number of motivation-related constructs such
as beliefs about causes of success (Duda and Nicholls 1992), self-motivation
(Biddle et al. 1996), and effort (Williams and Gill 1995), sources of competence
information (Williams 1994), competitive state anxiety (Newton and Duda
1995), enjoyment and satisfaction (Allen 2003), and motivation-related
behaviours (Boyd et al. 2002). Overall, the general pattern of results suggests
that a task-orientation tends to be more strongly related to adaptive motiv-
ational constructs such as self-confidence from the CSAI-2 (Newton and Duda
1995), self-referenced sources of competence information like goal attainment
(Williams 1994), experiences of learning and improvement (Williams 1994)
and incremental beliefs about ability (Sarrazin et al. 1996), personally control-
lable attributions of success/failure (Vlachopolous and Biddle 1997), intrinsic
motivation (Kavussanu and Roberts 1996), perceived control (Pensgaard
1999), self-esteem (Boyd et al. 2002), effort (Williams and Gill 1995), and
enjoyment (Allen 2003). An ego orientation has been shown to be signifi-
cantly and positively related to other-referenced and motivationally maladap-
tive outcomes such as cognitive and somatic anxiety (Newton and Duda
1995), indices of moral functioning in sport such as moral judgement, inten-
tions, and behaviour (Kavussanu and Roberts 2001), and personally
uncontrollable attributions of success/failure (Vlachopolous and Biddle 1997).
Goal involvement
Much of the research on achievement goal theory in sport and exercise has
focused on the effects of the dispositional and trait-like task and ego orienta-
tions on motivation-related constructs and behaviour. This research has
tended to emphasize the importance of a task orientation, alone or in conjunc-
tion with ego orientation, to foster motivationally-adaptive psychological
profiles in athletes. However, recent research has questioned the heavy focus
on task orientation as the individual difference that makes motivation in sport,

particularly competitive sport, adaptive (Hodge and Petlichkoff 2000).
Researchers are recognizing the need to account for the effects of situational
factors such as motivational climate (Seifriz et al. 1992) and involved
achievement goal states of an individual in achievement contexts (Harwood
2002). A distinction between situational, involved achievement goal perspec-
tives and the more traditional dispositional goal orientation has been pro-
posed. Consequently, there has been a shift towards a profile approach to goal
orientations and a more individual, ideographic basis (Harwood 2002). Har-
wood has criticized the focus of the majority of research adopting achievement
goal theory on nomothetic methods and individual difference data on
achievement goals from groups of athletes. While it is recognized that this
has offered some useful recommendations in terms of the coaching styles
and motivational climates afforded by coaches for their athletes, such
approaches are limited because much of the data is cross-sectional, which
limits causality inferences, but, more importantly, it neglects the import-
ance of goal involvement at the situational level which are state-like and
changeable over time.
Harwood supports this argument with data gained from involved measures
of task and ego orientation compared with dispositional data. For example,
Harwood used involved competition-specific measures of achievement goal
orientations to show that involved goal state profiles were markedly different
to goal orientations at the dispositional level as measured by the TEOSQ.
Athletes tended to report higher levels of involved ego orientation in competi-
tion than their levels on dispositional ego orientation towards their sport in
general. Results indicate the importance of over-reliance on dispositional
measures and the importance of taking into account individualized levels of
psychological constructs. Together these findings suggest that interventions to
change motivational orientation should target goal involvement at a con-
textual or even situational level. It may be that goal orientations in sport are
arranged hierarchically in keeping with other social cognitive theories of
motivation (cf. Vallerand 1997).
Motivational climate
One of the key contributions made by Ames’ (1992) research on achievement
goal theory is the role that situational factors, particularly the prevailing goal
structure operating in the achievement context, have on the situational goal
states observed in a given achievement situation. Ames (1992) and Nicholls
(1989) hypothesized that the motivational orientation experienced by an
individual performing a task in an achievement context was a function of their
dispositional goal orientation and the situational goal structure or motivational
climate operating in the context. Motivational climate reflects how com-
petence is typically evaluated with respect to tasks in a given environment and
is viewed as a function of the goals to be achieved, the role of competition or
relationships between individuals in that context, and the reward structure in

that environment. Ames created a mastery-oriented motivational climate by
presenting tasks so that effort was rewarded and the primary criterion for suc-
cess and found elevated levels of effort and persistence among school children
operating in such a climate.
In sport, motivational climate has been investigated alongside dispositional
goal orientations to examine the effects of personal dispositions and situ-
ational factors on motivational constructs. Seifriz, Duda, and Chi (1992) exam-
ined the effects of motivational climate and dispositional achievement goal
orientations on motivational constructs from the intrinsic motivation inven-
tory and beliefs about success. They developed an inventory, the perceived
motivational climate in sport questionnaire (PMCSQ, Seifriz et al. 1992) based
on items from Ames’s (1992) achievement goals questionnaire, and adminis-
tered it with the TEOSQ to high-school male basketball players. Results
showed that motivational climate positively predicted enjoyment and nega-
tively predicted tension in basketball players, but beliefs about success, com-
petence, and effort were accounted for by dispositional goal orientations
alone.
Subsequent research has shown that a mastery-oriented motivational cli-
mate predicts adaptive motivational and outcome variables in athletes such as
enjoyment and satisfaction (Boyd et al. 1995), attribution of success to effort
(Treasure and Roberts 1998), self-referenced sources of competence informa-
tion such as goal attainment (Halliburton and Weiss 2002), intrinsic motiv-
ation (Kavussanu and Roberts 1996; Petherick and Weigand 2002), perceived
competence (Sarrazin et al. 2002), self-efficacy (Kavussanu and Roberts 1996),
and problem-focused coping (Ntoumanis et al. 1999). Analogously, an ego-
oriented motivational climate tends to influence negative outcomes and be
related to maladaptive achievement pattern such as worry and stress (Pens-
gaard and Roberts 1995), attribution of success to ability (Treasure and Roberts
1998), other-referenced sources of competence information such as peer com-
parison and competition performance (Treasure and Roberts 1998; Halliburton
and Weiss 2002), extrinsic motivation (Petherick and Weigand 2002), and
emotion-focused coping (Ntoumanis et al. 1999). In terms of mechanisms,
Newton and Duda (1999) found a significant interaction between mastery
climate and a task orientation on satisfaction and effort. These findings indi-
cated that a situational goal structure that supports a task orientation results in
individuals reporting that they put in a great deal of effort and attribute their
success to effort.
Self-determination theory
Self-determination theory, particularly the sub-theory of cognitive evaluation

theory, has much to offer in the explanation of motivation in achievement
tasks as well as exercise. As seen in Chapter 3, cognitive evaluation theory
aims to explain variance in intrinsic motivation, most principally the

environmental contingencies that either promote or thwart intrinsically
motivated behaviour. In sport, just as in education and other achievement
contexts, these contingencies have to do with the nature of presentation of
rewards and the feedback given to individuals about their performance. This is
particularly relevant to those involved in the motivation of athletes in training
and competition such as coaches, managers, and parents as well as those pro-
moting exercise for health. This is because they are often in a situation where
they have the opportunity to foster the appropriate motivational context to
promote athletes’ intrinsic motivation. Intrinsic motivation is especially rele-
vant to athletes in achievement situations. Not only are the outcome psycho-
logical states of intrinsic motivation congruent with many athletes’ original
motives for participating in sport such as for enjoyment or pleasure, for com-
petence, and for affiliation (Ashford et al. 1993), but they are also relevant for
adherence and continued participation. Getting an athlete to self-regulate and
perform training behaviours and practices on their own without any external
reinforcement is essential for a coach, particularly for those who have limited
time to spend with athletes or who coach from a distance. Therefore strategies
to enhance intrinsic motivation are essential to a coach and these will be
reviewed later in this section.
The central premise of cognitive evaluation theory is concerned with how
rewards can potentially undermine intrinsic motivation, and how the nature
of the presentation of the reward and information regarding the reward
through interpersonal context may promote or enhance intrinsic motivation
(Deci and Ryan 1985). Interpersonal context refers to the manner in which
rewards are presented, usually through informational (autonomy supportive)
or controlling (autonomy thwarting) feedback. Rewards take a number of dif-
ferent forms; they can be tangible such as money, trophies, medals, distance
and proficiency badges, and certificates or intangible such as recognition,
praise, encouragement, enjoyment, satisfaction, and pride. Clearly, the set of
intangible rewards can be considered internal to the individual such as satis-
faction and enjoyment or external to the individual such as praise and recog-
nition. At the opposite pole to rewards are punishments and these can also be
either tangible, e.g. fines and withdrawal or removal of bonuses or intangible,
e.g. being dropped from the team, being dismissed from training, criticism,
guilt, and shame. All these rewards and punishments are types of reinforce-
ments that can be used to induce persistence in some sort of behaviour. Fur-
thermore, tangible rewards and punishments are very effective in maintaining
compliance and persistence in behaviour, but only to the extent that the
reinforcer remains omnipresent (Deci and Ryan 1985). Withdrawal of the
reward is likely to result in a decrease in motivation and reduced adherence to
the behaviour. This is particularly the case if the behaviour is made solely
contingent on gaining the reward. Cognitive evaluation theory maps how
non-contingent rewards may help motivate behaviour and maintain intrinsic
motivation (Deci and Ryan 1985).
Cognitive evaluation theory states that if the role of a reward has an infor-
mational function with respect to behavioural performance rather than the
behaviour being performed for the reward itself, then it will not undermine
intrinsic motivation (Deci and Ryan 1985). However, if a person engages in the
behaviour for the reward itself, the intrinsically motivating properties of
the behaviour will be lost to the individual and they will feel that their
engagement in the behaviour is outside their personal sphere of effectance or
‘locus of causality’. This has been labelled the undermining effect (Deci et al.
1999b) because the reward represents an additional reason for doing the
behaviour that is without explanation and above and beyond the intrinsic
reasons. Studies examining individuals solving interesting puzzles have sup-
ported this effect on a number of occasions. For example, Deci and Ryan
(1985) report a series of experiments in which participants solving a novel
puzzle were presented with a monetary reward. Half of the subjects were given
controlling feedback about their performance being told that they did as well
as they should according to the normative standard. The other half given
informational feedback about their performance; they were told that they did
well according to their own standards and the money was presented to
acknowledge their success. The group that received the controlling feedback
exhibited significantly lower levels of intrinsic motivation than the partici-
pants receiving the informational feedback, as indicated by free choice
engagement in the puzzle when the experimenter left the room. In addition to
the undermining effect, perceived competence also plays an important role in
cognitive evaluation theory. As reviewed in Chapter 3, feedback that supports
competence and is informational regarding performance (e.g. positive feed-
back, verbal praise) will enhance intrinsic motivation while feedback that
compromises competence and is controlling with respect to performance (e.g.
negative feedback, criticism) will reduce intrinsic motivation.
The undermining effect of rewards and other external contingencies such as
deadlines and feedback has been the subject of a large number of studies in a
number of achievement domains, but mostly in educational contexts. A num-
ber of meta-analytic studies have been conducted to examine robustness of the
findings for this effect across studies (e.g. Cameron and Pierce 1994; Deci et al.
1999a). Cameron and Pierce conducted a meta-analysis on 96 experimental
studies and found that verbal praise increased intrinsic motivation in
rewarded participants, but intrinsic motivation was not substantially under-
mined by the presence of a reward. They suggested that the undermining
effect is an exaggerated phenomenon only seen in certain contingencies, such
as when the reward was given for a behaviour with no prior expectation. They
also found that praise undermined intrinsic motivation when an extrinsic
reward was given. The authors concluded that the ‘negative effects of rewards
are limited and easily avoidable’ (1994: 29). However, Cameron and Pierce’s
analysis was criticized as confounding some interaction effects for rewards
and small effect sizes (Deci et al. 1999a). Deci et al. conducted another
meta-analysis of 128 experiments on the undermining effect and found
significant effect sizes for all types of reward contingencies (e.g. task-
contingent, task-completion, engagement contingent, and performance con-
tingent, see Chapter 3) for the undermining of intrinsic motivation. A strong
effect was also found for verbal praise on intrinsic motivation. The authors
concluded that these data provided strong support across a number of con-
texts for the undermining effect. On reflection, it seems that the balance falls
in favour of cognitive evaluation theory and that the undermining effect
may be a real one, and this may be especially so if the reward communicates
controlling contingencies.
In sport, cognitive evaluation theory has been investigated primarily with
respect to the effects of feedback on performers (Ryan et al. 1984). Self-
determination theory constructs in sport context have typically been meas-
ured largely by the sport motivation scale (SMS, Pelletier et al. 1995), which
has constructs common to measures of the perceived locus of causality in
other contexts and exhibits adequate construct validity and internal reliability
(Sarrazin et al. 2002). A number of studies have indicated significant effects of
positive feedback and verbal praise on athletes’ intrinsic motivation and com-
petence (e.g. Vallerand and Reid 1984; Whitehead and Corbin 1991). These
studies used intervention and control groups of athletes with the intervention
groups receiving varying levels of feedback from coaches or experimenters
about their performance of sport-related tasks. Furthermore, in these studies,
perceived competence seemed to be a mediator of the influence of intrinsic
motivation of sports performance, consistent with the notion that needs for
competence and self-determination are complementary, as has been shown in
research in other contexts such as education (Reeve 2002). However, to date,
no sports-related study has replicated the findings of Deci and colleagues in
terms of the effect of feedback on free-choice sport behaviour and interest, a
study that would unequivocally support self-determination theory in sport
contexts. Further, although many authors have cited the role that cognitive
evaluation theory has to play in examining the effects of rewards on sport
behaviour, few studies have examined the effect of a tangible reward on sports
performance.
Research in sport has also examined associations between perceived locus
of causality and other motivational variables. Pelletier et al. (1995) found
positive correlations between identified regulation and other hypothesized
determinants of sport behaviour such as perceived autonomy and com-
petence and outcome variables such as effort, intentions, and behaviour. Self-
determined forms of extrinsic motivation was found to influence persistence
among competitive swimmers over a 22-month period. Studies have also
pointed out positive relationships between perceived relatedness, autonomy,
and locus of causality (e.g. Ntoumanis 2001). This empirical evidence sup-
ports the construct validity of the perceived locus of causality in a sport
context and suggests that the motivational constructs in the sport motiv-
ation scales have a significant effect on key determinant and outcome
variables.
Self-determination theory and competition

Of supreme relevance to motivation in sport contexts is the application of
intrinsic motivation and cognitive evaluation theory to competition. In par-
ticular, theorists are interested under what conditions intrinsic motivation is
maintained in competitive situations regardless of objective outcome. For
example, is it possible to maintain intrinsic motivation in sports performers
whose performance outcome was a loss rather than a win? Deci and Ryan
(1985) have hypothesized that, consistent with cognitive evaluation theory,
competition would have the same undermining effect on intrinsic motivation
as extrinsic rewards because competition tends to make externally referenced
criteria for success salient to the individual. It was, however, also anticipated
that the undermining effect could be allayed by the presentation of the
competition as informational (e.g. gaining positive feedback about one’s own
performance) rather than controlling (e.g. trying to beat the others in the
competition). Experimental research in education (Reeve et al. 1985) and
sports contexts (Vallerand and Reid 1984) has suggested that participants for
whom the outcome of their behavioural engagement is failure or losing, gen-
erally report lower intrinsic motivation than those who are successful or win.
This suggests that competition may be inherently controlling. This conclusion
was supported by Deci et al. (1981) who found that participants solving puz-
zles together had significantly lower intrinsic motivation if they were told they
should try to beat the other person than when they were told to try to do their
best. Reeve and Deci (1996) attributed the inherently controlling and under-
mining nature of competition to the contingency of positive feedback on
winning and the negative competence feedback associated with losing.
However, it can be seen that it is the individual’s personal interpretation of
success and failure and the interpersonal context in which the competition is
presented that may also explain the undermining effect of competition on
intrinsic motivation. For example, McAuley and Tammen (1989) found that
basketball players who rated their perceived success highly had significantly
higher levels of perceived competence and intrinsic motivation compared
with those who reported lower ratings of perceived success. Importantly, there
were no differences in objective measures of winning or losing, suggesting that
it is the interpretation of the outcome that is important. The interpretation of
the competition outcome is likely to be dependent upon the degree to which
the structure of the competition was able to provide information about the
person’s performance that supported competence rather than being control-
ling and fostering a dependence on external cues. Vallerand and co-workers
(1986) found that participants in a competitive situation who were told they
would be compared to their other competitors exhibited lower intrinsic
motivation than those who were in a non-competitive situation and were
encouraged to do as well as they could on the task. Reeve and Deci (1996) also
found that when participants won a competition solving puzzles in an inter-
personal context that was non-pressuring, i.e. the participant was not told to
beat the other competitors, their intrinsic motivation was not undermined but
under pressuring conditions the undermining effect occurred. Importantly,
comparing these finding with a control group that had no competition and no
pressuring/non-pressuring feedback, Reeve and Deci found that the winning/
non-pressuring group participants had higher levels of intrinsic motivation
than the control group and the winning/pressuring group participants had
lower levels of intrinsic motivation.
The issue of competition becomes more complex when rewards are intro-
duced in conjunction with competition. However, the tenets of cognitive
evaluation theory are able to explain the effects of interactions between
competition, rewards, and interpersonal context on intrinsic motivation. In
competition, there are two types of rewards, competitively contingent rewards
and performance-contingent rewards. Competitively contingent rewards are
rewards that are attained for beating an opponent in direct competition, while
performance-contingent rewards are given when a normative standard is
reached. Thus, a performance-contingent reward can be attained by a number
of individuals while a competitively contingent reward can only be given to
the winner. In experiments studying competition and rewards, the effect of
three independent variables on intrinsic motivation can be seen: (1) the type
of the reward (competitive contingent or performance-contingent); (2) the
competitive outcome (winning or losing); and (3) the interpersonal context in
which the competitive task is presented (controlling or informational feed-
back). These groups are typically compared with groups that perform the task
in the absence of competition, receive no feedback, and receive no reward.
Research in this area has been limited, but findings suggest that when consid-
ered together, winners and losers receiving competitively contingent rewards
have lower levels of intrinsic motivation compared to a no-competition/
no-feedback/no-reward condition (Prichard et al. 1977).
Vansteenkiste and Deci (2003) investigated the effects of winning or losing
on intrinsic motivation in competitors who received competitive and
performance-contingent rewards. Winners were more intrinsically motivated
than losers, which partially reaffirms the premise that competition is inher-
ently controlling. However, it was found that among losers in competitive
situations positive feedback for meeting a specified standard (performance-
contingent feedback) went a long way to allay the negative effects of losing on
intrinsic motivation, while losers who received a reward for achieving a speci-
fied standard of performance (performance-contingent feedback) exhibited
reduced intrinsic motivation but their enjoyment of the task was unaffected.
The authors concluded that:
a focus on winning may indeed be counter-productive . . . with respect to

intrinsic motivation . . . If, instead of emphasizing winning above all
else, participants in activities and observers of the activities focused more
on good performance than on winning, the results for the participants’
motivation is likely to be far more positive. (2003: 298)
Thus, coaches would do well to help their athletes instill performance-related

goals that are personally relevant and informative about performance
rather than attending to goals relating to winning relative to others per se.
Furthermore, a coaching style that emphasizes the informational aspects of
competition and focuses on presenting competition as a means of measuring
performance against set criteria rather than against others is important.
Autonomy support in sport settings

Previously, in Chapter 3, the effect of perceived autonomy support of signifi-
cant others on the intrinsic motivation of students in a leisure-time exercise
context was discussed, particularly in the recently developed trans-contextual
model (Hagger et al. 2003b). Perceived autonomy support has also been found
to be a strong predictor of intrinsic motivation in educational (e.g. Koestner et
al. 1984; Reeve et al. 1999) and health (Williams et al. 1998; Williams et al.
2002) settings. Further, perceived autonomy support has been validated
against 20 core autonomy supportive behaviours in teachers in educational
settings (Reeve 2002). Given these findings and the research introduced in the
previous section on the importance of informative goals and feedback in the
maintenance of intrinsic motivation, the autonomy supportive behaviours
suggested by Reeve and colleagues (2002) are likely to have important effects
on intrinsic motivation and persistence in sport performance in training and
competition. This is because coaches’ autonomy supportive behaviours are
likely to promote intrinsic motives and enhance the informative nature of the
competition to their athletes. For example, Gagné, Ryan, and Bargmann (2003)
conducted a longitudinal diary-based study of the effect of perceived autonomy
support on the well-being of young gymnasts. They found that the gymnasts’
subjective well-being was determined by the relative satisfaction of psycho-
logical needs by the autonomy support gained from parents and coaches.
This preliminary evidence suggests that coaches’ autonomy-supportive
behaviours are strongly recommended because they tend to produce the desir-
able motivational orientations likely to maintain intrinsic motivation and
persistence in athletes.
Authoritarian and democratic styles

Legendary English soccer coach Brian Clough was notorious for his regi-
mented, authoritarian, and controlling coaching style and exhibited
behaviours on the training ground that exemplified those that would under-
mine intrinsic motivation. Yet on the pitch his players were among the most
committed, hard-working, and cohesive teams in the English league and
Clough steered them to successive European cup titles in 1979 and 1980. How
did the adoption of such an autocratic style, which seemed to go against
many of the predictions and recommendations of self-determination theory,
engender such seemingly intrinsically motivated behaviours in the team?
According to Iyengar and DeVoe (2003), the answer may lie in the structure
or underlying norms of the group and its effect on the way free choice and
the causality of behaviour is viewed, an explanation that is consistent with
self-determination theory.
Iyengar and DeVoe present evidence to suggest that people in collectivist
cultures tend to take into account the considerations of significant others
when making choices and decisions. As a consequence, people in collectivist
cultures tend to report higher intrinsic motivation when their choices are con-
sistent with the wishes of significant others or the cultural norm than when
making choices on their own. This may have a pervasive effect on how choice
and intrinsic motivation operate in a team context in which the pervading
environment is autocratic and no autonomy-supportive behaviours are dis-
played by the coach. Iyengar and DeVoe report experiments that examined the
effects of a collectivist culture on the intrinsic motivation. They used two
groups of children: European-Americans who had a predominantly individual-
ist cultural norm and endorsed personal choice and an independent view of
the self; and Asian-Americans whose prevailing cultural norm was collectivist
with an interdependent notion of self. Participants from each group were pre-
sented with the opportunity to solve a series of anagrams of equal difficulty
presented under three conditions. One group of children was told that the
anagrams had been chosen by an adult unknown to them (an experimenter)
while another was told that a significant other (their mothers) had chosen for
them. A third group was allowed to choose for themselves. In the experi-
menter choice condition, Asian-American and European-American reported
identical levels of intrinsic motivation. Most telling though was the finding
that European-American children reported significantly higher intrinsic
motivation than Asian-Americans children when personal choice was granted,
while Asian-Americans reported significantly higher intrinsic motivation
when they were told the significant other chose the anagram.
To explain these findings, it was proposed that the significant other choice
was more appealing to the preferences of children from the collectivist culture
because it reflected their tendency to respectfully account for significant
others’ views when making decisions. According to self-determination theory,
the participants from the collectivist group had internalized the group norm
to respect and accept the choices of other important members of the in-group,
and that is considered more important than personal choice. Thus, the choices
of others and those of the in-group are more salient to intrinsic motivation in a
collectivist culture than personal choice. Recent research suggests that differ-
ences in collectivist and individualist group norms also operate in social con-
texts not just in cultural groups (McAuliffe et al. 2003). Therefore, it is possible
that a collectivist group norm operated in Clough’s teams and they were
prepared to forego personal choice and volition because the group respected
and had internalized the instructions and autocratic style of the coach. In
such a context it is possible for intrinsic motivation to flourish in a team
where autonomy support is not evident and controlling leadership styles
pervade. However, it is always important for athletes to be given a choice when

working in an autocratic environment. If athletes are forced into accepting
authoritarian styles, then intrinsic motivation is likely to be undermined.
A hierarchical model of motivation

Vallerand (1997) proposed a hierarchical model of motivation that aimed to
integrate global, trait-like constructs related to motivation, such as psycho-
logical need satisfaction, with context- and situation-level motivation.
Vallerand’s model is an extension of self-determination theory that explicitly
hypothesizes links between the global, contextual, and situational aspects of
differing motivational styles adopted in the perceived locus of causality (see
Chapter 3). It also specifies links between social factors or interpersonal con-
text, motivational styles, and the consequences and outcomes of behavioural
engagement. In this model, global motivational constructs affect motivation
in specific situations via the mediation of context-level motivation (see Figure
5.3). Other than the proposal of three levels of generality, Vallerand states that
a complete model of motivation needs to account for intrinsic motivation
(IM), extrinsic motivation (EM), and amotivation (AM), the absence of inten-
tionality or goal-directed striving, and that motivation in a given context is
determined by situational factors and top-down effects from motivation at the
most proximal level. This model explicitly states that overarching motives that
reflect need satisfaction have wide-ranging and generalized effects across con-
texts and situations. However, having stated these hypotheses, the complexity
of the model has resulted in researchers seeking to support specific hypotheses
relating to its important postulates rather than testing the whole model
simultaneously.
One of the most comprehensive tests of the hierarchical model was provided
by Guay, Mageau, and Vallerand (2003). They aimed to examine the top-
down, bottom-up, and horizontal effects of some of the key constructs in the
model. Adopting a longitudinal design, the authors measured intrinsic motiv-
ation at the two levels of generality in two samples of school children at two
time points: one and five years apart respectively. They employed a cross-
lagged panel design, a powerful research design that enables the researcher to
better infer causality that cannot be done with data measured at a single time
point. The relationships tested are illustrated by the arrows in Figure 5.4. The
design permitted the researchers to establish whether global motivation influ-
enced situational motivation in a top-down fashion (arrow a in Figure 5.4),
whether situational motivation caused global motivation (arrow b in Figure
5.4), whether the relationships were reciprocal (arrows a and b in Figure 5.4),
or whether the majority of the variance in global and situational intrinsic
motivation was explained by the same variable measured at the previous
timepoint (arrows c and d in Figure 5.4). Results across both studies showed
that global motivation exhibited the greatest stability supporting the
proposition of the model that this construct reflects more generalized, less
Figure 5.3 Vallerand and Ratelle’s (2002) hierarchical model of intrinsic and
extrinsic motivation
Source: Vallerand and Ratelle (2002: 41)
changeable perceptions of motivation. Furthermore, a reciprocal effects model

best fit the data in both samples, suggesting that motivation at any level of
generality is caused partly by the stability of the motivation and perceived
motivation at the most proximal level.
In a sport context, Sarazin et al. (2002) provided a longitudinal test of the hier-
archical model in female handball players. In particular, they focused on the
proposed motivational sequence that determines behavioural consequences:
social factors→psychological mediators→type of motivation→consequences
(Vallerand 1997). The authors aimed to test whether global motives reflecting
players’ psychological needs for autonomy, competence, and relatedness (psy-
chological mediators) influenced intrinsic motivation (type of motivation)
and dropout intentions and behaviour (consequences). They found that the
contextual needs-related variables influenced contextual intrinsic motivation
Figure 5.4 Guay et al.’s (2003) hierarchical model of motivation, showing

top-down, bottom-up, horizontal, and reciprocal effects of global and context-
level motivation over time
Source: Guay et al. (2003: 994)
in keeping with the theory. Intrinsic motivation negatively predicted inten-

tions to drop out of handball participation and intentions to drop out signifi-
cantly predicted behaviour. This means that high levels of intrinsic motivation
offered a protective effect against specific intentions to drop out at the end of
the season. This study also examined the influence of social context on the
contextual motivation (Vallerand 1997). This provided a further test of
Vallerand’s hypothesis that an interpersonal context that promoted intrinsic
motives would have a positive effect on intrinsic motivation. In a leisure-
time context, this has been studied with respect to perceived autonomy
support (Hagger et al. 2003), but Sarrazin et al. (2002) capitalized on recent
evidence that a task-involving motivational climate (see previous section on
achievement goals) would enhance intrinsic motives. Indeed, the authors
found support for the sequence proposed by Vallerand (1997) from social con-
text (task-orientated motivational climate), to motivational styles (perceived
autonomy, relatedness, and competence), to intrinsic motivation, and situ-
ational decisions and behaviour (dropout intentions and behaviour). This
provides support for some major tenets of the hierarchical model in a sport
context.
Suggested reading
Biddle, S.J.H. (1999) Motivation and perceptions of control: tracing its development
and plotting its future in exercise and sport psychology, Journal of Sport and Exercise
Psychology, 21: 1–23. Award-winning review of motivational theories in sport and
exercise and control-related social cognitive constructs.
Biddle, S.J.H., Hanrahan, S.J. and Sellars, C.N. (2001) Attributions: Past, present, and
future, in R.N. Singer, H.A. Hausenblas and C.M. Janelle (eds), Handbook of Sport
Psychology (pp. 444–71). New York: Wiley. A review of attribution theory and its con-
tribution to exercise and sport psychology.
Chatzisarantis, N.L.D., Hagger, M.S., Biddle, S.J.H., Smith, B. and Wang, J.C.K. (2003) A
meta-analysis of perceived locus of causality in exercise, sport, and physical educa-
tion contexts, Journal of Sport and Exercise Psychology, 25: 284–306. Summarizes the
contribution of self-determination theory to sport.
Duda, J.L. and Hall, H. (2001) Achievement goal theory in sport: recent extensions and
future directions, in R.N. Singer, H.A. Hausenblas and C. Janelle (eds), Handbook of
Sport Psychology (pp. 417–43). New York: Wiley. Recent update of the role of achieve-
ment goal theory in sport.
Feltz, D.L. and Chase, M.A. (1998) The measurement of self-efficacy and confidence
in sport, in J.L. Duda (ed.), Advances in Sport and Exercise Psychology Measurement
(pp. 65–80). Morgantown, WV: Fitness Information Technology. A very informative
overview of self-efficacy theory in sport psychology.
Summary
• Motivation is a social cognitive construct that describes the intensity and

direction of an athlete’s activation or readiness to engage in a sports skill or
behaviour.
• The purpose of Heider (1958) and Weiner’s (1972) attribution theory was to
explain how athletes attribute the cause of their success or failure to various
sources. The sources are characterized according to three bipolar continua:
internal-external locus of causality, stable-unstable, and controllable-
uncontrollable.
• Social cognitive theory (Bandura 1977, 1997) has had a substantial impact
on sport psychology research and motivation. The theory proposes that
self-efficacy, as situation-specific self-confidence, is a strong predictor of
sport performance and salient outcomes. Self-efficacy is sourced from
performance accomplishment, vicarious experiences, verbal persuasion,

and physiological states.
• Achievement goal theory (Nicholls 1989) is one of the most frequently
cited theoretical approaches in sport psychology. The theory’s main prem-
ise is that athletes view their ability or success as either ego-oriented
(success is viewed as achieving performance outcomes like winning) or
task-oriented (success is viewed as achieving personal outcomes like
learning new skills). Achievement goal orientations have been found to
influence motivational constructs like intention, effort, and self-efficacy as
well as behavioural constructs like persistence, adherence, and sport
performance.
• Self-determination theory (Deci and Ryan 1985) aims to examine the events
that foster and maintain intrinsic motivation. Cognitive evaluation theory,
a sub-theory of self-determination theory, proposes that an athlete’s
intrinsic motivation is a function of whether their behaviour or perform-
ance is contingent on an extrinsic reward (e.g. money, verbal praise) and
whether the reward is interpreted as informational or controlling. External
rewards and controlling feedback/competition tend to undermine intrinsic
motivation while rewards and competition that are presented so as to give
informational feedback on personal success promote intrinsic motivation.
6
Athletes are emotional, too
In achievement contexts, considerable demands are placed on an individual

who has the potential to challenge their ability to cope and evoke a substantial
emotional response. Sport is an excellent example of such a context. Theories
of motivation in sport often cite positive affect – a positive emotional state – as
both an adaptive outcome of sport participation and a source of information
for future motivation to engage in the sport. However, the competitive nature
of sport also has the ability to evoke more negative or maladaptive affective or
emotional states. Sport, especially at the elite level, exerts considerable stress
on the athlete or performer. This is because at the highest level of performance
the stakes are very high; for example, professional sports performers depend
on success to earn their salary, prize monies, and win bonuses, as well as attain
intrinsic rewards such as personal satisfaction and self-esteem, rewards and
outcomes common to competitive athletes at all levels of sport. If there is a
mismatch between the demands placed on an athlete or sport performer by
their environment and their ability to cope with the concomitant emotive
states that arise from that demand, then it may interfere with their ability to
perform what Zajonc (1965) called the ‘dominant response’, i.e. the well
practised or trained movements and skills involved in sport performance. This
can often, catastrophically, manifest itself in sometimes embarrassingly poor
performances relative to performances in practice and training, even among
the most highly skilled athletes. This is a phenomenon often referred to as
‘choking’ (Baumeister 1984).
Many of us can think of occasions when this has happened in elite sport.
Think of France’s soccer team, the reigning World and European Champions
and tournament favourites at the 2002 FIFA World Cup in Japan, leaving the
championship after the first round in disgrace after an abysmal series of per-
formances in which they failed to win a match or score a goal. Think of Jean
van der Velde’s slump in the 1999 British Open golf championship when he
triple bogeyed the last hole after leading by five shots into the last. He needed
only a six to win, and he took seven in what is acknowledged as one of the
Athletes are emotional, too 131
greatest ‘chokes’ of all time. There are other examples, Martina Hingis resort-
ing to serving underarm against Steffi Graf in the 2000 French Open tennis
final after being unable to get any serve in and John Aldridge’s penalty miss
that handed the English FA cup to Wimbledon in 1988. Why should these
acknowledged champions fail to perform to anywhere near the high standards
they and others expect of them in high pressure situations? Anxiety, the set of
negative affective states associated with an inability to cope with stress placed
on an individual by environmental demands, is often the culprit. Elite and
professional athletes are schooled in the negative effects of anxiety and so-
called ‘negative’ emotional states on sport performance and many athletes
seek the help of sport psychologists for assistance with anxiety control.
Indeed, the majority of sport psychology consultations involve anxiety
management (Crocker et al. 1988). This chapter aims to evaluate social
psychological research into the role of emotion in sport and describe the
relationships between emotional states, psychological constructs, and sports
performance.
Why social psychological approaches to emotion and anxiety?
On the surface, research in cognition and emotion may not seem entirely
relevant to social psychological investigations into sport performance. How-
ever, an examination of the components of social cognitive theories will stand
as testament to the importance of emotion constructs in social psychology.
For example, as we saw in Chapter 2, affect has been shown to be an integral
component of the attitude construct in extensions of the theory of planned
behaviour (Hagger and Chatzisarantis, in press). Moreover, while models of
social cognition acknowledge that social information from the environment
(stimuli) and learnt personal belief systems are processed and serve as a basis
for motives, decisions, intentions, and behavioural responses, this does not
happen in a vacuum, devoid of feeling states or emotions (Eagly and Chaiken
1993; Perugini and Conner 2000). Emotions can also operate as response or
outcome states as well as sources of information for attributions, judgements,
beliefs, expectations, desires, intentions, and other social cognitive constructs.
Therefore, the study of social cognition and emotion in applied settings such
as sport is necessary, given the clearly complementary nature of these con-
structs, in order to explain the complex set of behavioural responses observed
in intense emotive sport situations, such as ‘choking’.
Defining affect, emotion, anxiety, arousal, and mood
Before embarking on a discussion of the role of emotions in social psycho-

logical research applied to sport performance, a prerequisite step is to define
emotion and emotion-related terms such as affect, emotion, arousal, mood,
132 Athletes are emotional, too
and anxiety. These terms are often used in a non-systematic manner and inter-
changeably making the interpretation and the exact nature of the explan-
ations offered by social psychological approaches of the role of affect-related
constructs in sports performance difficult. That stated, full agreement among
researchers in social cognition and emotion as to the distinction between
affect, emotion, and emotion-related terms is lacking and researchers state
that formal working definitions of these terms may be unclear due to substan-
tial overlaps between the concepts (Smith et al. 1993). Therefore, any def-
initions of affect-related concepts must indicate the boundaries, limitations,
and potential confounds.
Affect is a general or ‘umbrella’ term that encompasses all ‘mental feeling
processes’ (Bagozzi et al. 2002: 37), and therefore can account for the ‘felt’
aspects of emotion as well as the directive and motivational aspects, such as
the case of affective attitudes. A number of authors have suggested that affect
reflects ‘valenced feeling states’, a term that implies both directionality and a
number of specific emotion-related terms such as emotion and mood. There-
fore, emotion and mood can be considered as specific types of affective states
with anger, anxiety, guilt, and shame being specific examples of these emotion
types.
Formal definitions of emotion usually incorporate not only feeling states,
but also make reference to a ‘mental state or readiness’ arising from cognitive
interpretation of psychological and physiological states such as heightened
arousal (Smith et al. 1993). Furthermore, emotions are considered to have
direction towards a given object, person, or behaviour, much like attitudes,
particularly according to appraisal theorists (Smith et al. 1993). In addition,
emotion can also be described as having a behavioural or response function in
that it affects behaviour, such as facial expression of emotion and cognitive
and behavioural means to cope with the emotion (Bagozzi et al. 2002). Emo-
tions therefore have ‘action tendencies’. A vast array of emotions have been
identified in the social psychology literature, but ethological and cross-cultural
psychology research has identified six basic or core emotions: anger, fear,
sadness, happiness, surprise, and disgust (Ekman 1992). However, as we shall
see later, although not considered a core emotion, anxiety is recognized as an
important emotion in sport.
Mood is, by convention, considered different from emotion as it usually com-
prises a profile of different affective states, is less intense, more prolonged, and
with no action tendency. Mood is therefore less transient than emotion and
does not usually arise from the appraisal of specific events. However, the bound-
ary is sometimes less clear and both mood and specific emotions have been
implicated in sports performance. Indeed, some theorists claim that temporal
stability as a defining property of mood with respect to emotion is not valid
given moods and emotions can be both transient and prolonged (Frijda 1994).
Despite this lack of a clear distinction, mood is generally considered by emotion
theorists as different from emotion in its reduced ability to produce an action
tendency, its lower intensity, and its prolonged rather than transient nature.
Defining anxiety and arousal

More than any other single emotion, anxiety has been the focus of the vast
majority of research on emotion and social cognition in sports performance
(Gould et al. 2002). Anxiety is a specific emotion that has been described as an
unpleasant feeling of apprehension and distress, and is usually accompanied
by unpleasant physiological responses (Martens et al. 1990). Sensations such
as ‘sweaty palms’ (also known as ‘galvanic skin response’) and ‘butterflies in
the stomach’ (this may be the result of the shunting of blood from the stom-
ach due to the effect of catecholamines) are common physiological or ‘som-
atic’ symptoms of anxiety. Anxious athletes report these symptoms as well as
thoughts of negative performance expectations, a fear of failure, and inability
to concentrate (Jones and Hardy 1990). Modern theorists make the distinction
between state and trait anxiety. It follows that anxiety can be both a tendency
to respond with anxious symptoms in situations evaluated as being competi-
tive (trait-like) and a psychological state determined by environmental factors
such as competition and audience presence, as well as intrapersonal variables
such as the appraisal of the event as being important (state-like). In either case,
appraisal and cognitive-motivational-relational theories of emotion propose
that anxiety is a specific emotion with a specific pathology and characterizing
features.
Theorists also make the distinction between anxiety and arousal. Anxiety is
classed as having a somatic component (symptoms experienced physically e.g.
‘sweaty palms’, ‘butterflies in stomach’) and a cognitive component (symp-
toms felt psychologically e.g. ‘worry’, ‘inability to concentrate’) (Martens et al.
1990). Somatic anxiety is concurrent with some forms of physiological arousal
caused by changes in the sympathetic nervous system, and cognitive anxiety is
linked to the somatic form through the interpretive system that gives rise to
that heightened state of arousal. Of course, arousal itself is not anxiety, but it is
implicated in the anxiety process (Bagozzi et al. 2002). Somatic anxiety, for
example, is not physiological arousal but a person’s awareness of the symptoms
of arousal. Arousal is often considered a heightened state of activation in a
person’s physiological and psychological state. It is defined as a unidimen-
sional, ‘motivational construct’ (Landers 1980) and can be considered to oper-
ate on a continuum from very deep sleep to extreme excitement. Arousal is mani-
fested physiologically through changes in the autonomic nervous system
and hormones in the bloodstream that give rise to elevated heart rate, blood
pressure, perspiration rate, and muscle tension. A state of anxiety is often
accompanied by increased arousal, and, according to appraisal and cognitive-
motivational-relational theories of emotion, it is the interpretation of the
arousal that gives rise to specific emotions. Importantly, arousal is an intra-
personal variable that is likely to give rise to anxiety, but not all aroused
individuals become anxious, and the pathology of the arousal is such that
it may be a necessary but not sufficient condition for an anxiety response.
Arousal may therefore not always be accompanied by an anxiety response,
and early psychophysiology research showed that the interpretation of the

arousal could give rise to different interpretations of the accompanying
emotions (e.g. Schacter and Singer 1962). Recent appraisal theorists believe
that arousal is implicated in emotional responses such as anxiety, but there are
specific patterns of emotional responses according to the way in which the
arousing situation is appraised.
Trait versus state distinction

Early research with measures of anxiety considered only the trait aspect of
anxiety. Anxiety was viewed as a stable facet of personality and therefore con-
sidered trait-like in nature. In this view, anxiety was not directly like personal-
ity in the strictest sense because anxiety tendencies were considered to have
both an innate and learnt component. Early researchers such as Sarason et al.
(1960) produced scales that tapped anxiety as a general disposition that
determined anxiety responses in a variety of situations. It was thought that
individuals would exhibit characteristic behavioural patterns according to
their levels of trait anxiety (see Frijda 1994).
However, Spielberger, Gorusch, and Lushene (1970) noted that the explan-
ations provided by the conceptualization of anxiety as a trait did not yield
particularly satisfactory results. Spielberger et al. contended that anxiety
should have both state-like and trait-like properties. State anxiety was defined
as feelings of apprehensiveness and tension that were usually paired with
arousal of the autonomic nervous system (Spielberger et al. 1970). It was
contended that while trait anxiety may explain some variance in anxiety
states in given situations, it could not explain all the variance in the state
level of anxiety because such states were determined by more proximal situ-
ational factors and the individual’s interpretation of them. It is clear that
such a premise is a precursor of appraisal theories in cognition and emotion.
Trait anxiety therefore served as an indicator of an individual’s tendency to
interpret ambiguous situations as threatening (Frijda 1994). State anxiety, on
the other hand, is the actual level of anxiety in a given situation, all disposi-
tional and situational factors considered. Spielberger et al. subsequently
developed an inventory to measure both components; the State–Trait Anxiety
Inventory (STAI). Subsequent tests of anxiety appeared in the sport psych-
ology literature measuring both the trait (Sport Competition Anxiety Test;
SCAT) and state components of anxiety for competitive sport (Competitive
State Anxiety Inventory; CSAI) (Martens et al. 1990). The CSAI is particularly
interesting and important in this regard because it distinguishes between the
somatic and cognitive components of anxiety but also introduces a third
element, self-confidence to account for the ‘positive’ aspects of anxiety
extracted in factor analytic studies of emotion-related scales in sports per-
formance. A more in-depth review of the CSAI and its revisions is provided
later in this chapter.
Applying social psychology research on anxiety to sport
The academic study of anxiety in sport has undergone a series of evolutions

since Martens et al. (1990) introduced the first formal means of measuring the
construct. This evolution has mirrored the development of emotion research
in the general social and personality psychology literature. Theories of anxiety
in competitive sport were founded in early research in personality psychology
(Gould et al. 2002) and considered anxiety as a trait, or at least a trait-like
construct. However, such generalized theories had limited explanatory power
in predicting behaviour, a limitation attributed to the fact that such inventor-
ies were too far removed in their generality to have any bearing on action and
performance in specific situations. Spielberger and colleagues pioneered the
hypothesis that anxiety should be segregated into trait and state forms and
developed the SCAI for this purpose, and modern sport psychologists adhere
to this model of anxiety.
The state–trait distinction in anxiety research in sport psychology was
mirrored in the development of inventories to measure both components.
Martens developed the SCAT to measure trait anxiety which was defined as a
tendency to interpret competitive situations as threatening with concomitant
feelings of apprehension and tension (Martens et al. 1990). As a result of Spiel-
berger et al.’s distinction between state and trait anxiety, Martens et al.
developed the CSAI, and state anxiety was defined as an immediate, transient
feeling of tension and apprehension in a specific competitive sport situation.
The CSAI adopted the approach set out by Spielberger et al. and used many
items from the original STAI, but adopted items that had relevance to competi-
tive situations. Extensive development and validity research supported the
validity and reliability of the CSAI as a unidimensional measure of competitive
state anxiety. However, psychophysiological research and trends in test anx-
iety indicated that in competitive situations a distinction should be made
between the felt or experienced symptoms of an anxious state or somatic anx-
iety, often associated with heightened arousal interpreted as an anxiety
response, and the psychological disruption, tension, and worry termed
cognitive anxiety. In addition, seminal work by Burton (1988) suggested that
for a comprehensive evaluation of the anxiety–performance relationship,
researchers needed to adopt a multidimensional model incorporating the
cognitive-somatic distinction.
This compelled Martens and co-workers (1990) to develop a state anxiety
inventory that explicitly made this distinction. The CSAI-2 was therefore pro-
posed and researchers adopted a rigorous classical test theory approach to the
development of the inventory content. Martens et al. also aimed to tap other
components specific to sports-related distress such as physical harm and
generalized anxiety. Exploratory factor analyses of the initial pool of items
extracted three factors. One factor clearly contained items reflecting the
somatic component of anxiety (e.g. ‘my body feels tense’, ‘I feel jittery’).
However, the items purportedly measuring cognitive anxiety loaded on two
different factors. The content of the items loading on the first of these factors
reflected negative aspects of the cognitive anxiety pool of items (e.g. ‘I am
concerned about performing poorly’, ‘I am concerned about choking under
pressure’) while the item content of the remaining factor reflected positively
worded items from the cognitive anxiety item pool (e.g. ‘I feel self-confident’,
‘I am confident of coming through under pressure’). These factors were
labelled cognitive anxiety and self-confidence, respectively. The structure and
content of the resulting 27-item inventory representing Burton (1988) and
Martens et al.’s (1990) multidimensional model of anxiety have received
much attention in the literature (Burton 1998; Gould et al. 2002; Craft et al.
2003). However, while in the initial validity studies the use of the CSAI-2
supported its construct and factorial validity, discriminant validity, test–
retest reliability, internal consistency, and predictive validity in terms of
levels of the cognitive and somatic components prior to competition (Mar-
tens et al. 1990), recent research has questioned some of these initial
analyses.
The CSAI-2 has been further developed using confirmatory factor analysis
that has a number of advantages over the exploratory model used in the
initial development of the inventory (Cox et al. 2003). These analyses have
revealed that the inventory had a problematic structure that was mainly
attributed to the inclusion of items that displayed high residual variance, that
is, items that were not adequately representative of their hypothesized latent
factor. Cox systematically eliminated 10 items from the original inventory to
produce a more parsimonious 17-item revised CSAI-2 that exhibited good fit
with multiple samples in subsequent confirmatory factor analyses (Cox et al.
2003).
In addition to questions surrounding the construct validity of the CSAI-2,
studies have also indicated that the relationships between its components are
highly variable. The correlations between somatic and cognitive anxiety
components and self-confidence usually reflect a theoretically predictable pat-
tern with negative relations between self-confidence and the two anxiety
components and positive relationships observed between the two anxiety
components. Schwenkmezger and Steffgen (1989) meta-analysed a series of
studies examining these interrelations and suggested that the corrected correl-
ations were in the predicted direction and were significantly different from
zero. However, more recent meta-analytic studies on the CSAI-2 have indi-
cated that the relationships between the components of anxiety are signifi-
cant, relatively strong, and in the predicted direction (Craft et al. 2003). These
results lend support to the notion that the multiple components of competi-
tive anxiety exhibit discriminant validity. In summary, despite problems with
the factor structure that have been resolved through the modification of the
inventory by confirmatory factor analysis, the CSAI-2 appears to have
adequate conceptual and measurement properties. The next section will
review the efficacy of research examining the anxiety-sport performance
relationship.
Anxiety and the prediction of sport performance
Anxiety–performance hypothesis
One of the key criteria for the validation of the CSAI-2 and an important
hypothesis of multidimensional anxiety theory is the anxiety–performance
hypothesis. Martens et al. (1990) and Burton (1998) suggested that as competi-
tion approached, the three components of anxiety would exhibit a character-
istic pattern in terms of level and influence on performance. It was proposed
that self-confidence ratings would increase prior to competition and then be
subject to changes within competition. It was hypothesized that somatic anx-
iety levels would be relatively low until shortly before competition, would
rapidly increase immediately prior to competition, and then rapidly decrease
thereafter. It was expected that cognitive anxiety would be at an elevated state
prior to competition and decrease at the onset of competition, but be subject
to changes during competition, particularly for open skills. Research has gen-
erally supported the predicted temporal fluctuations in anxiety levels prior to
performance and has shown the detrimental effects of these levels of anxiety
on sport-related cognitive and motor tasks as competition approaches
(Martens et al. 1990). In addition, research has suggested that self-confidence
ratings tend to remain stable leading up to competition, but tend to decrease
just prior to and during competition (Martens et al. 1990).
Martens et al. also expected the strength of relationships between the differ-
ent components of somatic, cognitive, and self-confidence and performance
to vary prior to competition. They report a study examining relationships
between sport performance and cognitive and somatic anxiety in elite golfers
in non-competition (1–2 days before competition), pre-competition (1 hour
before), and mid-competition (after first 9 holes) indicated that non-
competition and pre-competition cognitive and somatic anxiety levels did not
determine initial performance (first 9 holes) but significantly predicted later
performance (last 9 holes). Mid-competition scores for both cognitive and
somatic anxiety did predict later performance. This provides some empirical
support that pre-competition state anxiety only interferes with subsequent
performance, which is contrary to anecdotal observations that elevated anx-
iety just prior to an event may hinder immediate performance. Furthermore,
findings from this study did not support the differential effects of cognitive
and somatic anxiety on performance, the effects were similar for both anxiety
components. However, subsequent follow-up studies showed stronger nega-
tive relationships between cognitive anxiety and sport performance for com-
petitive swimmers and a positive relationship between self-confidence and
performance and a curvilinear relationship with somatic anxiety (Burton
1988). These results have been supported in other studies, although the tem-
poral patterning of the correlations has not received consistent support, as
indicated by Martens et al. (1990).
Meta-analysis of the anxiety–performance relationship

A large number of studies have examined the anxiety–performance relation-
ship, and the vast majority have adopted the CSAI-2 to measure anxiety
(Burton 1998) and the multidimensional model of anxiety (Burton 1988;
Martens et al. 1990). Initial results from the validation studies were promising,
acknowledging the influence of cognitive anxiety on performance was nega-
tive, the impact of self-confidence on performance was positive, and that som-
atic anxiety exhibited a curvilinear relationship on performance (Burton
1988). Interestingly, the role of task complexity and duration on the somatic
anxiety–performance relationship did not yield consistent results. In addition,
results of subsequent studies have yielded inconsistent findings for the
anxiety–performance relationship for the three components of anxiety.
In an attempt to resolve these inconclusive findings, Craft et al. (2003) con-
ducted a meta-analysis of 29 studies examining relationships between the
multiple components of the CSAI-2 and sport performance. The authors
reported strong and significant intercorrelations among the somatic, cogni-
tive, and self-confidence anxiety subcomponents supporting their discrimin-
ant validity. Using the attenuation-corrected correlations in a multivariate
regression of the dependent variable of performance on the three anxiety
components permitted the authors to evaluate the unique effect of each com-
ponent on sport performance across all of the studies. Significant attenuation-
corrected beta-weights (βc) were observed between performance and cognitive
anxiety (βc = 0.13, p < 0.05), somatic anxiety (βc = 0.09, p < 0.05), and self-
confidence (βc = 0.36, p < 0.05). However, homogeneity tests for these
corrected correlations indicated that the relationships were heterogeneous
in all cases, indicating a significant proportion of the error variance in the
relationships remained unexplained after correcting for sampling error. This
indicated that these relationships were affected by moderator variables.
Craft et al. also examined the impact of several moderator variables: type
of sport (team/individual), skill type (open/closed), type of athlete (elite/
European/college athlete/college PE student), and time of CSAI-2 administra-
tion (< 15 mins/16–30 min/31–59 min/1–4 hours prior to performance). It was
hypothesized that individual sports performers would report higher levels of
anxiety (Beedie et al. 2000), that performance-based open skills such as team
sports would be affected more by anxiety levels due to greater interaction with
other competitors and a changing environment than closed skills like golf or
rowing (Terry and Youngs 1996), that elite level athletes may experience
stronger effects of anxiety on sport performance even though they may be
more accustomed to it than recreational athletes (Kliene 1990), and that anx-
iety would be a better predictor of sport performance the closer to competition
it was measured (Martens et al. 1990). Results revealed significant attenuation-
corrected beta-weights for all anxiety components on performance level for
studies on athletes in individual sports and for open skilled sports. These find-
ings support a previous theory that a continuously changing environment and
interactions with others perhaps place more demands on the performer and
therefore increase performance-related anxiety responses. Although, analo-
gously, individual sports performers seem to demonstrate a greater anxiety–
performance relationship, particularly for self-confidence, this is probably
because there are no teammates to moderate levels of anxiety and pressure is
greater when competing alone.
Anxiety levels and self-confidence ratings in elite level athletes, particularly
European club-level athletes, had the strongest impact on performance than
any other athlete group. Interestingly, at this level, cognitive anxiety and
somatic anxiety had a positive effect on performance, which has implications
for the facilitative anxiety model that will be reviewed later (see Jones et al.
1994). Paradoxically, the anxiety levels from the CSAI-2 seemed to have the
strongest influence at an intermediate time-point prior to competition (31–59
min) compared with longer and much shorter time intervals. The authors
suggested that proximal levels of anxiety have a time lag before they have an
impact on performance, possibly because self-reported anxiety levels are
unrealistic or inaccurately reported, while distal measures of anxiety are not as
relevant because the competition is a long way off. In summary, the meta-
analysis suggests that although cognitive and somatic anxiety are related to
sports performance, and moderated by type of sport, type of skill, athlete level,
and time prior to competition, they have only weak influences on perform-
ance. Self-confidence levels have a stronger impact on performance and are
much more consistently related to performance than the other anxiety com-
ponents. These results therefore point to the importance of the self-confidence
variable, corroborate a cognitive approach to the study of anxiety, and suggest
that means to promote more positive cognitive-affective states would be most
efficacious in improving sport performance.
The inverted-U hypothesis

From the outset it was stated that arousal is not an equivalent term for anxiety.
Indeed, some cognitive theories of emotion suggest that physiological changes
such as arousal are unnecessary for the expression of emotion and such arousal
states are artefacts of that felt emotion (Smith et al. 1993). However, psycho-
physiological theories and studies have implicated arousal in the expression of
emotions such as anxiety (e.g. Schacter and Singer 1962). The multi-
dimensional model of anxiety explicitly states that the symptoms of arousal
are separate from the cognitive component of the anxiety construct, and
called somatic anxiety. Therefore these theories imply that physiological
arousal is expected to accompany the performance situations that give rise to
competitive anxiety in an individual. The role of arousal in anxiety-evoking
competitive situations cannot be denied and an evaluation of the role of
arousal is essential to the understanding of the anxiety process in competitive
sport situations.
Early theories of arousal in sport drew on Yerkes and Dodson’s (1908)
proposed simple linear relationship between arousal and performance. This

was derived from experiments with mice that Yerkes and Dodson observed
were superior at negotiating complex mazes when more physiologically
aroused. They hypothesized that the heightened state of physiological arousal
triggered an increased intensity to satisfy innate physiological needs such as
thirst and hunger, hence it was termed ‘drive’ theory. In competitive sport, it
was hypothesized that heightened physiological arousal would increase psy-
chological and physical preparedness for competition, and the theory sug-
gested that the higher the arousal, the better the preparedness and hence the
better the performance (see Figure 6.1, broken line). However, observations
that very low or very high levels of arousal resulted in inferior performance of
fine motor skills and complex cognitive tasks when compared to intermediate
arousal levels in competitive sport led to several researchers proposing that an
optimal level of arousal was the most effective for performance (Oxendine
1970; Landers 1980). This relationship was referred to as optimal arousal the-
ory or the inverted-U hypothesis, named because of the shape of the curve
representing the relationship on arousal–performance axes (see Figure 6.1,
solid lines). Indeed, Oxendine (1970) suggested that for sports involving gross
motor activities and large muscle groups (e.g. boxing, athletics, weightlifting),
a high level of arousal was considered necessary for optimal performance.
However, in sports where very fine motor skills were prevalent (e.g. golf,
snooker, darts), a much lower level of arousal was deemed vital for optimal
performance, such that a level of arousal that may be associated with optimal
performance for a 100m sprinter would be debilitative to performance for a
table pool player. Thus different inverted-U relationships would exist for per-
formers of different sports, dependent on the demand characteristics of the
sport in terms of the gross or fine nature of the movements (see Figure 6.1,
solid lines for sprinter and table pool player).
While the inverted-U hypothesis is attractive because of its neat and clear set
of predictions, it has been criticized because of its over-simplification of the
role of arousal and the nature of skills within a variety of different sports (Jones
1990). Indeed, Landers (1980) suggested that optimal arousal theory did not
account for the complex blend of fine and gross motor skills that combine
most sports. For example, soccer comprises gross movements in terms of mul-
tiple sprints and sustained running required for movement about the pitch, as
well as fine skills involved in ball control, dribbling, and turning. Sports such
as this are therefore difficult to classify on a bipolar continuum of motor skills
ranging from gross to fine. Further criticisms of the optimal arousal theory
were levelled at the purely descriptive relationship between performance and
anxiety that did not offer an explanation as to why performance is sub-optimal
at lower and higher levels of arousal. Multidimensional anxiety theory has
attempted to provide a better explanation of the relationships between anx-
iety and arousal by incorporating somatic anxiety as a measure of the ‘felt’
symptoms of heightened arousal (Burton 1988). In addition, the shape of the
curve in the arousal–performance relationship was criticized as being
Figure 6.1 Optimal arousal and drive theories
inadequate and a more complex relationship has been proposed by Hardy and
co-workers (Hardy 1990; Hardy and Parfitt 1991) (see catastrophe theory, this
chapter). In summary, while arousal has been implicated in emotional re-
sponses such as anxiety in competitive situations, and optimal arousal theory
provided an early theoretical basis for the arousal–performance relationship,
limitations of these theories have demanded the development of more sophi-
sticated explanations of the arousal–performance relationship embedded in
cognitive theories of anxiety.
Antecedents of anxiety in sport: theoretical approaches
Thus far the focus of this chapter has been on examining the role of anxiety on
sport performance. However, just as anxiety is an antecedent of sports per-
formance, social psychological theories have identified a number of constructs
that give rise to elevated anxiety in competitive situations. Such antecedents
are particularly relevant for targeting key variables for intervention. For
example, if the reduction of cognitive anxiety will result in a concomitant
increase in performance as the hypotheses of multidimensional anxiety theory
predict, then the psychological variables that reduce cognitive anxiety would
be useful targets for intervention. In this respect such psychological constructs
would have an indirect effect on sports performance mediated by anxiety (see
Figure 6.2). Therefore, distal, trait-like constructs (such as competitive trait
anxiety, or goal orientations) would determine more proximal, state-like
Figure 6.2 Mediational model of hierarchical anxiety and performance
anxiety constructs, which would predict sport performance, as proposed by

authors of mediational models in the social psychology literature.
Trait anxiety
As mentioned previously, one of the means cited by Martens et al. (1990) in
multidimensional anxiety theory to account for baseline or ‘typical’ anxiety in
athletes would be to control for trait anxiety. Partialling out trait anxiety
would serve to control (statistically set all individuals in an interindividual test
of the anxiety–performance relationship to the same level of trait anxiety) the
effect of a general tendency to be anxious in competitive situations and permit
the examination of the unique effects of situated, state anxiety on perform-
ance. It is therefore of interest to researchers concerned with predicting per-
formance whether trait anxiety explains variance in state anxiety, and more
importantly if trait anxiety accounts for variance in sports performance and
whether the trait anxiety–performance relationship is accounted for by state
anxiety. This would give rise to a top-down, hierarchical model of anxiety
from a generalized, global construct to more specific situational judgements in
keeping with other hierarchical social cognitive models of behaviour (e.g.
Vallerand 1997), as shown in Figure 6.2.
Research has indicated significant relationships between trait anxiety and
the components from the CSAI-2 (Gould et al. 1984; Yan Lan and Gill 1984;
Crocker et al. 1988). In terms of the specific components, differential predic-
tions of cognitive and somatic anxiety are inconclusive with some researchers
finding strong correlations for trait anxiety with both cognitive and somatic
components (Gould et al. 1984), and some with either cognitive (Crocker et al.
1988) or somatic (Yan Lan and Gill 1984) components alone. In addition, trait
measures of self-confidence have predicted state measures of self-confidence
(Vealey 1986). However, few studies have controlled for trait levels of anxiety
when predicting sport performance from state anxiety levels, and this remains
a useful avenue for further research.
Goals and motivational orientations

Social cognitive theories of intention and behaviour in sport have recognized
the importance of social cognitive variables as antecedents of competitive anx-
iety. Swain and Jones (1992) and Hall and Kerr (1998) adopted approaches
from competitive orientations and goal orientations pioneered by Gill and
co-workers (Gill and Deeter 1988) and Duda and Nicholls (1992, see Chapter 5)
respectively to explain the interpersonal antecedents of competitive state anx-
iety. Swain and Jones (1992) measured the different competitive orientations
and related them to the three components of competitive anxiety from the
CSAI-2. The Sport Orientation Questionnaire (Gill and Deeter 1988) was used
to measure track and field athletes’ levels of competitive and win orientation.
It was expected that a win orientation and competitiveness would be most
strongly related to anxiety levels. Results showed that competitiveness was
strongly related to all three components of anxiety, but a win orientation was
not. Importantly, the relationships between competitiveness and the somatic
and cognitive components were negative, suggesting that athletes who per-
ceived themselves to be least competitive exhibited the greatest levels of anx-
iety. Athletes who viewed themselves as being more competitive were more
self-confident and therefore more able to handle competitive situations, as
indicated by the positive correlation between competitiveness and self-
confidence.
These findings have been bolstered by significant correlations between goal
orientation and competitive state anxiety, although relationships have been
inconsistent. For example, Duda et al. (1995) found that an ego orientation
determined the degree of cognitive anxiety in athletes across a number of
competitive sport contexts, while Ommundsen and Pedersen (1999) found no
relationship between ego orientation and cognitive anxiety. Instead they
found that a task orientation and perceived competence significantly and
negatively predicted cognitive anxiety, and perceived competence negatively
predicted somatic anxiety. Finally, Newton and Duda (1995) found that
expectations of performance outcome were the best predictors of cognitive
anxiety and self-confidence, rather than goal orientations. These results sug-
gest that the role of goal orientations may be inconsistent and that perform-
ance expectations in the form of perceived ability and competence better
account for variance in competitive anxiety levels.
The adoption of cognitive-motivational-relational theories of motivation
has provided more comprehensive mediational models of the appraisal and
coping processes implicated in the role of social cognitive variables such as
goal orientations on anxiety and emotion. Gaudreau, Blondin, and Lapierre
(2002) examined the role of coping potential in accounting for the relation-
ship between goal orientations and negative affect (including anxiety) in sport
performers two hours prior, one hour after, and 24 hours after competition.
Results indicated that the discrepancy between task and ego goal orientations,
a relative measure of how personal or performance-related an individual’s goal
orientation is, was significantly related to positive and negative affect. Import-
antly, active coping strategies like planning mediated the goal orientation
discrepancy–positive affect relationship, while behavioural disengagement
mediated the relationship between goal orientation discrepancy and negative
affect. This suggests that adaptive coping strategies tend to account for
cognitive appraisals of the situation as relevant to personal goals and positive
affective responses, and passive coping strategies account for the negative
relationship between goal orientation and negative affect. This suggests that
active coping strategies, like planning, have an adaptive function (Gaudreau et
al. 2002).
Goal orientations, motivational climate, and anxiety

Given the evidence that motivational orientations are predictive of competi-
tive state anxiety, researchers have also examined the characteristics of the
perceived competitive environment or motivational climate in giving rise to
anxiety responses. Since there is research to support the influence of motiv-
ational climate on goal orientations and motivational states (Duda 1993;
Seifriz et al. 1992, see Chapter 5), it has been hypothesized that goal orienta-
tions may mediate the effect of perceived motivational climate on anxiety
states in sport performers. Ntoumanis and Biddle (1998) supported this
hypothesis and illustrated no direct effects of motivational climate on the
components of state anxiety in team-sport athletes. Instead, the positive
effects of a perceived performance- or ego-oriented motivational climate on
cognitive and somatic anxiety were exerted through ego orientation and self-
confidence. Again the role of self-confidence is pervasive in the effect of
achievement goal orientations on anxiety.
In a further test of this hypothesis, White (1998) examined the role of
achievement goal orientations and perceived parent-initiated motivational
climate on competitive trait anxiety in adolescent athletes. Results indicated
that athletes who scored low on task orientation and high on ego orientation
viewed the motivational climate endorsed by their parents as one that fostered
low effort-contingent success and worry over making mistakes. This group
exhibited the highest competitive trait anxiety, indicating that athletes with
low-task, high-ego motivational orientations viewed the climate engendered
by their parents as performance-focused and were more likely to interpret
competitive situations as anxiety-provoking. Future directions of research will
examine the causal nature of these relationships and establish whether a per-
formance-involved motivational climate induces a motivational orientation
that is low-task and leads to a tendency to be more highly anxious in
competition.
In terms of interventions, a recent study by Yoo (2003) introduced an
intervention to change goal orientations and competitive anxiety levels in

adolescents attending tennis classes. Participants were assigned to a class that
had a pervading task-involving or performance- or ego-involving motivational
climate for six weeks. Performance levels significantly increased and cognitive
and somatic anxiety levels were significantly decreased for participants acting
in a task-involved motivational climate while anxiety did not change and
performance significantly decreased in participants receiving an ego-involving
motivational climate. These preliminary results support relations between
perceived motivational climate, achievement goal orientations, and levels of
anxiety in athletes and suggest that interventions can attenuate anxiety levels
if a mastery climate is engendered.
Recent theories of anxiety in sport
Intensity versus direction distinction

The previous section discussed the social cognitive antecedents of the sub-
components of anxiety, with a focus on climates and interventions that would
reduce the effects of the cognitive and somatic constructs. This assumes that
anxiety, particularly the cognitive component, has a detrimental or debilitative
effect on performance, and focuses solely on the level or intensity of athletes’
anxiety response. However, a recent branch of research has examined the
potential of heightened cognitive and somatic anxiety to have a positive or
facilitative effect on performance. Jones and co-workers (Jones et al.1993; Jones
et al. 1994; Jones and Swain 1995) hypothesized that it is not the intensity of
the athlete’s anxiety response per se that has a debilitative effect on perform-
ance, rather, it is the interpretation or direction of the anxiety response that
will determine its impact. They proposed that the appraisal of the competitive
situation as threatening as well as a secondary appraisal of coping ability, or
control over resources to cope with the threatening situation, would deter-
mine whether the anxiety response would be interpreted as facilitative or
debilitative.
Research support for the debilitative-facilitative conceptualization of anx-
iety has focused first of all on whether the directional component of anxiety
predicted performance and, second, whether control-related constructs would
moderate the directional component of the anxiety response. If the theory is
correct and in keeping with cognitive-motivational-relational theories of
motivation, then sports performers with high control and self-efficacy would
interpret anxiety as facilitative and demonstrate stronger relationships
between directional anxiety scores and sport performance. Jones et al. (Jones et
al. 1993, 1994; Jones and Swain 1995) developed a measure of competitive
state anxiety direction that was administered concurrently with the CSAI-2
which has become known as the Directional Modification of the Competitive
State Anxiety Inventory-2 (DM-CSAI-2, Burton 1998). The measure asks
respondents to label or interpret the anxiety symptoms on the CSAI-2 as

facilitative or debilitative, and research has shown that the intensity and
directional components for cognitive and somatic anxiety exhibit discrimin-
ant and predictive validity (Jones et al. 1993; Jones et al. 1996). Jones, Swain,
and Harwood (1996) found that sports performers who reported high positive
affect interpreted their anxiety as more facilitative, but a concomitant pattern
was not found for negative affect and debilitative anxiety. Mellalieu, Hanton,
and Jones (2003) found that athletes with a facilitative interpretation of
their anxiety or ‘facilitators’ labelled their pre-competitive affective states as
positive compared with ‘debilitators’ who interpreted their pre-competitive
affective experiences as negative.
Research examining the influence of anxiety direction on performance has
typically split groups of athletes into high and low anxiety on the direction
subscales. These studies have consistently indicated that the intensity of the
anxiety response for both the cognitive and somatic anxiety subscales is con-
sistent in performers regardless of their direction, highlighting the importance
of accounting for direction scores. Importantly, direction scores are strongly
correlated with performance, such that better performers report their anxiety
as more facilitative. For example, Jones et al. (1993) demonstrated that
gymnasts reporting ‘good’ performances on a balance beam trial reported
their anxiety to be facilitative while poorer performers reported their anxiety
to be debilitative. In addition, research has shown that a direct linear relation-
ship exists between cognitive anxiety direction scores and performance in
basketballers, while the intensity components cognitive anxiety and self-
confidence reflected an inverted-U relationship (Swain and Harwood 1996).
This indicates that interventions to positively affect sport performers’ anxiety
direction scores are likely to have a concomitant positive effect on perform-
ance. Thus, the first premise of the theory, that performers reporting high
facilitative anxiety will influence performance, has been supported and has
shown discriminant and predictive validity in sports with gross and fine motor
skills.
Subsequent research has also focused on the effects of competence-related
variables such as ability, goal attainment, and self-confidence on the relation-
ship between the directional component of anxiety and performance. For
example, elite swimmers and cricketers were found to interpret their anxiety as
more facilitative than their non-elite counterparts, even though the intensity
levels of anxiety did not differ across the groups (Jones et al. 1994). In addition,
Jones and Hanton (1996) found that swimmers with positive expectancies of
goal attainment reported higher levels of facilitative anxiety than swimmers
with vague or low expectancies of goal attainment. Self-confidence, another
competence-related variable, has been shown to be higher in athletes with a
facilitative interpretation of their anxiety (Jones et al. 1993, 1994; Edwards and
Hardy 1996). One possibility is that the self-confidence component may be the
mechanism by which facilitative anxiety direction scores predict performance
and there may be an interaction effect. Future research will further examine
the interaction between self-confidence and performance. Such a study

may test the hypothesis, based on previous findings, that a sports performer
reporting a facilitative anxiety interpretation and high self-confidence anxiety
pattern would exhibit the most optimal level of performance.
The question of how to promote a facilitative interpretation of anxiety
symptoms has been the subject of recent research. Researchers have attempted
to identify the strategies adopted by athletes with an adaptive anxiety direc-
tion profile and designed interventions based on these findings (Hanton and
Jones 1999a, 1999b). Hanton and Jones (1999a) followed up their empirical
work supporting the effect of facilitative interpretations of cognitive and som-
atic anxiety in sports performers by qualitatively analysing interviews with
elite swimmers who consistently reported facilitative anxiety interpretations.
Results of a content analysis of the transcripts revealed that the elite swimmers
had acquired a number of key strategies, either intuitively or through formal
psychological skills training, that were implicated in their facilitative inter-
pretation of anxiety symptoms including goal setting, imagery, and self-talk.
Based on this research, Hanton and Jones (1999b) developed an intervention
strategy to engender a more facilitative interpretation of anxiety in elite
swimmers that consistently reported debilitative interpretations prior to com-
petition. Using an idiographic, single-participant approach across a series of 10
competitions, three of the four swimmers involved in the intervention
reported facilitative interpretations, and this change was still evident in a
post-intervention follow-up. These results support the use of cognitive
interventions to change the interpretation of the anxiety response, and are
congruent with problem-focused and emotion-focused coping strategies from
cognitive-motivational-relational theories of emotion.
In summary, Jones’ (1995) control model of anxiety has advanced research
on the anxiety-performance relationship and has illustrated that the effects of
anxiety on performance may not be one of mere negative affect on perform-
ance. However, the theory has been criticized by a number of researchers.
Some have questioned the validity of the DM-CSAI-2, suggesting that the dir-
ectional measures have not received much formal validation in a manner
commensurate with the rigorous development of the CSAI-2 (Burton 1998).
Researchers have also questioned the concept of the directional measure, that
it may, instead be a measure of positive affect or ‘excitement’ that has been
noted by Jones (1995). Research is needed to further validate the directional
component, particularly with respect to measures of positive affect. Further-
more, given that the directional component of anxiety in the DM-CSAI-2 is
measured alongside measures of state and trait anxiety, some researchers
believe that the directional scale should interact with the intensity scale given
the large inter-individual variation of intensity scores (Burton 1998). In this
way, the directional component would moderate the effect of the intensity
component on performance, although this hypothesis has not been supported
by empirical evidence (Edwards and Hardy 1996).
Catastrophe theory
One additional limitation that has been levelled at Jones’ (1995) control
theory of facilitative-debilitative anxiety is that it does not explicitly account
for the level of arousal experienced by an individual. Indeed, Edwards and
Hardy (1996) have noted that anxiety intensity can have a facilitative effect on
performance when levels of physiological arousal are low and debilitative
effects when physiological arousal levels are high. These findings suggest a
more complex pattern of relationships between anxiety intensity, direction,
and performance, and imply that physiological arousal needs to be implicated
in any complete model of the anxiety process in sport performance. Following
Zeeman’s (1976) adoption of catastrophe models that aimed to explain the
discontinuous relationships evident in normally linear or continuous func-
tions, Hardy (1990) adopted the most common of the ‘catastrophe’ models –
the ‘cusp’ catastrophe – to model discontinuities in performance due to the
influence of physiological arousal and cognitive anxiety (Figure 6.3), i.e.
catastrophe theory.
The essential premise of the ‘cusp’ catastrophe model of anxiety and sport
performance is that at low levels of cognitive anxiety, the relationship bet-
ween physiological arousal and performance will follow an inverted-U relation-
ship – the continuous predictable relationship proposed by optimal arousal
theory. Conversely, at high levels of cognitive anxiety, the performance–
Figure 6.3 Hardy’s (1990) catastrophe theory

Source: Jones and Hardy (1990: 88)
physiological arousal relationship will increase in keeping with the inverted-U

relationship until optimal arousal is reached. However, an increase in physio-
logical arousal above the optimal point will result in a steep, ‘catastrophic’
drop in performance. The model is best illustrated using a set of three-
dimensional Cartesian co-ordinates (see Figure 6.3). In Figure 6.3, physio-
logical arousal, cognitive anxiety, and performance levels are expressed by the
x, y, and z axes, respectively. At low levels of cognitive anxiety, the arousal–
performance relationship follows a flattened inverted-U relationship, as
reflected by the far face of the graph. However, if cognitive anxiety is elevated
and arousal is increased performance will only follow the inverted-U relation-
ship until optimal arousal is reached, and this is illustrated by the arrowed line
on the upper part of the performance surface. The ‘fold’ in the performance
surface illustrates the catastrophic decline in performance once optimal
arousal is exceeded under conditions of high cognitive anxiety. The catas-
trophe theory is unique in that it is the only model that explicitly states that
cognitive anxiety and physiological arousal interact in their effect on perform-
ance. Moreover, the hypotheses from other theories, such as Jones’ control
model of anxiety can be incorporated in the model. For example, cognitive
anxiety does not always have a debilitative effect on performance, but can
actually facilitate performance at optimal levels of arousal, such that under
high levels of cognitive anxiety, performance will be substantially better than
low levels of cognitive anxiety provided arousal is at an optimal level (Edwards
and Hardy 1996). If arousal is too great, however, performances under high
levels of cognitive anxiety will be substantially worse than performances
under low levels of cognitive anxiety.
Importantly, once the performance catastrophe has occurred, the performer
cannot recover their performance until their physiological arousal and cogni-
tive anxiety state levels have returned to baseline levels, this is known as the
hysteresis hypothesis. This is illustrated in Figure 6.3. Under hysteresis, per-
formance will follow a different path when physiological arousal is increasing
compared to when it is decreasing. In this case, if, after the catastrophe, arousal
continues to increase, performance will not increase, and may continue to
decline, but at a slower rate. If arousal is decreasing, then performance may
start to recover, but will only do so once physiological arousal has decreased to
a much lower level than when the catastrophe occurred.
Although the original catastrophe models were notoriously difficult to test
using experimental designs (Zeeman 1976), the advent of specific cata-
strophes, like the cusp catastrophe, permitted the testing of specific hypoth-
eses relating to the cusp catastrophe model of anxiety and performance
(L. Hardy 1990). Hardy reports a series of studies in the initial development
of the theory to test its corollaries. For example, Hardy and Parfitt (1991)
tested the interaction of physiological arousal and cognitive anxiety on
performance and the hysteresis hypothesis in female basketball players. An
experimental manipulation of the three components of the catastrophe model
– physiological arousal, cognitive anxiety, and performance – was adopted.
Physiological arousal was manipulated by increasing the heart rate of an indi-

vidual by exercise to a given target zone. Cognitive anxiety was manipulated
using the time-to-event paradigm based on research that has shown high
levels of cognitive anxiety one-day prior to competition and low levels of
cognitive anxiety one-day after competition. Performance was tested using a
free-throw basketball test, an ecologically valid test of performance for the par-
ticipant-group. Results supported the catastrophe in performance under high
levels of cognitive anxiety and high-arousal. Findings also indicated that per-
formance under high cognitive anxiety was different for participants whose
level of arousal was increasing compared to when it was decreasing, support-
ing the hysteresis hypothesis. Subsequent studies have also provided support
for the facilitative effect of high cognitive anxiety on performance under con-
ditions of low or intermediate levels of physiological arousal and a debilitative
effect under levels of high physiological arousal (Edwards and Hardy 1996).
Hardy (1996) clarified the corollaries of catastrophe theory and presented
statistical procedures to provide an omnibus test of the theory. Hardy tested
the model with experienced golfers using self-reports of anxiety from the
CSAI-2, heart rate as a measure of physiological arousal, and golf putting as an
objective measure of performance. Hardy used direct differences methods to
test the fit of the catastrophe curves to the data and provided some evidence in
support of the cusp curve. Importantly, the inclusion of self-confidence as an
independent variable in the model increased the predictive efficacy of the
model, suggesting that self-confidence should still be implicated in any inter-
active effect of cognitive anxiety and arousal on performance (Gould et al.
2002). Recently, some qualitative support for the catastrophe model has been
promulgated. Edwards, Kingston, Hardy, and Gould (2002) conducted a
hierarchical content analysis of interviews on performances from eight elite
athletes about phenomenological experiences of ‘catastrophic’ drops in
performance during competition. Two themes emerged, one representing a
sudden drop in performance and the other continued performance deterior-
ation. The sudden, steep decrease in performance followed by a shallow con-
tinued decrease was characteristic of the catastrophe predicted by Hardy’s
(1990, 1996) theory and the hysteresis hypothesis. Suffice to say that catas-
trophe theory is an important addition to the literature, but tests of the model
have been limited, perhaps due to its complexity, and further investigation is
required, particularly with respect to omnibus tests of the model.
Individual zones of optimal functioning (IZOF)

Many of the approaches thus far have adopted quantitative, empirical investi-
gations into the social psychological approach to anxiety and emotion in
sport. The focus of these has been the examination of inter-individual differ-
ences in social psychological constructs and emotion variables. These investi-
gations have been predominantly nomothetic (group-oriented) in approach
with the advantage that findings are assumed to be generalizable across
athletes within the confines of the validity of the methodology and the homo-
geneity or representitiveness of the sample. Hanin (1995, 2000) proposed the
individualized zones of optimal functioning (IZOF) model, an alternative
approach to the study of emotion in sport performance. The model offers an
integrated perspective on emotional experience and sport performance that
adopts hypotheses from person–environment interactions theory, appraisal
theories of emotion, idiographic versus nomothetic views of personality, gen-
eral systems theory, trait–state distinctions, and psychological readiness for
competition. This approach is largely from an individualized or idiographic
approach, but also attempts to generalize to athletes on the basis of trends in
intra-individual and inter-individual emotional experiences of successful and
poor performances.
The aim of the IZOF model is to provide a holistic view of the emotion–
performance relationship in sport and it includes a number of specific hypoth-
eses (Hanin 2000). First, emotions arise from the cognitive appraisal of the
probability that individual goals will be achieved in a given sport context.
Second, athletes learn specific emotional responses to given competitive situ-
ations over time. Third, emotional responses are specific to the individual, the
context, and time frame of the competitive event. Fourth, there is a reciprocal
relationship between emotion and performance. And finally, different emo-
tions exert optimal or dysfunctional effects on performance. In summary, the
IZOF model focuses on the effect of an individual’s personal emotional
responses to a given competitive situation on performance and the effect of
performance on emotional responses.
Much of the research on the IZOF model has focused on pre-competitive
anxiety because it has provided useful descriptions, particularly in an indi-
vidual setting, of the relationship between multiple dimensions of emotion,
including anxiety, and sport performance (Hanin 2000). Key principles of the
IZOF with respect to anxiety, and indeed other emotional states related to
sport, is that every athlete has an optimal level or range of emotional intensity
(high, medium, or low), such as anxiety and other emotional states, that will
lead to successful performance in sport. An emotional profile that falls out
with this range, i.e. above or below the optimized levels, will be debilitative
towards performance. This is known as the in-out of zone principle. An indi-
vidual’s optimal level of anxiety is established using recall methods of success-
ful (‘best ever’) and poor (‘worst ever’) performances to establish the optimal
zones of anxiety necessary for successful performance. Hanin (1995) found
that participants were surprisingly accurate in recalling their pre-competition
anxiety levels even after a substantial time lag of up to four months. These
studies supported the IZOF model indicating performance decrements when
sports performers’ anxiety levels fell outside their recalled optimal zones.
However, the cognitive-somatic anxiety distinction was not clearly met in
early work with the IZOF model, and researchers found they have had equal
success in using somatic anxiety measures to identify optimal zones of func-
tioning (Morgan et al. 1988). Hanin’s (2000) reconceptualization of the model
to include a full complement of emotional responses in competition addressed

the multidimensionality of competitive state anxiety and subsequent studies
have illustrated that separate zones could be established for the cognitive and
somatic aspects of competitive state anxiety. Studies have supported the basic
premises of the IZOF model, indicating that a wide variability in performance
is exhibited at given levels of pre-competitive state anxiety (Hanin 1995). A
more sophisticated model was adopted by Gould et al. (1993) using the inter-
active effects of both cognitive and somatic anxiety to produce a unitary opti-
mal zone. This research found strong support for the IZOF model, stronger
than studies that have used the cognitive or somatic anxiety dimensions
alone.
Narrative and meta-analytic reviews have also supported the basic premises
of the IZOF model for the anxiety–performance relationship (Gould and
Tuffey 1996; Jokela and Hanin 1999). Gould and Tuffy (1996) conducted a
critical review of IZOF model research and concluded that performance on the
basis of the in-out of zone principle was generally supported, although the
authors had some theoretical and methodological reservations. Jokela and
Hanin (1999) conducted a meta-analysis of 19 empirical studies of the IZOF
model with 146 effect sizes and found fairly good support for the in-out of
zone principle such that athletes in their optimal zone performed significantly
better than athletes out of their zone with a medium effect size (Cohen’s d =
0.44). In addition, the study also provided additional support for athletes’
ability to recall and anticipate their levels of competitive state anxiety. The
authors concluded that the IZOF model was efficacious in discriminating
between high and low performing athletes, and claimed that the specific
optimal levels of anxiety were superior in the prediction of performance than
individually unspecified levels of anxiety intensity from other models using a
nomothetic approach (e.g. Kliene 1990).
One advantage of the IZOF model is that it has intuitive or face validity,
which is useful for sports psychologists who typically adopt measures such as
performance profiling to gauge the success of their interventions. It is one of
the few theories in sport psychology that has been developed with specific
recommendations for intervention in naturalistic settings in mind. In add-
ition, most applied sports psychologists work using an individualized or
idiographic approach, so a model developed in a naturalistic setting is advan-
tageous because it is supremely relevant to practitioners. This makes intuitive
sense for sports psychologists focusing on the pragmatic issues of resolving
maladaptive anxiety patterns in their performers.
However, the model is not without its critics. One criticism levelled by
Gould and Tuffey (1996) is that the theoretical underpinning of the IZOF
model is questionable because it does not explain the antecedents or pre-
dictors of optimal anxiety, but instead focuses on the individual nature of the
anxiety–performance relationship. Moreover, many of the studies have
adopted small sample sizes and methodological inadequacies such as lack of
objective measures of performance and longitudinal measures (Gould and
Tuffey 1996). In addition, Hanin (2000) claims that the model is able to gener-
ate further hypotheses through an inductive, grounded-theory approach.
While such an approach has its advantages, the co-existence of a hypothesis-
driven, quantitative, deductive framework with an inductive, data-driven,
qualitative approach does not make for a happy union. The main reason for
this is that one cannot generate a theory at the same time as one tests a theory.
For example, with this ‘combined approach’ findings that serve to disconfirm
the theory on the basis of the falsification of a proposed set of hypotheses can
lead to a researcher formulating a theory to support the falsified results, which
does not sit easily with the principle of falsification. Finally, the model has
been criticized for being based or developed on successful performance and
unsuccessful performance or failure, which are both qualitatively and quanti-
tatively different, and therefore the model may not have efficacy in predicting
anxiety–performance relationships across a range of intensity scores.
Other theoretical accounts of emotion in sport
Mood states in sport

Researchers in the field of emotion in sport have criticized the strong focus on
competitive state anxiety in studies in sport and have called for the study of a
broader range of emotional and feeling states in sport contexts (Hanin 2000).
Research on mood states has provided insight into the effect of a more varied
range of feeling states and their effects on sport performance. Mood states are,
by definition, different from specific emotions like anxiety in that they are
more enduring and less intense feeling states but their effect on sports per-
formance is thought to be substantial. Due to the number of feeling states that
constitute mood, much research has adopted a profile approach (McNair et al.
1971). The profile approach suggests that while individual mood or feeling
dimensions can be distinguished at the subordinate level, it is the pattern-
ing or overall profile of the levels of these moods that gives rise to specific
cognitive and emotional outcomes in sport performance.
Investigation into mood states in sport owes much to the work of Morgan
and colleagues who adopted the profile of mood states to investigate perform-
ance levels in athletes (Morgan 1980; Morgan et al. 1988). They adopted the
profile of mood states (POMS, McNair et al. 1971) inventory to distinguish
between successful and unsuccessful performance in elite and non-elite
performers. The POMS has 30 items that measure six subscales: tension,
depression, anger, vigour, fatigue, and confusion. Morgan et al. found that
elite athletes exhibited a characteristic profile of depressed levels of all of the
POMS subscales with the exception of the vigour subscale which tended to be
rated very highly prior to competition (see Figure 6.4). This characteristic
‘spike’ in the mood state pattern that distinguished elite performers was
termed the iceberg profile.
Figure 6.4 Morgan’s (1980) iceberg profile
While Morgan’s iceberg represented a neat and intuitively clear patterning

of mood states that is predictive of performance, other researchers have been
critical of its true value. Cockerill, Neville, and Lyons (1991) indicated that the
POMS exhibits little or no variance in elite and non-elite performers in the
hours prior to performance, while there is considerable variance in the days
leading up to competition. They suggested that as an immediate predictor of
performance, mood was relatively poor, but a better prediction could be made
in longitudinal analyses of the POMS over the days leading up to competition.
Analogously, Rowley et al. (1995) conducted a meta-analysis of 16 studies that
had adopted the iceberg profile to predict sport performance. Findings indi-
cated that although the mood profile significantly predicted sport perform-
ance, the effect size was small (Cohen’s d = 0.15) and only accounted for less
than 1 per cent of the variance in sport performance. It was concluded that
despite a significant relationship, the explanatory value of the iceberg profile is
questionable and explains a degree of variance in sport performance that is
largely unsubstantial.
In addition, subsequent research adopting an alternative perspective to the
explanation of elite sport performance has provided findings contrary to
Morgan’s original results. For example, studies in sport training have found
that the POMS could not readily distinguish between elite and less successful
athletes but it could readily distinguish between high-fit and low-fit athletes
and was a useful tool for establishing the effects of overtraining in endurance
athletes (Berglund and Safstrom 1994). In addition, mood states have been
shown to be efficacious in predicting successful and unsuccessful perform-
ances in fitness tests in adolescents (Lane and Lane 2002). This is in keeping
with the clinical context in which the POMS was developed. Indeed, Prapaves-
sis (2000) reviewed the conceptual issues and research associated with POMS
and mood state investigation in sport and concluded that the POMS was an
inappropriate model to apply to the study of pre-competitive mood states
because of its lack of explanatory value, inconclusive findings, and limited
theoretical underpinning.
Recent research in mood states has provided strong, psychometrically sound
instruments to measure mood states in sport contexts and advanced research
in the prediction of sport performance from mood states. For example, Terry
and colleagues (2003) have developed a revised inventory that has been shown
to have satisfactory construct, predictive, and concurrent reliability, and
adequate reliability in a number of athletic populations. The advent of such
inventories resolved some of the critiques of the methodology in mood state
research (Cockerill et al. 1991; Berger and Motl 2000). Indeed, research with
these mood state instruments has yielded successful explanation of perform-
ance in a number of sport contexts (e.g. Lane and Lane 2002; Terry et al. 2003).
In addition, Beedie, Terry, and Lane (2000) conducted a meta-analysis of the
POMS in sport contexts, but, unlike Rowley et al., made the distinction
between studies in which the dependent variable was level of achievement
and studies where the dependent variable was performance outcome. In this
distinction, sport achievement was considered measures of absolute attain-
ment in sport such as winning a gold medal while sport performance measures
reflected relative attainment such as gaining a personal best time. In keeping
with Rowley et al.’s findings, the effect size for the influence of mood on sport
achievement was small (Cohen’s d = 0.10). However, the effect of mood on
sport performance outcomes yielded a moderate effect size (Cohen’s d = 0.31).
In addition, Beedie et al. examined the effects of the individual subscales from
mood state measures and found a moderate and positive effect size for vigour,
and moderate and negative effect sizes for tension and depression with sport
performance outcomes, but small or no effect sizes for the other subscales. The
authors concluded that these findings supported the model proposed by Mor-
gan and co-workers (Morgan 1980; Morgan et al. 1988) and provided support
for the importance of the vigour subscale in the prediction of performance
outcomes.
Importantly, these findings have stressed the salience of examining and dif-
ferentiating between the individual subscales from the POMS, and not the
adoption of aggregate scales, a finding that has been corroborated elsewhere
(Hanin 2000; Lane and Lane 2002). Terry and Slade (1995) also recommend
that not only should individual mood subscales be the focus of research, but
the mood subscales may interact in the prediction of sport performance. They
indicated that depression, for example, might moderate the effects of other
mood state subscales like anger and tension on sport performance. For
example, Terry and Slade studied the influence of both competitive state anx-
iety and mood states on karate performance. Results indicated that the som-
atic, cognitive, and self-confidence components of competitive state anxiety
and the individual mood states of vigour and anger were able to discriminate
between successful and unsuccessful performers. Finally, Beedie et al.’s (2000)
meta-analysis also illustrated the importance of the definition of the perform-
ance parameters and measures to be studied. Clearly, the prediction of athletic
achievement may be too insensitive and crude a measure of sport perform-
ance, and the measurement of relevant performance outcomes is recom-
mended when examining the efficacy of mood states in predicting
performance.
Reversal theory
Reversal theory is a relatively new approach applied to the study of anxiety
and emotion in sport performance, but recent applications have been promis-
ing and have received empirical support. Reversal theory provides a general
framework for the understanding of the relationships between arousal and
emotion, and how these influence motivational constructs and behaviour
(Apter 1982). A key premise of reversal theory is that an individual can inter-
pret his/her state of arousal as pleasant or unpleasant, and this is known as
hedonic tone. Given that an individual may also experience high or low levels
of arousal, this gives rise to a 2 × 2 formation of hedonic tone and arousal level.
In this paradigm, high levels of arousal interpreted as pleasant result in an
excited affective state, high levels of arousal interpreted as unpleasant produce
an anxious state, low levels of arousal interpreted as pleasant generate a
relaxed state, and low levels of arousal interpreted as unpleasant give rise to a
boredom response (Figure 6.5).
Reversal theory predicts that an individual’s metamotivational state will
determine the relationship between their hedonic tone and arousal level. A
metamotivational state is a person’s interpretation of their motives or goals in
a given context and at a given point in time. An individual can either be
goal-focused or ‘serious’ in their pursuit of their outcomes, known as a telic
metamotivational state or be activity-orientated or ‘playful’ in their approach,
known as the paratelic metamotivational state. When in arousal-evoking situ-
ations, a person can switch between metamotivational states that will deter-
mine whether they will be on the relaxation–anxiety hedonic tone-arousal
curve (solid line, Figure 6.5) or the excitement–boredom hedonic tone-arousal
curve (broken line, Figure 6.5). This ‘switch’ is known as a reversal, and as
metamotivational states are state-like constructs and subject to changes in
the environment, reversals can happen at any time during an activity. A telic
metamotivational state is often viewed as arousal-avoiding while a paratelic
metamotivational state is often viewed as arousal-seeking or thrill-seeking.
Figure 6.5 Relationship between arousal and affect in reversal theory

Source: Kerr (1985)
Evidence for this has been shown in the tendency for people who approach
or participate in thrill-seeking activities (e.g. frightening funfair rides like
rollercoasters) to exhibit a high paratelic state and increased preferred arousal
with positive emotional states like excitement and positive emotions com-
pared with individuals who perform ‘safe’ activities. Therefore, both situ-
ational factors and habituation (known as ‘satiation’) of the activity and
metamotivational state may result in a reversal (Kerr 1997). Thus, reversal
theory, like catastrophe theory, attempts to integrate affective states with
activation or arousal states. The theory proposes that an individual’s inter-
pretation of arousal as pleasant or unpleasant is dependent on whether their
metamotivational state is arousal-avoidant/telic or arousal-seeking/paratelic.
There have been numerous tests of reversal theory since Kerr’s initial appli-
cation of the theory in sport contexts, and many provide support for its
hypotheses. Bellew and Thatcher (2002) tested the factors affecting reversals
from telic to paratelic states in a naturalistic setting with rugby players and
found that reversals generally occurred as a consequence of factors external to
the athletes, or frustration caused by external factors. The authors concluded
that internal factors such as satiation or feeling the same state for an extended
period of time were not responsible for changes in metamotivational state.

This finding is supported by previous research that has showed that
unexpected environmental events like errors are an important influence on
inducing reversals from telic to paratelic metamotivational states in athletes
(Kerr and Tacon 2000). Research has also focused on the relationships between
metamotivational states and motivation in sport. For example, Lindner and
Kerr (2001) found that participation motivation was predicted by metamotiva-
tional states from general or life-oriented and contextual or sport-oriented
sources.
Other studies have provided only limited support for some of the premises
of reversal theory and questioned some of its predictions. For example,
Legrand and LeScanff (2003) used an idiographic approach to the study of a
javelin thrower’s metamotivational states before competition. They found
that hedonic tone did not distinguish between high and low achievement
across the season, but did find that when individual mood components were
considered rather than global emotional groupings, the mood states of placid-
ity, anger, boredom, and provocativeness were found to fluctuate significantly
between the thrower’s best and worst performances of the season. Similarly,
Kerr, Fujiyama, and Campano (2002) found that serious (telic) and hedonistic
(paratelic) recreational tennis players did not differ on measures of tension
and effort stress, but significant decreases in overall ‘negative’ emotions and
tension stress was evident in both groups after the completion of a practice
session, but especially so in the serious group. Taken together, the findings of
reversal theory are not unequivocal. However, like catastrophe theory,
research on reversal theory in sport has not been the subject of intensive
research, unlike multidimensional anxiety theory, and thus support in terms
of the number of findings is limited. Clearly, more investigation needs to be
done, but laboratory- or field-based studies in particular are required to test the
existence of reversals and whether they can be induced experimentally.
Suggested reading
Craft, L.L., Magyar, T.M., Becker, B.J. and Feltz, D.L. (2003) The relationship between
the Competitive State Anxiety Inventory-2 and sport performance: a meta-analysis,
Journal of Sport and Exercise Psychology, 25: 44–65. Very recent cumulative analysis of
the impact of competitive state anxiety on sport performance.
Hanin, Y.L. (2000) Emotions in Sport. Champaign, IL: Human Kinetic. Hanin’s detailed
reader on his IZOF model with important contributions from Vallerand on the
importance of emotions in models of social cognition in sport.
Jones, G. (1995) More than just a game: research developments and issues in competiti-
tive anxiety in sport, British Journal of Psychology, 86: 449–78. All about the control
model of facilitative and debilitative anxiety in sport.
Summary
• Affect is an umbrella term that includes all ‘feeling states’, with emotion
and mood as specific examples. Emotions are single, intense, and change-
able ‘feeling states’ that tend to have ‘action tendencies’ while mood tends
to be conceptualized as a profile of affective states that is considered more
enduring, less intense, and with no action tendency.
• Competitive state anxiety is considered to be multidimensional compris-
ing cognitive anxiety, somatic anxiety, and self-confidence (Martens et al.
1990). Cognitive anxiety, in particular, is consistently and negatively linked
with sport performance (Craft et al. 2003), but may be positively related to
performance if it is interpreted as facilitative (Jones 1995).
• Arousal may be an outcome of increased state anxiety, and is implicated in
the anxiety-performance relationship in catastrophe theory (Hardy and
Parfitt 1991). The theory states that the arousal–performance relation-
ship will follow an inverted-U relationship under conditions of low
cognitive anxiety, but a catastrophic drop in performance will occur under
conditions of high cognitive state anxiety and high arousal.
• The individual zone of optimal functioning model (IZOF, Hanin 2000) and
reversal theory (Kerr 1997) focus on an individual-based approach to emo-
tional processes in sport. The IZOF identifies specific limits in which anxiety
is ‘optimal’ and athletes within their zone exhibit better performance.
Reversal theory charts how the interpretation of arousal and hedonic tone
give rise to specific metamotivational states that are related to motivation
and behaviour in sport.
7
Group processes in sport
‘There is no I in team’ and ‘the whole is greater than the sum of its parts’; these
oft-cited clichés provide a useful prologue to this chapter that aims to evaluate
the salient group processes that occur among athletes in sport contexts. In
team sports, success is often contingent on the team working effectively and
collectively toward a common goal. Often this means accepting given roles
and responsibilities within a team, foregoing personal ambitions, investing
effort for the common good, having extensive knowledge of other team mem-
bers’ abilities, and providing support and feedback for the members of the
team. Within sport teams, there is a wealth of information available to each
member regarding the team’s operation and effectiveness, as well as informa-
tion regarding their own personal performance within the team. Therefore,
individuals in a team behave on the basis of information from the group –
about their membership and role in the group – as well as their own beliefs,
personality, and other intrapersonal constructs. Social psychology has a lot to
offer in terms of explaining how individuals respond and interact in groups
and how they affect individual and group behaviour and performance. This
section will examine some of the key processes that underlie group or team
performance in sport and will examine the theoretical and empirical work that
has attempted to explain these processes.
Definition of a group and a conceptual framework
A group is not a mere collection of two or more individuals. The mere presence
of others is a necessary but not sufficient condition for a group to be estab-
lished (Zajonc 1965). As we shall see in the next section, the presence of others
when performing motor tasks like sports skills does have an influence on cog-
nition, behaviour, and performance, but this form of social influence is differ-
ent from that which operates in a group. Membership of a group gives rise to a
specific set of social cognitions or beliefs among the group members brought
Group processes in sport 161
on by the interaction between group members, the notion of a common goal

or desired outcome, a specific structure within the group, and the presence of
important group processes such as communication and cohesion (Carron and
Hausenblas 1998). Therefore, a group comprises one or more people, involves
interaction between people, demands an awareness of some form of common
fate or goals – although this may not be clearly defined, has a specific structure
known to all members such as the role and status of individuals within the
group, and group norms.
Why group membership?

One of the key motives often cited for participation in exercise and sport is
one of affiliation (Ashford et al. 1993), and this can be considered indicative
of a basic human psychological need to form social bonds or relationships
with others. For example, Baumeister and Leary (1995) cited the need to
form attachments to others and form interpersonal relations as a funda-
mental human motive. These authors found that the motivation to form
interpersonal relations satisfied a number of essential criteria to establish it
as a fundamental motive; it operates across a wide variety of contexts; it
influences a variety of important social cognitive and emotional constructs;
it can predict behaviour in a variety of contexts; it is independent of other
psychological and biological needs; it can affect psychological adjustment
and well-being if absent; and it is essential for all humans. Organismic theor-
ies of motivation also share the view that the need for affiliation or related-
ness is an essential human need. Deci and Ryan (1985, 2000) suggest that
relatedness is a basic psychological need, and is complementary but
independent of other psychological needs for competence and autonomy.
Thus, membership of groups is something that is essential for psychological
health and satisfies our innate needs. Sport is a social setting in which group
affiliation processes abound and is a prime example of people satisfying their
basic needs to form interpersonal relationships or relatedness. Further, not
only does group affiliation assist in satisfying group needs, it is considered
an essential part of the way in which we define ourselves. Often an indi-
vidual’s self-esteem is, in part, derived from the social groups to which he or
she belongs (Tajfel and Turner 1986). It is therefore important when con-
sidering group behaviour in sport that such behaviour is determined by
innate needs and that a person’s sense of self is likely to be intrinsically tied
in with group behaviour.
A conceptual framework
Carron and Hausenblas (1998) provide a conceptual framework to study the
major factors that influence group behaviour in sport. The proposed model is
given in Figure 7.1. The model outlines the major influences on group struc-
ture and formation, group processes, and group outcomes. Group structure
162 Group processes in sport
Figure 7.1 Carron and Hausenblas’ (1998) conceptual framework for the
study of sport teams
Source: Carron and Hausenblas (1998: 20)
comprises the norms within the team (i.e. what is acceptable and unaccept-
able behaviour) and the roles (e.g. leadership, status, and function) of the
group members. Member attributes such as abilities and personalities and
group environment such as size and territory contribute to the group struc-
ture. Group structure influences important group processes such as team
goals, collective efficacy, and cooperation within the group. This is pro-
posed to be mediated by the important variable of cohesion, or the degree to
which the group members act in unison in the pursuit of common goals.
Cohesion is considered to be an extremely important social psychological
construct in the group dynamics in sport because it has a pervasive effect
on a number of interpersonal variables at the group processes stage such
as collective efficacy, cooperation, effort, and motivation. These group
processes affect individual and personal outcomes such as performance and
satisfaction.
The aim of this chapter is not to provide an extensive overview of Carron
and Hausenblas’ (1998) model. Indeed, Carron and Hausenblas have provided
an excellent and lucid text on these elements. Instead, the chapter focuses on
the social psychological processes that underlie the key elements of this
model. It will examine the effect of a specific member attribute, namely player
ability, on group structure. Factors relating to group environment will be
visited in greater detail in the section on social influence and home advantage.
The role that group norms have to play in determining group structure will be
reviewed. Much emphasis will be placed on group or team cohesion as an essen-
tial influence on group processes and outcomes such as collective efficacy and
sport performance respectively. This will provide an in-depth overview of
some salient variables involved with group dynamics research and practice.
Finally, some practical solutions in terms of team building will be presented, to
provide some pragmatic solutions to enhancing team cohesion and team
performance in sport.
Ability of team members
One assumption often made in sport is that the teams with the best individual
players are often the most successful. To some extent, this is borne out by
observation of the win–lose statistics of successful teams in a number of pro-
fessional sports. However, this is juxtaposed by the reality that some teams
perform very well with a collection of members whose individual ability may
not be of the calibre of others, indeed, they may outperform teams that have
individuals of high ability but perhaps do not function well as a unit. In such
cases, the ability of a team of high-ability individuals may be limited by group-
level variables like lack of cohesion, while the team of lower ability individuals
performs above standard because their levels of group-oriented constructs like
cohesion are optimal. The issue of cohesion will be visited in detail in sub-
sequent sections, but this illustrates that individual ability is important but its
relative contribution to team performance is moderated by group-level
constructs.
Research in social psychology has focused on problem-solving tasks to study
the effects of individual ability on group performance. Many researchers have
shown that when solving problems at a group level, group performance can be
heavily dependent on the individual resources at the disposal of the group,
and in particular the relative abilities of the group members (e.g. Devine 1999).
Reviews of studies on problem solving have found significant effects of indi-
vidual cognitive ability on group performance (Devine 1999). However, there
are occasions when the ability of the individual team members does not
always result in optimal team performance. Studies on fine motor tasks found
that individual team members did not perform as well in the group situation as
they did alone, even though the overall team performance was superior to the
individual team members’ performance (Devine 1999). This was attributed to
the lack of individual performance feedback and a subsequent social loafing
effect. Research has suggested that other factors not related to ability may
affect the relationship between individual ability on group performance such
as the nature of the task, the clarity of the goal, and the size of the group. In
addition, research has shown that tacit and task-related knowledge, skills, and
abilities may be important in determining an individual’s ability to work in a
group situation (Miller 2001). These findings indicate that the ability of indi-
vidual team members is an important artefact in team performance, but it does
not account for all the variance in team performance.
In sport, a number of studies have examined the effect of individual skills and
abilities on the effectiveness of team performance (Widmeyer 1990). In a review
of these studies, Widmeyer (1990) concluded that the relationship between
individual ability parameters and performance was consistently significant
across all studies. Suggested moderators of this relationship included attributes
of the sport that limit the productivity of the individuals in the team and team
psychological characteristics. For example, in sports like basketball, where a
great deal of coordination among players and greater cohesion is necessary for
success, ability accounts for less variance in performance than in sports

like baseball where success is less dependent upon coordination of resources
(Carron and Hausenblas 1998). Team characteristics that are likely to affect the
efficacy of individual team members’ abilities in producing optimal perform-
ance are aspects of group cohesion such as task motivation and the presence of
clearly defined performance goals and role structure within the team.
Gill (1984) suggests that given the consistent relationship between indi-
vidual ability and team performance, there is no reason to select team mem-
bers on the basis of any attribute other than their individual ability. However,
to attain maximal productivity from the players at a coach’s disposal, whose
ability may be limited, the moderator variables relating to team effectiveness
such as cohesion, team goals, and roles, particularly in sports that demand
coordinated play and strategy, need to be maximized. Indeed, coaches’ percep-
tions of players’ ability are also significant in this regard. Clearly, the coach
must make executive team decisions based on his/her players’ ability but these
will also be mitigated by the demands of the team, i.e. whether a player will be
able to contribute to the cohesiveness of the team. Research suggests that
coaches’ efficacy expectations regarding their team, which included estimates
of perceived ability as well as perceived opponents’, ability and control over
outcomes, were significantly related to performance indicators in basketball
players (Chase et al. 1997).
Group norms
One of the key constructs in theory and research in group dynamics is group
norm. Group norms describe which behaviours are acceptable and unaccept-
able, condoned and shunned, within a group. Group norms are a powerful
influence on behaviour because, according to social identity theory, a player’s
self-esteem is intrinsically intertwined with their membership of the group
and their acceptance by other group members (Tajfel and Turner 1986). There-
fore, a group member who behaves in a manner that is contrary to the
accepted norms within the group risks being vilified by the other group
members. In addition, behaving contrary to the accepted norms will also
violate the member’s own self-stereotype as a member of the ingroup. As a
consequence, their behaviour does not match up to their perceived expect-
ations of what a prototypical group member would do, and this would create
cognitive dissonance (Festinger and Carlsmith 1958). Dissonance is a feeling of
cognitive discomfort when cognition is incongruent with behaviour. Festinger
and Carlsmith state that individuals strive to reduce dissonance and create a
state of consistency or consonance between their personal beliefs and social
behaviours. As a result, non-conformity to group norms is often checked by
the individual, and the dissonant action corrected in order to restore conson-
ance within the dissenter between their role as a group member and the types
of normative behaviours that are condoned within the group.
In sports teams, a number of group norms may be considered valuable for

success. In interviews with team sports players, Colman and Carron (2001)
identified the team norms perceived to exist in teams in different contexts. In
competition, effort, support, and punctuality were considered important,
while in training punctuality, productivity, and attendance were considered
virtuous. This is hardly surprising since successful team performance is gener-
ally dependent upon attendance, and hard work is viewed as essential for
success. However, while a team may endorse a team norm that is adaptive, the
performance of the group is only likely to be successful if it is accompanied by
adaptive group properties like group cohesion. Interestingly, a norm for prod-
uctivity, a highly salient performance-related variable in team competition,
has been found to interact with the degree of cohesion viewed among team
members to produce optimal performance. Therefore, if a team norm of high
productivity is prevalent in a cohesive team, their performance is likely to be
optimal. But if the norm for productivity is low, then a dramatically poorer
performance will prevail because a cohesive group is likely to endorse the
prevailing group norm more vociferously (Colman and Carron 2000).
Coaches can use team norms to enhance and maintain unity and cohesion
within a team. Means to instil long-term group norm in naturalistic settings
like a sport teams may come from the attitude change (Eagly and Chaiken
1993) and elaboration-likelihood model (Petty and Cacioppo 1986) literature.
Techniques of persuasive communication use information giving and per-
suasive messages to evoke dissonance and precipitate changes in attitudes.
Social learning theory uses exemplars, modelling, and vicarious experience to
change behaviour (Bandura 1977). Research has suggested that such messages
may have an effect on both individuals and groups. Coaches should therefore
focus on persuasive communication and use role model sports teams with
effective group norms as examples in order to promote favourable group
norms such as productivity (Carron and Hausenblas 1998).
Collective efficacy
Bandura’s (1977, 1997) construct of self-efficacy, regarded as one of the most

important interpersonal variables in social psychology (see Chapter 5), also
operates at group level (Spink 1990). Just as self-efficacy is a set of beliefs
regarding ability to produce outcomes in a given situation, collective efficacy
reflects the beliefs shared by a group of individuals in their team’s abilities to
function as a unit and to successfully achieve outcomes in group behaviours
(Carron and Hausenblas 1998). Importantly, measurement of collective effi-
cacy not only involves measuring the belief in the ability of the team of each
individual team member but also in the consensus of these beliefs shared by
the team. Thus, a team is said to have high collective efficacy if the majority of
team members cite high levels of beliefs in the team’s ability. Zaccaro et al.
(1995) draw on Bandura’s (1977) work on self-efficacy and state that collective
efficacy comprises two sets of beliefs: collective resources and coordinative

capabilities. Collective resources are beliefs in the extent of the abilities (skills,
knowledge, strategy etc.) of the team to produce salient outcomes (Spink
1990). Coordinative capabilities are beliefs in the team’s ability to draw upon
the resources in the collective or team and act cooperatively towards the
common goal or outcome. Collective efficacy also has the property of situ-
ational specificity, that is, it reflects beliefs about team abilities in specific
situations.
A number of antecedent variables of collective efficacy have been identified:
verbal persuasion and leadership, group cohesion, group size, vicarious experi-
ence, and prior performance. Teams likely to have high collective efficacy are
those that have effective leaders who provide active encouragement, positive
reinforcement of group goals, and positive feedback on the abilities of the
team. Widmeyer and Ducharme (1997) state that accomplishing team goals
will foster collective efficacy, so it is important that team members have their
goals reinforced through feedback from team leaders and coaches. Vicarious
experiences, also a key influence on the social learning of competence in
Bandura’s conceptualization of self-efficacy, will also enhance collective effi-
cacy. Thus watching and modelling successful teams can enhance collective
efficacy. Prior experiences of success will also have a positive effect on collect-
ive efficacy. For example, Feltz and Lirgg (1998) found that perceived team
efficacy was reduced after a loss and increased after a win, suggesting that
collective efficacy fluctuates according to performance outcome. This does not
mean that teams that always lose will have low collective efficacy. Rather, it
means that there may be important mediators of this relationship. What is
more important than match or game results is the way in which the outcome is
perceived and portrayed by the team and the people around the team like
coaches and parents. Success in team sports can be defined beyond the match
or game result. It can be losing but playing well, or successfully completing a
target number of passes or rebounds, or keeping a clean defensive record.
The role of group size as a positive or negative influence on collective effi-
cacy is controversial, and yet to be fully investigated. While a larger group has
more resources that can be made available, it may also lack cohesion because it
opens up the possibility of sub-groups to form that undermines the super-
ordinate group. Furthermore, larger groups may make individual performance
and contribution more ambiguous and difficult to quantify; a performer’s per-
sonal contribution may be ‘lost’ in the collective. This may result in reduced
effort and motivation in these members when performing due to their percep-
tion that their performance is of reduced consequence to the team; a phenom-
enon known as social loafing. Finally, group cohesion exerts a powerful effect
on collective efficacy. Teams that act like a unit, are coordinated, have a clear,
shared notion of the collective goal, and are more likely to be efficacious.
Research into the role of group cohesion as an antecedent of collective efficacy
and other salient group performance-related variables will be reviewed in the
next section.
While collective efficacy has been shown to be related to the performance of

sports teams (Feltz and Lirgg 1998), studies have focused on establishing the
mechanisms by which collective efficacy determines performance. In particu-
lar, studies have indicated that group goals are implicated in the effect of
collective efficacy on performance. For example, Greenlees, Graydon, and
Maynard (2000) found that athletes with high collective efficacy and appropri-
ately set goals were able to maintain their personal performance. Bray (2004)
found that collective efficacy significantly predicted group performance in a
muscular endurance task and this influence was mediated by group goals. In
summary, these results illustrated that collective efficacy regarding future team
performance in sport was significantly related to team performance and this
relationship was mediated by the nature of the goals set by the team.
Recent studies have examined the psychological processes that determine
and induce collective efficacy in sports settings. Magyar, Feltz, and Simpson
(2004) examined the effects of constructs from achievement goal theory and
self-efficacy theory on the collective efficacy of competitive rowers. It was
found that task self-efficacy predicted individual perceptions of collective effi-
cacy and a mastery-oriented motivational climate determined collective effi-
cacy at group level. Similarly, research has shown significant correlations
between task self-efficacy and collective efficacy (Bray et al. 2002). It seems
that individual perceptions of ability as well as the appropriate motivational
climate that cultivates these senses of high ability and skill are contributory
factors to the development of collective efficacy of the team. Therefore, just as
the individual ability of team members is predictive of team performance,
individual perceptions of task-related efficacy lead to high levels of collective
efficacy.
Group cohesion
Previously we saw that the individual abilities of members in a sports team

usually determined a successful sports team (Widmeyer 1990). However, it was
also emphasized that while a team may have an abundance of resources to
draw on and a productivity potential (possible performance level) that was
very high, group-level variables may limit or even inhibit these. One construct
that may serve to moderate the relationship between potential productivity
and performance is group cohesion or team cohesion. It seems a given that
more cohesive teams are more likely to be successful in attaining optimal per-
formance, and there are many anecdotal examples of teams that have rela-
tively low levels of individual ability but high cohesiveness outperforming
teams with patently higher levels of potential productivity but low cohesive-
ness. This section will examine group cohesion in team sports by defining
group cohesion, examining a proposed model of group cohesion, studying the
effect of cohesion on performance and other psychological outcomes in sports
teams, and the predictors and determinants of cohesion.
What is group cohesion?

The study of group cohesion in sport arose from the group dynamics literature
in social psychology and is defined as the social forces that maintain the attrac-
tion between members of a group and the resistance of the group to disruptive
forces. Carron and colleagues (Carron 1982; Carron et al. 1985; Carron and
Hausenblas 1998) proposed a conceptual framework of group cohesion in
sport that accounts for the factors that contribute to the formation of group
cohesion and the effect that group cohesion has on team performance out-
comes. The model is shown in Figure 7.2. Team cohesion is viewed as a func-
tion of personal, team, leadership, and situational or environmental factors.
Personal factors reflect the shared personal psychological attributes of the team
members such as shared goals and motivation to succeed. These can be a
shared at a person level such as each member reporting high task self-efficacy
or at a group level such as the collective efficacy levels of the team as a whole
and the endorsement and maintenance of group norms such as productivity
and performance norms. Team factors include psychological variables that
operate at group level, such as group norms and collective efficacy. These are
associated with cohesion but may act alongside or interact with personal fac-
tors such as task self-efficacy. Coaches have the potential to influence the
cohesiveness of the group, and these leadership factors may influence cohesive-
ness directly and indirectly through the mediation of personal and team fac-
tors. Coaches are responsible for the pervading motivational climate and
group norms operating in the team’s training and competitive situations,
Figure 7.2 Carron’s (1982) model of antecedents and outcomes of group

cohesion in sport
Source: Carron (1982: 131)
which have been shown to affect personal variables like task self-efficacy and
collective efficacy. Finally, situational factors include the physical and func-
tional proximity of team members such as closeness on the pitch or field, but
also in terms of living and socializing. Research has shown that teams spend-
ing more time together in residential training camps are more likely to be
highly cohesive (Rainey and Schweickert 1988), and this is reflected in the
popularity of mid-, end-, and pre-season training camps in which team players
live, train, and socialize together in an environment they share.
Together, these factors, termed ‘inputs’ by Carron (1982), are viewed as
influencing task cohesion, which mediates the influence of the input variables
on group and individual outcomes or ‘outputs’. These outputs include group
outcomes such as actual performance and team-level variables such as stability,
but are also likely to affect individual performance outcomes and psychological
outcomes such as satisfaction. The next section will identify the properties of
group cohesion in sports teams and examine the relationships between the
key factors in Carron’s (1982) framework of group cohesion.
A conceptual model of group cohesion and its measurement
Carron et al. (1985) identified two related but independent dimensions of

cohesion, namely, individual attraction to the group or group integration and task
or social reasons for involvement with the group. This gave rise to a 2 × 2
framework for group cohesion, shown in Figure 7.3, in which an individual’s
Figure 7.3 Carron et al.’s (1985) 2 × 2 framework for group cohesion

Source: Carron et al. (1985: 248)
perception of team cohesion could be characterized by a profile of scores giv-

ing rise to the individual attractions to group-social (ATG-S), individual attrac-
tions to the group-task (ATG-T), group integration-social (GI-S), and group
integration-task (GI-T). These dimensions were measured on scales from the
Group Environment Questionnaire (GEQ), a standardized psychometric
instrument validated for the measurement of group cohesion in sports teams.
This scale has demonstrated adequate construct validity and reliability (Car-
ron et al. 1985). In the GEQ, the group integration items ask respondents
about the general unity of the team with respect to task or social motives for
involvement. The individual attractions to group items reflect personal
motives to maintain affiliation to the team.
Cohesion–performance relationships
The hypothesis that cohesive teams will be more successful in accordance with
Carron et al.’s (1985) conceptual model has been tested in a meta-analytic
study of 46 studies that adopted the GEQ to measure cohesiveness in sport
settings. A large average corrected effect size correlation was evident for the
group cohesion-group performance relationship. In addition, group cohesion
also predicted the performance of individuals in the team (Bray and Whaley
2001). Furthermore, ethnographic studies adopting a qualitative approach
have also supported the perception that team cohesion will result in superior
performance (Holt and Sparkes 2001). Overall, the weight of evidence from
these studies seems to suggest that teams high on cohesiveness will exhibit
better performance.
Research has also sought to examine the mechanisms that may explain the
group cohesion-performance relationship. Using a simple manipulation,
Grieve, Whelan, and Meyers (2000) found that they were able to change the
level of cohesiveness in triadic basketball teams. The manipulation success-
fully created teams of high and low cohesion, but the cohesiveness of the team
had little effect on performance. However, performance had strong effects on
team cohesion, so that winning teams had higher cohesion. This is supported
by studies that have examined the reciprocity of group cohesion on perform-
ance (e.g. Mullen and Copper 1994). These studies addressed the question:
Does successful performance engender cohesiveness or does cohesiveness gen-
erate successful performance, or both? These studies adopted a cross-lagged
correlation design in which cohesiveness and performance are measured
simultaneously at one point in time and again at a second time point. Cross-
lagged partial correlations are then calculated for the proposed reciprocal rela-
tionship while holding the effect of the time 1 variables on the dependant
variables constant. This is illustrated in Figure 7.4. Reciprocity is supported
if the cross-lagged correlations are equal, and the direction of causality
is inferred if one correlation is significantly higher than the other. A
meta-analysis of longitudinal studies on the group cohesion–performance
Figure 7.4 A cross-lagged model of the relationship between group cohesion

and performance
relationship across many competitive situations supported a direction of caus-

ality from performance to cohesion (Mullen and Copper 1994). However,
cross-lagged panel studies have found only weak support for the causal rela-
tionship between cohesion and performance in sport contexts (e.g. Slater and
Sewell 1994). More studies are required to test this relationship using robust
designs to examine the true nature of this relationship in the sport psychology
literature.
Much of the research on the group cohesion–performance relationship has
focused on team sports where the group environment involves constant inter-
action between the members of each team, and the opposing team. Some
research has focused on the role that cohesiveness plays in team sports where
there is little or no interaction between team members or that of the oppos-
ition. For example, Matheson, Mathes, and Murray (1996) compared changes
in the levels of group cohesion of interacting (lacrosse, basketball) and co-
acting (swimming, gymnastics) sports over pre-, mid-, and post-season time
points. There was a significantly greater change in the attraction to group-task
dimensions for the coacting group mid-season, suggesting that experience
across the season for coacting sports tends to foster greater cohesion, while
there is less scope for cohesion improvements in already well-functioning
interactive sport teams. In a follow-up analysis, Matheson et al. (1997) exam-
ined the effects of outcomes (winning and losing) on cohesiveness throughout
a season for interacting and coacting teams. Scores for the attraction to group-
task scales were higher on all occasions in the coacting teams, and this was
especially pronounced in losing situations. Together these results suggest that
cohesiveness seems to be influential to performance in coacting sports, and
levels may be higher than those in team sports for attraction to group-task
dimension. This is perhaps because this aspect of cohesiveness is most relevant
to sports participants who do not interact within the confines of their sport,
so they characterize their relations with team mates on the attraction and
task-related aspects of cohesiveness.
Cohesion–outcome relationships
Team cohesion has influenced a number of outcome variables other than indi-
vidual and group performance. According to Carron et al.’s (1985) model,
cohesion will have an impact on both performance-related and psychological
outcomes. Teams high in cohesiveness often report elevated levels of motiv-
ational psychological constructs at the group level such as collective efficacy
(Spink 1990), and at the personal level such as motivation (Williams and
Widmeyer 1991). Thus, it seems that cohesion is motivationally adaptive and
predicts constructs related to persistence in behaviours in interacting and
coacting sports teams, and in non-competitive situations. Further, cohesion is
also related to psychological well-being constructs in team members like goal
and performance satisfaction (Carron 1982). Analogously, teams that are high
in cohesion are also resistant to disruptive elements (Brawley et al. 1988) and
have low levels of undesirable outcomes among team members such as
drop-out (Robinson and Carron 1982), social loafing (Everett et al. 1992), and
anxiety (Eys et al. 2003). In summary, it seems that group cohesion is consist-
ent with adaptive outcomes and affective states among team members and
minimizes disruptive elements.
Predictors of cohesion
In the conceptual framework for cohesion and group dynamics, Carron (1982)
suggests that a number of influential factors will determine a team’s level of
cohesiveness: personal, team, leadership, and environmental/situational fac-
tors (see Figure 7.2). Focusing first on personal factors, research has shown that
significant variance in group cohesion could be explained by perceptions of
individual and team sacrifice (Prapavessis and Carron 1997), so it seems that
knowledge that others are foregoing personal gain and resources contributes
to the formation of a cohesive team. Turning next to team factors, prior suc-
cess and team goals are proposed to have an effect on cohesion (Widmeyer and
Ducharme 1997). Another team factor thought to influence team cohesion is
communication. For example, Sullivan and Feltz (2003) identified that three
dimensions of effective communication – acceptance, distinctiveness, and
positive conflict – were positively related to all of the GEQ dimensions. This
preliminary evidence suggests that the adaptive aspects of communication
may be an antecedent of group cohesion.
In terms of leadership factors, perceptions of leadership and coaching
behaviours have been shown to affect cohesion. Cohesive team members tend
to rate their coaches as displaying high levels of democratic behaviour, posi-
tive feedback, social support, and training and instruction behaviours, and low
in autocratic behaviours (Gardner et al. 1996). Turman (2003) identified the
coaching behaviours that promoted team cohesion such as motivational

speeches, athlete-directed techniques, team prayers and dedications, and
behaviours that undermined cohesion such as ridiculing and embarrassing
players. Finally, in terms of environmental factors, there is evidence that
group size affects cohesiveness. Widmeyer, Brawley, and Carron (1990) found
that there was a significant inverse relationship between team size and team
cohesiveness. Together these findings provide a support for influential factors
on group cohesion proposed by Carron’s (1982) framework.
Team-building interventions and group cohesion

Interventions and programmes to improve levels of group cohesion in sports
performers have been met with success. Team-building exercises usually adopt
a number of strategies to increase the key variables thought to increase
cohesiveness among teams such as collective efficacy, communication,
cooperation, and acceptance. Widmeyer and McGuire (1996, cited in Carron
and Hausenblas 1998) offer a season-long approach to team building, which
begins in pre-season training and has four phases. The educational phase in
which players’ awareness of the key principles involved in goal setting is raised
and the importance of setting team goals emphasized. The goal-development
phase is essentially a planning phase in which players form the essential com-
ponents of the team such as attack and defence and discuss strategies to
address problematic areas in their game flagged by previous seasons statistics.
These strategies then form the basis of five key team goals, which are agreed in
a plenary session. During the implementation phase, a team attains feedback
from match statistics to evaluate the effectiveness of their team goal pursuit
over a period of six games or competitions. Finally, the renewal phase is an
overall evaluation of the team’s effectiveness in addressing the team goals for
the six-game period against the statistics and an opportunity to produce a
revised level for the next six games. The effectiveness of this team building
programme has been supported with increases in team cohesiveness being
noted after completion of the programme (e.g. Voight and Callaghan 2001).
Role ambiguity, role efficacy, and team performance
Role involvement is implicated in Carron’s conceptual framework for the

study of group processes as a team factor that has a pervasive and direct influ-
ence on group cohesion, team effectiveness, and performance. According to
Carron and Hausenblas: ‘A role is the pattern of behaviour expected of an
individual in a social situation’ (1998: 157). Theorists in group dynamics have
identified two types of roles in sports teams: formal and informal. Formal roles
are those assigned to players and team personnel within the team structure
such as forward, full-back, marker, captain, and so on, and arise from team
structure. Informal roles arise from social interactions among team members
and emerge as the team develops. They have no functional purpose within a
team’s strategy for performance. Examples of informal roles are team police-
man, spokesperson, social director, and team ‘joker’. Team performance is
dependent upon team personnel performing in accordance with their formal
role in the team, and much of the literature has focused on the influence of
formal roles in sport teams. This does not mean that informal roles should be
discounted, and they may very well be significant determinants of other
important group processes.
Formal roles in sports teams affect team characteristics such as group cohe-
sion and team performance. An indirect property of team cohesion is the
degree to which a player accepts, assumes, and performs in accordance with
his or her role. Further, within teams there are likely to be roles that are more
high-profile in terms of recognition and prestige. However, team cohesion
and, indirectly, its effectiveness and productivity are dependent on all players
performing in accordance with their role, even if it is at the expense of assum-
ing a more prestigious role. Indeed, it is a property of group cohesion that
sacrifices such as these are accepted and recognized by the team members.
Beauchamp et al. (2002) have identified three key role-related constructs
which determine a team player’s performance in their role. Role performance,
a key outcome variable in the study of team player’s roles and team perform-
ance, is the extent to which an individual behaves consistently with their
assigned and expected role. Successful role performance is an important out-
come because team performance and other team properties such as team cohe-
sion are dependent on it. Role performance is influenced by role conflict, role
ambiguity, and role efficacy (Figure 7.5). Role conflict refers to the degree to
which a player experiences an inability to meet the demands of their estab-
lished or assigned role. Role ambiguity reflects a lack of understanding in the
player of the expected behavioural demands of the role, full role clarity is
necessary for optimal role performance. Role efficacy is a player’s estimate of
their ability to perform the behaviours expected for successful role perform-
ance. It has similar sources and properties of other efficacy beliefs such as task
self-efficacy and collective efficacy (Bandura 1997), but it has been found to be
conceptually and empirically distinct from these efficacy beliefs (Bray and
Brawley 2002). It is thought that high role ambiguity and low role efficacy may
lead to role conflict. Put simply, if a player is unsure of the nature of his/her
role in a team and they are not confident in their abilities to perform their role,
they are likely to experience conflict between different behavioural demands
related to their role. The effect of role conflict on role efficacy is mediated by
role ambiguity (see Figure 7.5; Beauchamp and Bray, 2001).
Beauchamp et al. studied the effect of role ambiguity on role performance in
rugby players, and proposed a triadic mediational model in which the effect of
role ambiguity on performance was mediated by role efficacy (see Figure 7.5).
Research has suggested that the effect of role ambiguity on role performance
differs for offensive and defensive players, but the defensive model supported
the mediational model. This suggested that an individual player’s lack of
Figure 7.5 Beauchamp’s model of role performance

Source: Beauchamp (2004)
clarity regarding his/her role in the team negatively influenced his/her role
performance, but this relationship was extinguished by each player’s belief in
his or her own abilities to perform the role adequately. This emphasizes the
need to promote high role efficacy, using similar strategies to promote self-
efficacy proposed by Bandura (1977, 1997), as it mitigates any potential for
role ambiguity to affect performance. In addition, role ambiguity is also related
to other adaptive and maladaptive outcomes within team sport. Research in
sport teams has shown that role ambiguity is positively related to both cogni-
tive and somatic competitive state anxiety and that clarity of scope of
responsibilities is related to satisfaction with team ability, strategy utilization,
training and instruction, team task contribution, and group integration at
both early and late season (Beauchamp et al. 2003; Eys et al. 2003). Together
these results provide preliminary evidence to suggest that lack of clarity in
individual players’ roles in sport teams can have a disruptive effect, and is
related to maladaptive outcomes and group processes like role conflict.
Several important practical guidelines arise from this research (Beauchamp
and Bray 2001; Beauchamp et al. 2002). First, coaches would do well to foster
role efficacy in team players. Mastery experiences in training are important in
this regard, to provide players with practices designed to emphasize the
importance of their role in the team. In addition, coaches should incorporate
in their team building programmes educational and information sessions

aimed at clarifying: (a) team players’ scope of responsibilities; (b) their role
behaviours; (c) their evaluative criteria; and (d) the consequences. Given the
findings that different aspects of ambiguity affect role performance in offen-
sive and defensive roles, coaches should also take into account the contexts in
which roles are likely to be fulfilled. It is also recommended that coaches be
explicit and clear in outlining why players have been given specific roles
(emphasizing specialties and specific abilities) and how players’ roles fit in
with overall team strategy. Most of all, coaches should not keep their strategies
as secret designs to add ‘mystique’ to their methods but should instead
emphasize common goals and the fact that team success is dependent upon
players fulfilling their roles.
Social influence
One of the earliest studies in social psychology was conducted in a sport con-
text and examined the effect of the presence of other people on the perform-
ance of motor skills and cognitive tasks, a phenomenon that became known as
social facilitation (Triplett 1898). Triplett’s (1898) study examined the effects
of the presence of others in cycling racing and in children participating in a
fine motor skill (‘turning fishing-line reels’). Triplett found that cyclists’ times
were faster when they raced in competition with others compared with paced
and unpaced ‘solo’ races. Analogously, children were much faster when turn-
ing reels when in direct competition with another than when on their own.
Triplett suggested that the presence of a competitor ‘served to liberate latent
energy not ordinarily available’ (1898: 507) in pursuit of the task. This pioneer-
ing study stimulated much research on the effects of the presence of other in
skills and tasks, including sports skills. However, research in the area was met
with inconsistent findings, with some studies finding a significant effect of the
presence of others on performance (e.g. Dashiell 1930) and some studies find-
ing an inconsistent or significant effects at all (e.g. Allport 1920). These incon-
sistent results led to a lull in the theoretical and empirical advances in the area
until Zajonc (1965) proposed a theoretical explanation of the effects of an
audience on behaviour that rejuvenated and intensified research in social
facilitation (Cottrell 1972).
Zajonc’s (1965) theory of social facilitation

Zajonc’s (1965) theory of social facilitation provided an explanation of the
inconsistent findings in social facilitation research. Zajonc made a number of
important observations regarding the nature of the audience and the way it
was interpreted by the performer. The theory was based on findings that organ-
isms have a set of behavioural tendencies or responses that are dominant and
override all other responses. These dominant responses tend to be well-learned
skills or response patterns that have been incorporated into an individual’s

repertoire through experience. These are opposed to non-dominant responses
that usually delineate novel, difficult, or unpractised skills. Zajonc noted from
studies in drive theory that high levels of arousal tended to result in an indi-
vidual adopting on his or her ‘dominant’ response to a stimulus and the per-
formance of the response was exacerbated, or ‘facilitated’, by the increased
arousal (Yerkes and Dodson 1908). It was hypothesized that under conditions
of coaction or in the presence of an audience an individual would experience
increased arousal that would give rise to the dominant response being dis-
played when the individual engaged in a task. If the task was a simple or well-
practised one, a social facilitation effect was observed because the dominant
response was the successful execution of the skill. However, if the task was
complex, novel, or unpractised, the dominant response tended to be one of
failure, and therefore the opposite of a social facilitation effect was observed,
or social inhibition. This theory was able to account for the observation of a
social facilitation effect in tasks that were simple, well-learnt, or instinctive
such as a basic motor skill or driving a car, while a negative social facilitation
effect was found when participants were engaged in novel or complex tasks
such as a complex motor skill or solving puzzles and mathematics problems.
Importantly, Zajonc’s notion of social facilitation considered the presence of
others as passive and that had no interaction with the performer. Thus,
Zajonc’s proposed social facilitation effect gave rise to the mere presence
hypothesis. That is, observers of an individual performing a task will evoke
arousal and the dominant response by virtue of fact that they are just in
attendance and research in social facilitation using a passive audience
corroborates the mere presence hypothesis. However, it must be noted that
‘physical presence’ may not be necessary to evoke a social facilitation effect,
information that others are performing the task co-operatively but at a differ-
ent location may be enough for a social facilitation effect to be observed
in some individuals (Dashiell 1930). This finding has been corroborated in
studies in which participants were told they were being monitored by
‘electronic’ surveillance (e.g. Aiello and Douthitt 2001).
Evaluation apprehension and social facilitation

Cottrell (1972) proposed that the mechanism behind the social facilitation
effect did not lie in mere presence alone, but in whether the individual per-
forming in front of the audience believed that the audience were judging or
evaluating their performance. This evaluation apprehension was perceived
to be the source of the increased arousal in the performer, and thus evoked the
dominant response. The mediating effect of apprehension evaluation has been
supported studies on individual and group and performers that have adopted
passive or non-evaluative, evaluative, and no-audience conditions (e.g. Bray
and Sugarman 1980). Although many of these studies have indicated that
evaluation apprehension seems to explain a great deal of the variance in the
social facilitation effect, it does not completely account for the effect and
authors have suggested that mere presence has a unique effect on social
facilitation (Bond and Titus 1983).
Bond and Titus conducted a meta-analysis of 241 studies that examined the
social facilitation effect. A significant, albeit small, average corrected effect
sizes for the influence of an audience on performance was found, accounting
for between 0.3 and 3 per cent of the variance in the difference in performance
in relation to non-audience conditions. Further analyses revealed that arousal
was only heightened by the presence of an audience if the task was complex.
The speed of a simple task was facilitated by the presence of an audience, while
complex tasks showed a social inhibition effect. It was also found that the
presence of an audience compromised the accuracy of a complex task, but had
a small facilitative effect on the accuracy of a simple task. Importantly, the
social facilitation effects were unique and were not affected by evaluation
apprehension. These findings therefore suggest that evaluation apprehension
may be a methodological artefact and cast some doubt on Cottrell’s (1972)
modification of Zajonc’s theory.
In a sport context, social facilitation effects have been confirmed in a num-
ber of studies, with results tending to lend support to the evaluation apprehen-
sion hypothesis than mere presence (Strauss 2002). Strauss suggests that mere
presence effects tend to be weak, and the effects of an audience of different
types of motor task tend to be inconsistent. Several studies have supported the
role of an evaluative audience or co-actor on sport performers. For example,
Smith and Crabbe (1976) found that active experimenter participation
increased the learning of a balancing motor task in participants compared
with passive or no experimenter conditions, supporting the evaluation appre-
hension hypothesis. In a novel adaptation of the social facilitation experi-
ment, Paulus et al. (1972) studied the performance of novice and experienced
gymnasts in the presence of an audience. One group of gymnasts in each
performance level were forewarned about the presence of an audience while
others were not. Those who were not warned exhibited high quality perform-
ances compared to the forewarned gymnasts who incurred a decrement in
performance. The authors suggested that the anticipated evaluation of the
audience evaluation was responsible for the decrement in performance and
undermined the dominant response in the skilled performers. In the presence
of an evaluative audience, Bell and Yee (1989) found that karate experts’ per-
formance was unaffected in terms of the speed and accuracy of a kicking task
relative to a solo condition. However, unskilled karate performers reduced the
performance of their kicking to avoid errors, indicating that the presence of an
audience has a slight social inhibitory effect causing a speed–accuracy trade-
off. This supports the findings found in previous studies in which complex
tasks for novice performers tend to incur an accuracy decrement (Allport 1920;
Bond and Titus 1983).
While these studies found effects largely consistent with Zajonc’s (1965) and
Cottrell’s (1972) hypotheses, Kozar (1973) found no difference in supportive,
non-supportive, or no audience conditions in learning a gross motor skill.

Kozar also found no differences in the learning performance of high- and low-
anxious participants, negating the potential moderating effect of evaluation
apprehension. However, it must be stressed that this study focused on learning
rather than performance and used a gross, relatively simple motor task, not a
complex task.
Role of cognition in social facilitation effects

Social cognitive constructs may also have important effects on social facilita-
tion. Research examining cognitive variables on the social facilitation effect
has given rise to explanations based on attention conflict in which the distrac-
tion of an audience competes for cognitive resources or attention from the
individual (Baron 1986). For example, Hall and Bunker (1979) found a signifi-
cant interaction between social facilitation and the social cognitive variable of
locus of control. Participants engaged in a novel task that reported an internal
locus of control did not exhibit a decrement in performance before an audi-
ence while those reporting an external locus did. Similarly, Forgas et al. (1980)
found social inhibition effects for expert squash players and social facilitation
effects for novice squash players when playing as a pair before an audience. To
explain this, Forgas et al. suggested that under solo conditions the match was
viewed as a straightforward competition, so it was acceptable for players to
play to their ability. However, when playing before an audience, the expert
players may have felt the need to curtail their performers as there was an
increased need to be seen to be participating co-operatively. These results
suggest that the social facilitation effect is more complex than the premises
put forward by Zajonc (1965) and Cottrell (1972) over four decades ago. Future
research on social facilitation will account cognitive and address hypotheses
relating to the mechanisms involved in this complex phenomenon.
Social loafing
While the social facilitation literature provides a theoretical explanation for
the conditions under which co-actors may facilitate or debilitate performance,
researchers have identified an additional outcome of performance of a task in a
co-acting or collective situation in which an individual exhibits a clear motiv-
ational decrement or performance loss. This process is called social loafing,
and individual performance of a sport-related task in a team environment is an
ideal naturalistic setting to examine this phenomenon. The social loafing
effect was observed in early research in group influence on individual per-
formance toward a collective task by Ringelmann (1927, cited in Latané et al.
1979) in which people exhibited worse performance when engaging in a task
in co-operation with others than when they were working in solitude. In
essence, these observations were the opposite of the social facilitation effects
in co-acting conditions (working alongside others rather than in front of an
audience) observed in the studies of Triplett (1898), Allport (1920), and

Dashiell (1930), and this has since become known as the Ringelmann effect.
Since, researchers have argued that this effect should be studied in conjunc-
tion with the social facilitation effect to provide a unified theory of social
facilitation (Aiello and Douthitt 2001).
One of the first studies to integrate these findings and offer an explanation
for the two seemingly contradictory social influence effects was conducted by
Jackson and Williams (1985). Using the completion of computer mazes, the
research indicated that individual performance was optimal when working
collectively on difficult mazes and when working individually on simple
mazes. Importantly, only when the performance of the group on the maze was
readily identifiable and individual performance was not, were individual per-
formances on the maze inhibited. When individual performance was
distinguishable from that of the group, individual performance when working
collectively was improved. Sanna (1992) provided evidence that high self-
efficacy moderated the social loafing effect. Theorists have also suggested that
the social loafing effect was observed when performers did not receive any
feedback regarding their own performance or the group performance. In sum-
mary, social loafing seems to arise when high-self-efficacy individuals working
toward the collective task are aware that they were not being evaluated and
were not given feedback relating to this performance.
In sport, social loafing research has focused on identifying the factors
responsible for the reduced motivation in individuals competing in team
sports for which the collective rather than the individual is held responsible
(C. J. Hardy 1990). In keeping with theory, identifiability, defined as the extent
to which an athlete’s performance is known to him or her and others, of an
individual within a team, moderated the social loafing effect in collegiate
swimmers (Everett et al. 1992). The sport ability of an individual performer in a
team relative to his/her teammates has also been shown to magnify the social
loafing effect, suggesting that perceptions of incompetence may account for
motivational decrements probably because of the athlete’s perception that
they would have little impact on collective performance (Hardy and Crace
1991). This effect was also seen in collective team situations when faced with
opposition that was vastly superior to themselves (Heuze and Brunel 2003).
Recent research has focused on group-centred social cognitive variables that
may affect social loafing. Lichacz and Partington (1996) found that partici-
pants with high collective efficacy and with prior experience of the group
exhibited fewer inhibition effects due to social loafing. Indeed, research has
indicated that the cohesiveness of the team is negatively related to the social
loafing effect (Everett et al. 1992), however, it must be noted that team cohe-
sion and collective efficacy alone may not negate the social facilitation effect.
For example, Hardy and Latané (1988) found that members of established
teams performing intrinsically motivating (i.e. interesting) tasks exhibit social
loafing, an effect that could be explained by complacency in the absence of
evaluative feedback. Interestingly, when athletes were given information
about the social loafing effect before group competition, there was still evi-
dence of a decrement in motivation and performance in a team competition
compared with a solo condition (Huddleston et al. 1985). This suggests that
prior knowledge does not interfere with situational characteristics that may
influence social loafing such as presence of evaluative performance feedback,
collective efficacy, and competence levels.
Future directions in social facilitation research

Social facilitation research has met with considerable success in identifying the
phenomenon, providing explanations for the effect, and identifying the dif-
ferent conditions and influences on performance under conditions of social
facilitation. However, some have criticized the lack of an integrative or unified
approach to the different explanations and facilitating conditions in the
extant literature (Aiello and Douthitt 2001). Aiello and Douthitt claim that
this lack of a framework limits research in the area due to a lack of clearly
defined hypotheses. These authors suggest that theories of social facilitation
could benefit from further clarification of some key tenets or features of the
effect: (1) the definition of social facilitation; (2) identification of the salient
dimensions of social facilitation; and (3) the predicted effects of the presence
of other under a given set of situational and psychological conditions. Such an
integrative approach may assist in directing future investigations toward the
appropriate research hypotheses necessary to fill gaps in theory.
An integrative model was presented by Aiello and Douthitt in an attempt to
draw together the various theoretical strands and research findings into a con-
ceptual framework to explain social facilitation effects. The integrative model
is presented in Figure 7.6. The model is presented with individual influences
on social facilitation at its heart, indicating that intrapersonal variables medi-
ate the effects of other, more extraneous, influences on performance factors
and outcome variables portrayed at the base of Figure 7.6. Three main areas of
extraneous influences on the individual or personal factors that are associated
with social facilitation effects are identified: (1) situational factors; (2) presence
factors; and (3) task factors. Situational factors represent the characteristics of
the climate in which the audience is presented, whether they are seen or
heard, how close they are, whether feedback on the situation is available, and
the structure of the motivational climate. A set of presence factors map the
features of the audience likely to have an effect on social facilitation, namely,
the type of presence or make-up of the group, the role of the audience, how
salient they are to the individual and their relationship to them (e.g. parents
vs. strangers), and the length of the presence. Then the model accounts for
the task factors that feature most prominently in social facilitation research:
the difficulty (simple vs. complex) and type of task (cognitive vs. motor),
and the time demand of the task.
These variables are proposed to affect the individual factors. Central to this
are the perceptions and reactions of the individual to the extraneous factors.
Figure 7.6 Aiello and Douthitt’s integrative model of social facilitation
Source: Aiello and Douthitt (2001: 74)
These include evaluation pressure, social comparisons, self-awareness or self-

presentation, whether the extraneous factors are a challenge or a threat, or
invasive. These initial perceptions and reactions are envisaged to mediate the
influence of the extraneous influences on subsequent reactions such as physio-
logical arousal, cognitive conflict, self-monitoring, and self-efficacy. Personal-
ity and other individual characteristics such as motivation, intelligence and
ability also play a moderating role in the effects of extraneous and situational
perceptions and reactions on subsequent reactions and performance. It is pro-
posed that subsequent reactions, especially physiological arousal, will affect
the performance factors, many of which have been dependent variables in
social facilitation research: speed and accuracy. Aiello and Douthitt (2001)
claim that other responses to social facilitation such as aggressiveness, cooper-
ation vs. competition, and effects of social facilitation of the performance of
others need to be studied in greater depth. While a number of these aspects
have been studied, increases in the sophistication of the methodology in social
facilitation research may provide future studies that incorporate more elem-
ents of this proposed model. This is particularly pertinent in examining the
effects of social facilitation in sport, as the findings to date have been
inconsistent.
Home advantage
One of the most pervasive, consistent, and oft-cited social effects in sport
psychology is the home advantage. The effect is so well observed that it has
been elevated into the lore of sports statistics as a principle rather than mere
theory. Further, the effects of home advantage have been the subject of intense
research interest by statisticians, sport scientists, and sport psychologists. Sport
psychologists’ interests stem from the view that the home advantage is a
largely psychological phenomenon, particularly if physical parameters
between two teams are considered to be relatively equal and any differences
largely insubstantial. Since competitive sport is often undertaken by com-
petent individuals, both arousal and cognitive explanations of social facilita-
tion lend credence to the expectation that team players at home will have their
dominant response reinforced and enjoy a facilitative effect on performance.
However, as the previous section testifies, situational, presence, task, and indi-
vidual factors can alter these effects. The aim of this section is to provide an
overview of the home advantage in sport and to examine the theories that
have been put forward in applied social psychology to explain these effects. In
particular, the premise that crowd influence will have the most pervasive effect
on the home advantage in accordance with social facilitation theory will be
addressed.
Prevalence, methods of investigation, and the home advantage

The home advantage effect has been noted in a variety of sports including
American football (Schwartz and Barsky 1977), baseball (Schwartz and Barsky
1977; Courneya and Carron 1991), basketball (Schwartz and Barsky 1977;
Silva and Andrew 1987; Varca 1980), cricket (Jones et al. 2001), cross-country
running (McCutcheon 1984), field hockey (Russell 1983), ice-hockey
(Schwartz and Barsky 1977; McGuire et al. 1992; Agnew and Carron 1994),
soccer (Brown et al. 2002), alpine skiing (Bray and Martin 2003), and wrestling
(Gayton and Langevin 1992). These studies have indicated that teams win
between 52 per cent and 88 per cent of their home games (Schwartz and Barsky
1977; Courneya and Carron 1991; Gayton and Langevin 1992; Bray 1999).
Studies have also indicated that teams in outdoor sports such as American
football and baseball seem to have only a marginal home advantage, while
teams in indoor sports such as ice-hockey and basketball seem to enjoy more
of a home advantage (Schwartz and Barsky 1977). However, there has been
debate as to whether the home advantage occurs all of the time. Some studies
have shown that in championship series and play-off matches in baseball and
basketball, the final games at the end of the season that decide champion-
ships, home teams won the majority of early games in the series, but lost
the majority of the final games (Baumeister and Steinhilber 1984). As a con-
sequence, the hypothesis has been disputed and is a topic of debate among
researchers in the area (Baumeister 1995; Schlenker et al. 1995), suffice to say
that on some occasions, perhaps in high pressure matches, a home advantage
may be negated or even overturned. The next section will visit the methods
and theories proposed to explain these inconclusive findings.
A number of methods have been adopted to examine the home advantage
effect. Schwartz and Barsky were among the first to adopt formal statistical
procedures and demographic information to explain the home advantage in
indoor (ice-hockey and basketball) and outdoor (American football and base-
ball) sports. The authors used official published archival statistics from all
league teams to explain the win and loss ratios on home compared to away
grounds. Many studies have since adopted this methodology (e.g. Varca 1980;
Gayton and Langevin 1992; McGuire et al. 1992; Agnew and Carron 1994).
Other studies have examined the home advantage from the perspective of
individual team statistics over the course of the season rather than league
average (Bray 1999). Some studies have used observational techniques by
trained observers to provide additional information regarding performance
not based on published statistics, such as Greer’s (1983) observation of crowd
protest over the course of the season. Studies have also used these obser-
vational techniques to collect data from televised sports matches (Salminen
1993). Finally, the most challenging studies from an empirical point of view
are those that collect additional data from players and participants in the
competitions throughout the course of the season (Bray and Martin 2003;
Neave and Wolfson 2003). These studies are usually data-rich, but are very
expensive and time-consuming to conduct and are often plagued by the prob-
lems typically experienced in applied social psychological research such as
small sample sizes and missing data due to participant drop-out. Together,
these methods of investigation have provided converging evidence for the
influential factors and consequences of home advantage in sport.
Explanations and theories of home advantage

The home advantage has been attributed to a number of factors, some demo-
graphic such as age, others situational such as size of crowd and distance to
venue. More sophisticated theories have been proposed to explain the mech-
anisms behind the influence of these factors, and how they affect the social
psychological processes that underlie performance. Of all these factors, it
seems that the make-up of the crowd and its effects on the arousal levels and
perceptions of the home team players are the most pervasive. This section will
examine a variety of theories and factors thought to influence and explain the
home advantage.
Territorial/ethological theories
One of the more controversial but compelling explanations for home advan-
tage is based on ethological observations in organisms relating to marked terri-
tory and their defence of it. Russell (1983, 1993) argues that organisms are
more fervent in their defence of their territory because it represents their live-
lihood for breeding and feeding. Therefore, organisms attach greater value in
the defence of contested territory when it is their own because they have a lot
to lose. Futhermore, this has an evolutionary advantage as organisms that
cannot defend their territory successfully are selected out of the species and
there is evidence in support of this ‘home advantage’ in animals (e.g. Rajecki et
al. 1979). According to Russell (1993), the theoretically greater fervour exhib-
ited in the defence of their home territory could be attributed to a greater
display of aggressive behaviours by the home team and this has been sup-
ported in some studies (e.g. Varca 1980). Further, recent evidence has provided
some additional evidence to support a territorial explanation of home advan-
tage in ‘human competitive encounters’ like sports competition. Neave and
Wolfson (2003) found that testosterone levels were higher in soccer players
before their home games, and especially higher when facing ‘extreme’ rivals
compared with ‘moderate’ rivals. The authors suggest that elevated hormone
levels may be concomitant with a greater innate propensity to defend home
territory and display aggressive behaviours. However, sport research has not
conclusively supported increased aggressive behavioural displays in home
teams. While ethological and territorial explanations may potentially account
for the home advantage effect, Russell (1993) claims that these theories pro-
vide a more philosophical than empirical explanation for the home
advantage.
Crowd size, density, and hostility

As proposed earlier, one of the key explanations offered for the home advan-
tage is the presence of a partisan home audience or crowd during play. A
number of studies have examined the effect of crowd size and density on
performance of teams at home. As is almost exclusively the case in team sports
particularly at elite level, the majority of home spectators support the home
team, even in so-called derby games. Therefore it is reasonable to assume that
the size of the home audience is a reflection of the number of home supporters
present. Schwartz and Barsky (1977) were among the first to examine the effect
of crowd size on home advantage in American football, basketball, baseball,
and ice-hockey, and their results found that winning percentage of home
teams increased in proportion to crowd size. In addition, Russell (1983) stud-
ied the effect of crowd size on indicators of performance (e.g. goals scored) in
home teams in field hockey, and found no significant correlation between
these for the home team. However, a significant negative correlation between
crowd size and performance parameters of the away team was found. This
suggests that the home advantage may be more due to the inhibitory effect of
crowd size on away performers. This supported findings that suggested the
home advantage was really an ‘away disadvantage’ (Silva and Andrew 1987).
Some researchers have hypothesized that it may not be the size of the crowd
per se, but its density. Agnew and Carron (1994) studied the effect of crowd
density on home performance of junior ice-hockey teams and found that dens-
ity was significantly related to the home advantage, but only accounted for a
small percentage of the variance in winning percentage. These results lend
some support to the hypothesis that the density rather than size of the crowd
matters in home advantage, while the size of the crowd may be implicated in
visitor disadvantage.
Given the finding that visitors may be adversely affected by the audience,
researchers have sought to examine whether the nature of the support offered
by the crowd has a detrimental effect on away teams. Greer (1983) examined
the effect of spectator protest (‘booing’) on the performance parameters of
home and away teams in college basketball games. Results indicated that the
performance gap between the home and away teams increased dramatically
during episodes of protest. However, although there were marginal increases
in the performance of the home team, it was decreases in the performance of
the away teams (greater number of fouls committed and reduced overall per-
formance) that contributed to the increase in the performance advantage
already enjoyed by the home team.
By way of explanation, Silva (1979) suggested that hostile and angry
behaviours exhibited by the crowd towards players may disrupt concentration
and inhibit execution of motor skills. This ‘distraction’ effect is consistent with
the social facilitation and social loafing literature that suggests that the audi-
ence draws cognitive resources and attention that should be committed to the
performance of the task (Baron 1986). However, other studies have not sup-
ported a negative effect of spectator protest on visiting team performance.
Salminen’s (1993) study of basketball, ice-hockey, and soccer teams revealed

that, contrary to hypotheses, when the home crowd supported the visiting
team, the home team also scored more points/goals. These findings question
the conclusion that away teams suffer a decrement in performance before a
hostile crowd, and suggest that any display of support, regardless of direction,
is likely to have a positive influence on home team performance. In sum,
research examining the size, composition, and nature of the crowd on the
home advantage makes intuitive and theoretical sense. A partisan home sup-
port, for example, that is spread out across a large stadium is less likely to
facilitate the performance of the home team because support is likely to be
diluted, compared with an audience of comparable size that is closely packed.
Explanation of the facilitative and inhibitory effects in home advantage
research can be explained by social facilitation theories. However, the research
suggests that the effects of the crowd on home advantage or visitor disadvan-
tage demands a more complex explanation than Zajonc’s (1965) theory of
mere presence. Some researchers suggest that the arousal effect of the presence
of a crowd may evoke more ‘assertive behaviours’ in athletes as suggested
previously in ethological explanations of home advantage. Varca (1980) sug-
gested that the home advantage evoked the types of behaviours consistent
with aggressive but performance-enhancing play. Examining these ‘functional
assertive behaviours’ in school basketball teams supported this hypothesis.
Further, visiting teams exhibited a significantly greater number of ‘dys-
functional assertive behaviours’ such as fouls, a finding consistent with the
visitor disadvantage. Varca supposed that the presence of these aggressive
behaviours partly explained the home advantage effect. McGuire et al. (1992)
corroborated these results in league ice-hockey players. They found that
aggression was advantageous in home players but disadvantageous to away
team players. One possible mechanism for this aggression is the increased
arousal evoked by the partisan crowd (Sanna 1992).
Sport type
Schwartz and Barsky (1977) identified indoor sports like basketball and ice-
hockey as being most influenced by home advantage, while outdoor sports
had fewer effects. This may be due to the proximity and density of crowds in
indoor arenas. However, there has been little research to investigate the
parameters that might affect the extent of the home advantage across different
sports. However, some interesting research has examined whether the home
advantage is as pervasive in individual sports as it is in team sports. Studies
have found a significant home advantage in individual sports like wrestling
(Gayton and Langevin 1992). Results were attributed to increased feelings of
security and dominance and attributed to a ‘prior residence effect’ in which an
initial resident in a geographical area has a social dominance advantage over
an intruder. However, others have found no home advantage was found in a
study of individual downhill skiiers (Bray and Martin 2003). The studies of
home advantage in individual sports are limited, and further research will
reveal further trends. It is possible that the extent of the home advantage will
vary according to the sport itself as observed across different team sports.
Home venue familiarity

Competitor’s familiarity with the site and facilities of the home venue may
account for the variance in performance over home and away fixtures. In a
novel study, Loughhead et al. (2003) investigated the effect of a change in
home venue on home performances of professional basketball, hockey, and
soccer teams. They found no differences in the teams’ performance immedi-
ately after relocation and the games following relocation compared to before
location. However, when the quality of the team was introduced as a moder-
ator, it was found that high quality teams were unaffected by the move, while
low quality teams benefited, in terms of performance, from the relocation. A
possible reason for this may be that the improved quality of the facilities for
the lower quality teams may outweigh the detrimental effects of unfamiliar
surroundings, possibly because the gap in standard between the old and new
facilities was greater for the lower quality teams. Jones et al. suggested that
familiarity with facilities was a plausible explanation for a home advantage
effect found in cricket teams, a sport in which audience influence was min-
imal. One possible explanation for the effect of familiarity on home advantage
may be that athletes are more likely to perform better when performing the
skills in the environment in which they originally learnt them, as opposed to a
novel one (Russell 1993).
Distance and travel

One of the most well-researched factors thought to contribute to the home
advantage is the distance travelled by the away team. Early research supported
an effect of distance travelled on home advantage (Schwartz and Barsky 1977).
However, the effect of distance travelled alone provides only a partial explan-
ation for the effect of distance on home advantage, and there are a number of
distance-related variables that may influence home advantage. Courneya and
Carron (1991) found only a very small effect for distance travelled, presence of
home travel for either the home or the away team, and time of season (number
of games into season) on home advantage in baseball players. The effect of
further distance-related parameters, including number of time zones crossed
by the visiting team, on home advantage in professional ice-hockey players
was investigated by Pace and Carron (1992). While the number of time zones
crossed and preparation time were inversely related to visiting team perform-
ance, these factors also only accounted for a small amount of variance in the
home advantage. Converging evidence from these studies suggests that
there is a minimal, albeit significant effect of visiting team travel on home
advantage, but the variance this artefact explains in performance is small.
Recently, an interesting new perspective on the effect of visiting teams
travelling across time zones on home advantage has been proposed and has
promise to account for inconsistencies in research on travel effects. Steenland
and Deddens (1997) used archival data from professional United States basket-
ball (U.S. National Basketball Association) and American football (U.S. Ameri-
can Football League) teams to examine the effect of players’ circadian times of
day on home advantage. As hypothesized, they found that teams based on the
West Coast of the United States playing away games at East Coast locations
performed significantly better in the those particular away games because they
were playing at a time of day closer to their theoretical physiological peak.
These results suggest that the effect of distance and travel across time zones per
se may not fully explain the effect of visitor team travel on home advantage. It
is possible that these results could be extrapolated to travel of international
visiting teams from west to east, and suggests that an adequate recovery period
should be included for teams travelling from east to west.
Referee bias
Opposition supporters to successful teams often state, albeit slightly tongue-
in-cheek, that their opponents seldom get sanctioned with penalties (e.g.
fouls, free kicks) against them and often get penalties in their favour from
match referees and officials when playing at home. Referee bias has therefore
been cited as one factor that may contribute to home advantage. In his study
on crowd protest, Greer (1983) did not attribute the significantly greater num-
ber of fouls incurred by visiting teams after protest episodes to referee bias,
rather, this seemed to be due to an overall decline in performance not just
increased incidence of fouls. Recently, however, Nevill et al. (2002) conducted
an experimental study in which soccer officials assessed videotaped game situ-
ations with and without crowd noise. Results showed that the officials watch-
ing the games with crowd noise were more uncertain in their decision-making
and awarded significantly fewer fouls to the home team. These results sug-
gested that crowd noise had a pervasive effect on referee bias, but questions
still remain over the ecological validity of the experiment and the partisan
nature of the crowd noise. Interestingly, in a study on county cricket, a sport
with virtually no crowd influence, Jones et al. (2001) found no evidence of
match umpire decisions for home and away teams, which may lend further
weight to the influence of the crowd in biasing decisions. Another factor that
may affect match official or referee bias is the quality and profile of the team
members. In a study on a professional basketball team, Lehman and Reifman
(1987) also found that while officials awarded equal number of fouls to ordin-
ary, ‘nonstar’ players in home and away teams, the well-known ‘star’ players
on the home team incurred significantly fewer penalties than the away team
players.
A number of explanations for referee bias have been put forward. One prom-
inent effect observed in social psychology known as interpersonal bias, often
referred to as the Pygmalion effect, may account for referee bias. Referees may
unconsciously bias their decisions toward the home team and star players
because they have an expectation that the home team and star players will
perform better. Evidence for this is given by two studies on judging officials in
gymnastics and figure skating respectively. Scheer and Ansorge (1979) studied
the effect of prior expectation on gymnastics judges’ decisions. In gymnastics,
it is commonly known that coaches in each team always send their gymnasts
out to the apparatus in reverse order of ability, as judges have an expectation
that the best gymnasts will compete last in the rotation. Scheer and Ansorge
required gymnastics judges to score videotaped Olympic qualifier routines, but
reversed the order of presentation of the routines such that the first gymnast in
each team appeared last, and so on. The judges’ scoring was found to be biased
to favour the last performer, even though they were the least skilled in each
team. Interestingly, this effect was found to be moderated by the personality
traits of the individual judges. Judges who reported having an external locus of
control and thought that events were subject to extraneous factors and luck
were more likely to be biased toward the order of the competitors than judges
with an internal locus of control. This evidence suggests that biased decisions
may therefore be a function of the psychological profile of the referee
themselves rather than their expectations alone. In a study on figure skater
judges, Findlay and Ste-Marie (2004) found that judges gave significantly
better marks to skaters who were known to them, suggesting an influence of
reputation bias. Together this evidence suggests that prior expectations and
reputation are likely to bias decisions of referees and judges rather than
audience effects per se.
Home disadvantage
In the introduction to this section, we visited Baumeister and Steinhilber’s
(1984) experiments on the paradox of the home advantage in high pressure,
last-game situations. In situations of high audience expectation it seems that
the home advantage is negated. A re-analysis of the findings and addition of
new data suggested that the disadvantage effect was not as pervasive as previ-
ously cited (Schlenker et al. 1995). However, Baumeister (1995) insists that the
effect still exists even though it may be weaker than previously stated. Explan-
ations for these effects come from the social facilitation literature and the
potential for the high-pressure, anxiety-provoking situation to interfere with
the ‘dominant response’. The increased expectations of the home crowd may
inhibit performance for several reasons: (1) increased arousal due to the audi-
ence presence may cause a distraction or an inability to focus attention on the
task at hand (Baron 1986); (2) a heightened self-awareness causes players to
move their attention away from the appropriate cues for action (Baumeister
1984); or (3) the player may focus too greatly on the execution of well-learnt
skills and this exertion of cognitive control results in disruption (Baumeister
1984); or (4) fearing failure, the competitor becomes self-aware and again
experiences distraction from appropriate cues. It seems that the ‘champion-
ship choke’ is a phenomenon that is difficult to explain, but it likely to have
a pervasive effect on the expectations of home teams in high pressure
match situations. Home teams may therefore not assume or rely on any home
advantage in last-game or championship series.
Future directions in home advantage research

The home advantage effect is found to be a robust and consistent effect across
a number of sports. Investigations into the causes of this effect have mainly
been centred on the influence that the audience has on the players of the
home team, and theoretical explanations have been based on models of social
facilitation, the arousing effects of an audience, and their effect on social cog-
nitive variables like self-efficacy and self-confidence. Other explanations have
been proposed including travel fatigue, home venue familiarity, referee bias,
sport type, team quality, and age. However, much of the research for these
effects has not led to consistent results, suggesting that one universal model or
theory is unlikely to account for all of the variance in performance between
home and away games. This is just an illustration of the complexity of the
phenomenon and regression models to explain home advantage may exist for
each different sport. One criticism that can be levelled at the home advantage
research is their bias towards culturally-specific sport disciplines. American
football, baseball, and, to a lesser extent, basketball and ice-hockey tend to be
sport disciplines played predominantly in North America, and the majority of
the studies have come from this region. Studies on European and international
sporting disciplines are becoming more prevalent (e.g. Jones et al. 2001; Brown
et al. 2002), but further research needs to be done in order to examine the
extent of the home advantage in different cultures and sports disciplines. Fur-
ther, future research needs to examine the effect of the home advantage on
social cognitive variables in order to form a comprehensive theory for the
home advantage.
Suggested reading
Aiello, J.R. and Douthitt, E.A. (2001) Social facilitation from Triplett to electronic
performance monitoring, Group Dynamics, 5: 163–80. Recent review that charts the
progress in social facilitation research and presents a new and exciting conceptual
model.
Baumeister, R.F. and Steinhilber, A. (1984) Paradoxical effects of supportive audiences
on performance under pressure: the home field disadvantage in sports champion-
ships, Journal of Personality and Social Psychology, 47: 85–93. Baumeister’s fascinating
classic paper on the paradox of the home disadvantage.
Carron, A.V. and Hausenblas, H.A. (1998) Group Dynamics in Sport, 2nd edn. Morgan-
town, WV: Fitness Information Technology. The leading text on the influence of
group processes in sport.
Schwartz, B. and Barsky, S.F. (1977) The home advantage, Social Forces, 55: 641–61. The
paper that started it all; the citation classic on home advantage in sports teams.
Summary
• Social psychologists define a group as a collection of people who interact

with each other, have a common goal, have a clear structure, and have
shared communication processes and behavioural outcomes. Sports teams
are one example of a group and the interactions and other processes
operating within the team have a pervasive effect on the behaviour of the
individual members that make up the team.
• Carron and Hausenblas’s (1998) conceptual model provides a framework
for understanding group processes in sport. In the model, group structure is
determined by team norms and the roles of the group members. Group
structure influences important outcomes in sports teams such as team
goals, collective efficacy, and cooperation. Group cohesion mediates these
relationships.
• Group cohesion is an important construct in group processes and has
profound effects on team outcomes and performance.
• Role ambiguity is a threat to cohesion and coaches should promote clarity
and efficacy in sports performers regarding their role in a sport team.
• Social facilitation describes the effects of the mere presence and perceived
evaluative apprehension of other people such as an audience or co-actors on
sport performance. Social facilitation is moderated by degree of evaluation
apprehension.
• The lack of personal agency, role clarity, self-efficacy, and clear feedback on
individual performance can give rise to reduced effort and performance
level in team members, known as social loafing.
• The home advantage for sports teams is influenced by the size, density, and
hostility of the audience, reputation of performers in team, age of team
members, and home facility familiarity. There is research to suggest that the
home advantage effect is actually due to impaired performance by visiting
teams, known as away disadvantage.
• Home advantage tends to be least affected by distance travelled to match
and referee bias. Further, the home advantage seems to be overturned in
matches that lead to athletes performing at their optimal time of day and
in the later games of championship series matches.
8
Aggression and crowd violence
Aggression in sport
There is perhaps no more reprehensible and harrowing sight than groups of

athletes fighting each other over an infraction, a point, a disputed tackle, or
some other relatively minor penalty during a sports game or competition. To a
neutral observer, the cause of the melee may seem trivial, almost insignificant,
and therefore unjustified. However, move closer to the action, and the fans,
crowd, or spectators can all be seen to hold strong views about the matter and
may even endorse such aggressive action (Russell 1979). Further, coaches often
side with their players in matters of dispute, invariably supporting their teams,
the politician ‘inside’ them often the only thing preventing them from joining
the fracas in defence of their players. The players themselves are often united
in their aggravated feelings and provocative gestures towards the opposition,
but are often observed stopping short in participating in actual aggressive
behaviours. Alongside the players, the coaches, and the spectators, the media
watches with morbid fascination, selecting, recording, and presenting the
events that ensue and their intricate details prominently in their headlines
often with a ‘holier than thou’ sense of moral outrage.
Why do players resort to such violent conduct when they know there are
clear penalties for doing so? Why do coaches defend their players when they
can clearly see that such aggression is counter-productive and morally dubi-
ous? And what role do the media play in perpetuating such aggression? Do the
media present such hostilities in such sensationalist fashion through an exag-
gerated sense of moral obligation? Or is it because the consumers of such
media are captivated by such aggression? This chapter will seek to address
these questions and evaluate the social psychological research into aggression
and crowd violence in sport. Definitions of aggression and key social psycho-
logical theories will be reviewed, and several important studies of aggression in
sport will be visited. Crowd violence and a specific form of fan violence,
hooliganism, will also be reviewed. The chapter will conclude that aggression
194 Aggression and crowd violence
and crowd violence are frequently occurring phenomena in sport and are
affected by situational, interpersonal, and inter-group factors.
Definition of aggression
Aggression is an often misinterpreted construct in social psychology. It is
commonly thought that non-descript violent shouting and other outbursts
such as fist-shaking and provocative gestures at the opposition or equipment
abuse such as throwing a tennis racket on the ground in disgust constitute
aggression. However, in terms of a formal definition, these violent or aggra-
vated behaviours are not aggression. Aggression is a set of behaviours that are
likely to, or have the potential to, cause harm to others, and are intended to
cause harm, and are therefore goal-directed (Berkowitz 1993). Just as gesture
and equipment abuse tend to lie outside the realm of aggression, some
behaviours in sport are often misinterpreted as aggression. Examples are assert-
ive behaviours which are merely actions that enhance the competitiveness of a
team by imposing dominance over the opponent, but without injuring him,
her, or them. Often coaches want players to be more ‘aggressive’ in their
approach to their opponents. What they mean is that they want their players
to display more assertive behaviours such as hard tackling in soccer and rugby,
or hitting or pitching the ball at the opponent in tennis and baseball respect-
ively. Such behaviours are physically vociferous, but have a purely pragmatic
goal toward being successful rather than deliberately injuring an opponent.
The definition of aggression can also be considered relative to what is tolerated
by rule-producing governing bodies and what is not. For example, apparently
provocative, ‘aggressive’ behaviour such as the Haka performed by the New
Zealand All Blacks rugby teams prior to matches is endorsed by the authorities
of the game. However, other violent acts are punished, albeit leniently, but for
other reasons other than their causing injury to others, such as racket or verbal
abuse in tennis. These acts are not aggression, so while aggression is usually
against the regulations of given sports, some violent infringements in sport
cannot be defined as aggression unless they are coupled with the defining
criteria listed previously. To summarize, aggression is a set of behaviours that
are intended to harm others, but are not to be confused with nondescript
verbal outbursts or violent conduct towards inanimate objects and assertive
behaviours in sport.
A number of authors have identified two types of aggression that can oper-
ate in sport contexts, hostile aggression and instrumental aggression (Silva 1980;
Berkowitz 1993). Hostile aggression has the primary goal of harming or injur-
ing another person or player. It is often a response to provocation, coupled
with high emotional arousal, and an accompanying negative emotion often
appraised as sourced from the focus of the aggressive act (see Figure 8.1). In
hostile aggression, aggressive acts supersede other goals of the sport such as
scoring points and will distract from the aggressive players’ role in the team
(see Chapter 7, Beauchamp and Bray 2001). For example, in April 2001, Roy
Aggression and crowd violence 195
Keane, mid-fielder for Manchester United soccer team, deliberately caused

injury to Alf Inge Haaland from the opposition team, Manchester City. ‘I’d
waited long enough. I hit him hard. The ball was there (I think). Take that . . . I
didn’t wait for Mr Elleray [the match referee] to show the [red] card. I turned
and walked to the dressing room’, commented Keane of the incident in his
autobiography, for which he was later censured. Keane had no goal other than
to injure the opponent, and, in getting himself sent off, actually jeopardized
the chances of his team’s success. Thus, the probable cause of hostile aggres-
sion can be clearly traced and the goal of the resulting violent behaviours is
clearly focused on bringing retaliatory harm to the perceived perpetrator and
not on other instrumental outcomes.
Figure 8.1 Silva’s diagram of the different types of aggression in sport

Source: Silva (1980: 205)
Instrumental aggression, on the other hand, is intent to harm another

but with a superordinate goal to achieve an outcome that is of benefit to the
athlete, player, or his or her team. Thus, the effect of injuring the opponent
or source of the aggression mediates the effect of the athletes’ intention
on their primary goal or outcome. Unlike hostile aggression, therefore, intent
to harm is only peripheral to the overall goal (see Figure 8.1). Stories exist
that in some team environments players are encouraged to deliberately
injure opposing team players, at a risk to themselves in terms of personal
safety or sanction from the match officials and governing bodies, in order to
achieve a team goal. Instrumental aggression tends to be less spontaneous

and is not necessarily provoked by elevated levels of arousal or anger,
although it can be seen that hostile aggression can be calculated rather than
spontaneously provoked, as in the Keane example previously. There are
therefore ‘grey’ or ambiguous cases in the classification of assertive behaviours
and hostile or instrumental aggression. The distinctions and overlap between
these three behavioural types are conceptualized in Silva’s (1980) taxonomy,
illustrated in Figure 8.1.
Silva’s conceptualization of the aggression and assertive behaviours not only
illustrated the core features of the three behavioural types, but also indicates
that there are ambiguities in all three. Thus, these definitions may not
encapsulate all instances of such behaviours in sport, and some types of
behaviours may share features with more than one definition. Examples of
these types of behaviours are gamesmanship behaviours, where players perform
actions that are not illegitimate by the rules of the game but are considered
‘unsportsmanlike’ in nature. For example, a tennis coach may actively instruct
a player able to play powerful forehand volleys to hit the ball directly at their
opponent’s body at close range. The purpose of this tactic is to give them a
winning advantage as it is clearly more difficult for an opponent to return a
ball passed to the body and it is, strictly speaking, within the rules of the game.
However, the underlying intention may be to injure, physically or psycho-
logically, the opposing player. Thus, the intent to harm may exist, but it is
difficult to overtly accuse someone of this, given that the behaviour is legitim-
ate within the rules of the game. Therefore, in this case, the behaviour shares
some properties of assertive behaviours, particularly legitimate force (see Fig-
ure 8.1), but also with instrumental aggression, particularly intent to harm.
Theories of aggression
Frustration–aggression hypothesis
One of the earliest social psychological explanations of aggression was put
forward by Dollard et al. (1939), known as the frustration–aggression hypoth-
esis. Aggression was supposedly the product of an anger response to the frus-
tration of goals and desires of the individual. It predicted that frustration led
an individual to select and direct their frustration towards the perceived source
of that frustration in the form of aggression. Therefore, a highly skilled field
hockey forward may become frustrated because she is not getting sufficient
service in the form of cross and through balls from the mid-field players to
score sufficient goals. However, she is also marked out of the game by an
opposing team defender. According to the frustration–aggression hypothesis,
the frustration felt by the hockey player will ultimately lead to aggression, and
she may decide to take out that frustration by assaulting the defender. How-
ever, while this may happen occasionally, frustrated hockey players seldom
assault their markers and often do little more than create some minor fouls or
infringements. Indeed, there is little support for the frustration–aggression
hypothesis in sport. Theorists have recognized that the competitive environ-
ment of sport does lend itself to frustration, but seldom aggression (Russell
1993). This was a major limitation of the frustration–aggression hypothesis
because frustration did not lead inevitably to aggression. It was therefore inef-
fective in explaining the circumstances under which frustration did not lead to
aggression. Recent revisions of the theory have incorporated hypotheses from
social learning theory (Bandura 1977, 1997) to delineate the factors that affect
the translation of frustration into aggression (Berkowitz 1993). Berkowitz pro-
posed that the arousal induced in stressful situations gave rise to an emotional
response such as frustration. However, frustration would only be manifested
in aggression if previously learned patterns of behaviour linked frustrated
emotions in that situation with an aggressive response.
Social learning theory

Bandura’s (1977, 1997) social learning theory has already been considered in
Chapters 2 and 5 as an influential theory of motivation. The observational
learning aspects of social learning theory can also be applied to the genesis of
aggressive behaviour in social situations. Bandura’s seminal work on aggres-
sion as an imitation of modelled behaviour or vicarious experience by children
serves as both an important but also controversial and distasteful comment on
the development of human aggression. Bandura’s seminal ‘Bobo doll’ experi-
ments not only supported the hypothesis that aggressive behaviour is acquired
through the observation of aggressive acts, but could also explain the circum-
stances most likely to result in their manifestation and their persistence or
maintenance.
Social learning theory has had a profound effect on how aggression is por-
trayed in the media and its effects on children. However, Bandura outlined
that there were moderating factors on the extent of imitation of aggressive
behaviour. These factors included: (1) past experiences with the aggressive
behaviour, both personally and the observation of others; (2) previous ‘suc-
cess’ with aggressive behaviour in relation to personal goals; (3) the pattern of
reinforcement expected on the display of aggression – is it likely to be
rewarded or punished?; and (4) psychological, social, and environmental
factors such as personality, verbal encouragement, and presence of significant
others, respectively. Bandura’s theory is therefore more comprehensive and
far-reaching than the frustration–aggression hypothesis because it explains the
development and conditions under which aggression occurs.
In sport, social learning theory has been used to explain the effects of overt
aggression in professional sports on hero selection among fans. Using inter-
views and archival data, Russell (1979) found that goals scored and penalties
incurred predicted ice-hockey fans’ favourite player and team choice. In par-
ticular, teams were selected more on the basis of penalties conceded than goals
scored, suggesting that aggressive behaviours have a more pervasive effect on

attitudes than performance indicators. Russell attributed this to the media’s
disproportional focus on aggressive conduct, resulting in such acts becoming
so ubiquitous in the coverage that they are considered an integral part of the
game and are associated with success. The suggestion is that the participants
learnt to associate success with aggressive behaviours in the professional game
and used it as a criterion for their selection of their favourite team. However,
Russell (1993) subsequently reported that there was no relation between the
level of aggressive play in the favourite teams and the aggressive behaviours
displayed in their own play. This suggests that the players may have learnt that
success among professionals is concomitant with aggression, but this was not
something that was necessary for success in their own play.
Social learning theory has also been applied to help drive interventions
to promote assertive behaviours in sport, but to avoid the learning of
aggressive behaviours. Connelly (1988) suggests that modelling of aggressive
players’ behaviours is an appropriate technique to modify non-assertive play-
ers’ behaviour in competition towards being more intense. However, she
claims it is important when making such modifications to clearly distinguish
between increased intensity of sport behaviours that are legitimate and goal-
directed and the features of the aggressive behaviours such as those outlined in
Silva’s (1980) model (see Figure 8.1). Therefore, it is important when using
modelling and techniques to enhance assertive but not aggressive behaviours
that an overt distinction is made between desired behaviours and overly
aggressive play that may be penalized and result in injury to other players.
Personality and individual difference explanations

Some personality characteristics have been linked to aggression, although
there is no evidence to suggest that there is one sole trait that characterizes
those who engage in aggressive behaviours. Research examining the associ-
ations between personality and aggression has shown that one personality
factor, agreeableness, is not only strongly and negatively associated with
aggressive behaviour in adolescents, but also predicts the social cognitions
that are hypothesized to give rise to aggression such as frustration. This is an
important finding because agreeableness is the personality dimension that is
most associated with maintaining positive and harmonious relationships with
others. Recent research, however, has identified aggression as a distinct and
unique personality factor (e.g. Zuckerman et al. 1993). Zuckerman et al. isol-
ated an aggressiveness–hostility factor that was distinct from the other person-
ality dimensions. Future research will further examine the predictive validity
of the relationships between these personality dimensions and aggressive
tendencies, particularly in sport contexts.
In addition to personality factors, researchers have also identified a stable
behavioural pattern known as ‘Type A’ personality (Matthews et al. 1982).
Type A personalities are highly-driven, competitive, and extrovert in their
behaviour and interactions with others. There is some evidence to suggest that
Type As display higher levels of aggression and hostility towards others in
competitive situations. It is likely that the interaction between the situational
climate, e.g. performance-oriented or competitive, is likely to interact with
personality factors such as a Type A pattern in contributing to aggressive
responses. There are few studies examining Type A personality in aggression in
sport. Biasi (1999) found that Type A personalities are more than twice as
prevalent in athletes and dancers than a normal comparison group. Despite
their possible Type A personality, dancers were much less likely to express
negative emotion and more likely to avoid interpersonal conflict. This seems
to contradict the expected behavioural pattern of a Type A personality. How-
ever, it can be seen that emotional control and low interpersonal conflict are
very important to dancers who have to act cooperatively for success and for
whom emotionality is a sign of weakness and fragility, both undesirable
behaviours in their profession. To speculate, it may be that social cognitive
variables such as group cohesion and self-efficacy towards success may
moderate the effect of Type A personality on aggressive behaviours.
A major criticism of personality theories of aggression is that they are
limited because they ignore the influence of individual and group-level social
cognitive constructs. While the effects of personality constructs such as agree-
ableness and Type A personality on aggression indicate that personality
constructs account for a modest amount of variance in aggression, a large
proportion of the variance goes unexplained. More sophisticated approaches
would take into account situational and social cognitive influences on
aggression that would identify the mediators and moderators of the
personality constructs on aggressive behaviour. This may provide answers to
the questions about the relative contribution, if any, that personality makes in
the explanation of aggressive acts in sport performers.
Factors that influence aggression
The cathartic hypothesis

It is often said that aggression is instrumental in venting ‘pent-up’ frustration
and unexpressed emotion. The act of expressing such emotions and the con-
comitant feeling state of release are known as catharsis, and the premise that
this occurs in social situations where aggression serves this purpose is known
as the cathartic hypothesis. The belief in this instrumental purpose of aggression
is prevalent in sports, particularly among spectators and observers of sport. For
example, Wann et al. (1999) found that participants believed in the cathartic
function of watching aggressive sports and said that they were less likely to be
aggressive themselves because of this effect. Further, the belief that participa-
tion in sports with aggressive characteristics is healthy for young people
because of its cathartic function is prevalent among sports coaches and school
teachers (Bennett 1991).
Despite these deep-seated beliefs, there is very little evidence to support the
cathartic hypothesis, and research suggests that persistent exposure to aggres-
sion does not reduce the likelihood of it being expressed elsewhere but, para-
doxically, such exposure increases it. For example, Bushman, Baumeister, and
Stack (1999) gave experimental participants one of three ostensible newspaper
articles written in support of the cathartic hypothesis (pro-catharsis), against
the cathartic hypothesis (anti-catharsis), and neutral point of view. Partici-
pants were then asked to write an essay from the point of view expressed in the
article they read, which was to be criticized by a student in another room. After
hearing very harsh criticism of the essay, participants were asked to fill out an
emotion questionnaire to measure anger and then given a choice of task, one
of which was a punching bag exercise. Angry participants in the pro-cathartic
condition were more likely to choose the punching bag exercise, indicating
that media attention may propagate the belief that aggression has a cathartic
effect. A follow-up study showed that participants in the pro-cathartic condi-
tion were more likely to administer a high level of punishment to their critic,
even those who had spent time in the punching bag task. Therefore, while
participants believed in the cathartic purpose of aggression, they were more
likely to commit an aggressive act (administer punishment) when they
punched the bag and received the cathartic essay. These data suggest that the
cathartic hypothesis is a belief rather than an actual effect, although research
suggests that the belief is as present as ever despite little empirical support
(Bennett 1991).
Gender
It is often cited that men are more likely to engage in aggressive behaviours
than women (Wrangham and Peterson 1996) and more likely to express
aggressive attitudes and beliefs (Eagly and Chaiken 1993). These differences
have been attributed to: (1) elevated levels of androgens (sex hormones) in
males; (2) the notion that aggression has an evolutionary benefit in demon-
strating dominance and status; and (3) the socialization of aggressive tenden-
cies in men during development. In sport, the majority, but not all, of the
aggressive acts seen in team competitions are committed by male athletes, and
men seem more likely to endorse aggression in sport than women (Tucker and
Parks 2001). Young girls have been shown to express high levels of moral
behaviour in sports participation (Stephens and Bredemeier 1996), but studies
suggest that females do endorse aggressive acts towards opponents if the group
norm endorses it (Stephens and Bredemeier 1996; Tucker and Parks 2001).
Interestingly, though, females engage in verbal assault as readily as men even
though they do not engage in subsequent aggression, verbal or physical
(Harris 1992).
Morality issues
Prosocial behaviours are representative of the general moral conduct and con-
formity observed towards others in everyday life. However, as mentioned
previously, social identity theory suggests that personal moral behaviour can
be rendered insignificant in contexts where group membership is salient and a
group norm pervades. In such situations, people become ‘deindividuated’ and
assume the social identity of the group members. In doing so, the individual
internalizes the attitudes of the group and has a tendency to evaluate ingroup
members positively and outgroup members negatively. Bredemeier and
Shields (1986) suggest that assuming a group identity in sports teams could
result in the prosocial moral attitudes normally pervasive in everyday life
being suppressed and individuals undergoing a moral transformation towards
a more egocentric view of morality. This results in a bracketed morality in which
the ‘usual moral obligation to equally consider the needs and desires of all
persons’ (1986: 257) is suspended. This bracketed morality can lead to a legit-
imization of injurious or aggressive acts in a sports setting. Bredemeier and
Shields illustrated this by presenting moral dilemmas to establish the level of
moral reasoning in basketball players and non-athletes. The moral dilemmas
pictured situations in which aggressive acts in sport were legitimized and par-
ticipants were asked to make a decision about what the correct choice would
be and their reasons why. Findings indicated that basketballers cited more
egocentric and less prosocial reasons in their resolutions to the dilemmas,
indicating a clear egocentric bias in sports participants’ moral reasoning. These
findings illustrate a moral element to the acceptability of aggressive
behaviours in sporting contexts and a legitimization of these acts in the eyes of
athletes.
Arousal
Other factors that contribute to aggression in sport may be external physio-
logical factors such as heightened physiological arousal. By the nature of
intense effort, athletes tend to be in a state of heightened autonomic arousal
during competition. Further, this heightened state of arousal may seem to
abate after a short recovery period, but can be elevated for hours afterwards
(Zillman et al. 1974). For example, there is evidence to suggest that heightened
physiological arousal can be misinterpreted by an individual and expressed as
an emotional response like anger or frustration. Further, a participant may not
be aware that their heightened physiological state is responsible for their
emotional response (Zillman et al. 1974). This was particularly pervasive in
Zillman et al.’s experiment in which participants in an aroused state due to
physical exertion expressed more hostility towards an actor who insulted
them when they were still aroused from the exercise. Importantly, the hostility
was exaggerated when their awareness of that arousal was negated by the
introduction of a time gap between the insulting situation and the bout of
exercise. In sport, particularly multiple sprint sports like hockey, rugby, soccer,
and lacrosse that have periods of intense effort followed by periods or rest,
players are likely to be in a physiologically aroused state that may contribute to
feelings of anger and, in a group situation where a norm legitimizes it, result in
aggression.
Hormones and steroid abuse

Other external factors that may play a role in influencing the tendency of
individuals to display aggressive behaviour in sport may be the prevalence of
substance abuse in players. While the level and type of substance abuse vary
across sport type and level, this cannot be denied as a potential contributory
factor to aggression. Pharmacological studies have reported a clear link
between the abuse of anabolic steroids and aggressive behaviour (Pope and
Katz 1994). The abuse of such substances by athletes changes the androgen
profile of the athletes which results in behavioural changes due to increased
levels of testosterone. Indeed, testosterone has been implicated in the higher
incidence of aggressive behaviours expressed by male athletes when compared
with females (van Goozen et al. 1994). Although the relatively low prevalence
of drug abuse may mean that the overall effect of such hormonal changes on
aggressive behaviours may be minimal, it is important to note that situational
factors may interact with physiological changes caused by drug abuse and
predispose such individuals to aggressive conduct.
Crowd violence, collective aggression, and hooliganism
The aggression between athletes in sport encompasses only a small part of the
aggressive acts that occur in sporting contexts. Aggression and violence,
particularly among spectators, crowds, and fans, have been a prevalent,
notorious, and unpleasant occurrence at sports venues and events worldwide,
particularly in Western Europe, and North and South America. It also seems
that particular sports are plagued with crowd violence (e.g. soccer). Although
its incidence seems to be abating in some areas, sports authorities, govern-
ments, and anti-hooliganism groups have yet to fully eradicate this scourge of
sport. This section aims to outline the problem and get a measure of the scale
of crowd violence in sport, introduce and critically analyse some of the major
social psychological theories that have been proposed to explain these insur-
gencies, identify the influential factors that predispose crowds to be aggressive,
and provide a balanced view regarding possible solutions to the problem.
In society, one should be under no illusion as to the popularity and import-
ance that many members attach to sports and sport success. Sport is ingrained
in our social fabric, and sports fans’ moods, emotions, motivations, and per-
sonal relationships are sometimes fully ensconced in the fortunes of the local
sports team. In addition, it is clear that professional sports teams are not only
part of the ‘fabric’ of the local community and representative of its success,
they also generate a vast amount of revenue and are integral to many local
industries such as the service industry and tourism. Thus, sports teams have
great responsibility, given their power to affect people at an individual and
greater social level. It also means that when things go wrong and the team
experiences failure, such as relegation from a league competition, it may be
implicated in social ills such as violence and aggressive behaviour.
A neutral observer could cynically dismiss the influence of sport and sports
teams on aggression and aggressive behaviour in everyday life, but research
suggests that such a dismissal would be unwise, given some compelling pre-
liminary evidence of the powerful effect sports teams can have on social
behaviour. For example, Fernquist (2000) studied two indices of violent
behaviour: (1) homicide, a direct measure of aggression toward others; and (2)
suicide relative to the successes and failures of professional sports teams in 30
North American cities between 1971 and 1990. Making play-off tournaments
(end of season championship matches) was significantly related to declines in
both homicide and suicide rates, while championship wins were negatively
correlated with suicide rates only. Fernquist (2000) cited Gabennesch’s (1988)
broken promises theory to explain the results. He suggested that the frustra-
tion and hopelessness created by the unfulfilled hopes presented by the sports
teams may lead to social manifestations of those negative emotions in homi-
cide and suicide. These findings suggest that sports teams may have a pervasive
effect on social behaviour, and this must be considered when examining
crowd violence and collective aggression. However, it must also be noted that
crowd violence may not necessarily be a function of the outcome or fortunes
of the supported team, there are numerous instances in the literature where
the sport event itself and fortunes of the team are peripheral to the lived
experience of those involved in a crowd of supporters.
Theoretically, this section shifts focus away from the individual athlete and
aggression in sport contexts, and takes a broader view, examining aggression
among teams, sports crowds, spectators, and sports fans. These explanations
demand a different perspective from the approaches to individual violent
conduct and aggression which tend to focus on micro-level explanations on
situated behaviours and individual differences (Weed 2000). Instead, crowd
violence needs to be studied at group level, identifying the broader social con-
text as well as group-level perceptions held by individuals. However, it is
important to note that many social psychological theories applied to the study
of group interaction and conflict maintain an empirical, theory-testing,
information-processing approach, and use the individual as the main unit of
analysis. For example, social identity theorists, one of the major overarching
social psychological paradigms that will be visited to explain collective vio-
lence, cite the necessity to examine the group attitudes and perceptions at an
individual level (Marsh and Harré 1978; Stott 2001).
Definitions of group aggression, crowd violence, and hooliganism

Research on group processes has indicated that focusing on individual-based
factors alone in group situations will result in a misleading and incomplete
explanation of social behaviour. This is because many social psychological
theories on group processes recognize that people behave differently in groups
and tend to assume the psychological and behavioural characteristics of other
group members. A sports crowd is one of the many clearly identifiable groups
in society, and its members represent a common goal and share a common
bond in their attributes as football fans and supporters of a particular team.
One of the major social psychological frameworks adopted to explain crowd
violence, social identity theory, predicts that the attributes of fellow crowd
members, the way other group members are evaluated, the way members of
rival teams fans are portrayed, and the importance of the crowd to individual
members’ self-esteem are all important elements of crowd behaviour.
Crowd violence is a form of collective aggression. Collective aggression can be
defined as violent, unified behaviour with intent to cause injury or harm to
another individual or group of people. In some cases, such unified aggression
may occur even when the defining characteristics of the group are very weak
and the individuals within the group may not know each other very well
(Tajfel and Turner 1986). Much research on crowd violence has been centred
about the behaviour of soccer fans in Europe, known as hooliganism (Marsh
and Harré 1978). There is little consensus over the precise definition of a
hooligan or hooliganism and the meaning of these terms depend on the
source and academic approach to the hooligan events (Weed 2001). Although
researchers tend to view hooliganism as the embodiment of a crowd of
sports supporters displaying negative behaviours associated with collective
aggression and violence, the definition also encompasses non-violent,
extroverted behaviours displayed in a sports crowd such as singing, chanting,
jeering, and shouting. These non-violent behaviours often constitute an
alternative definition of ‘hooligan’ within a group of supporters, referring to a
person performing extroverted or outrageous acts (Marsh and Harré 1978;
Weed 2001). Such behaviours among sports supporters would be condoned
and actively encouraged, such that the self-esteem and group identification
of individuals making such behavioural displays would be enhanced
and reinforced through the positive feedback given by other group members.
On this basis, one could say that, in some groups of sports fans, the label
‘hooligan’ would be positively evaluated and worn like a ‘badge of honour’,
rather than have the negative, violent connotations often associated with the
term. Often the media are responsible for whipping up adverse public opinion
regarding hooligans through a hyperbolic stereotype of a typical sports
‘hooligan’ and the behaviour of crowds of hooligans (Weed 2001).
Theoretical accounts of collective aggression

Numerous theoretical approaches from the social psychology literature have
been put forward to explain collective aggression and crowd violence. Some
of these theories focus on individual differences on group-level variables using
an empirical, hypothesis-testing framework from an information-processing
perspective to explain collective aggression. Prominent among these theories
are social learning theory (Bandura 1977), reversal theory (Kerr 1997), and
social identity theory (Tajfel and Turner 1986). Conversely, theories that stem
from a more sociological tradition tend to examine the effects of broader
social contexts on crowd violence (Ward 2002). These approaches include
Dunning and co-worker’s (Dunning et al. 1988), also known as the ‘Leicester
School’, figurational approach to soccer hooliganism that focuses on cultural
norms, conflict theory whose roots are in Marxism and tell of the working-
class resistance to the embourgeoisement of football (Coakley 1981), and func-
tionalist theory that outlines the purpose of institutions such as sport to
maintain a well-functioning society (Marsh and Harré 1978). Other theor-
etical accounts come from innate biological bases for behaviour such as
ethological theory or instinct theory and hooligan addiction theory (Ward
2002). The aim of this section is not to provide a comprehensive overview
of these theories, readers interested in learning more about the various
approaches are directed to the numerous lucid reviews available on the prevail-
ing social psychological (e.g. Tajfel and Turner 1986; Kerr 1997) and socio-
logical (e.g. Dunning et al. 1988; Weed 2001; Ward 2002) theories of crowd
violence. Instead, the present section aims to provide a comprehensive over-
view of four key approaches from the psychological social psychology (social
identity theory and reversal theory) and sociological social psychology (the
figurational approach and conflict theory) traditions. The major premises of
these approaches will be critically reviewed, the empirical research in these
areas evaluated, and their effectiveness as accounts of crowd violence and
group aggression examined.
Social identity theory

Social identity theory was first introduced in Chapter 7 as an explanation for
inter-group processes and group cohesion. In this section we will outline how
major premises of social identity theory serve to explain prejudice and collect-
ive aggression among sports spectators. The aim of social identity theory is to
provide an explanation of the mechanisms behind social behaviour in a group
context by studying how the influence of group membership affects the
behaviour of individuals within the group. The theory supposes that indi-
viduals forego their own identity and assume the shared beliefs, attitudes, and
expectations of the group. This leads to a number of processes that can result
in complex group and individual patterns of behaviour such as prejudice and
aggression towards other sports teams.
One of the key hypotheses of social identity theory is that individuals’

personal attitudes are superseded by the attitudes shared by the group. An
individual in a group will assume the beliefs, judgements, and behavioural
patterns shared by other group members. They do this because their self-
esteem is tied in with their membership of the group. Readers will recall in
Chapter 4 that self-esteem statements form the basis of a self-concept. In the
case of assuming a social identity, self-concept statements will, in part, reflect
the attributes and values that person shares with the group(s) to which they
belong. In order to maintain a positive sense of self, group members have to
experience a feeling of belonging (identification) with the group, so common
interests and attitudes are important, and they have a need to receive
ongoing endorsement of their self-esteem and so view themselves as a proto-
typical group member. This results in self-stereotyping where an individual
in a group or crowd categorizes him or herself as part of group. This self-
categorization process results in group members conforming to the norma-
tive behaviours (what is acceptable to the group) and people look to the core
group members for guidance or as models. This self-categorization process is
met by a social comparison process in which the person forms judgements
regarding members of his or her own group, the ingroup, relative to members
of other groups to which they are not members, or outgroups. As we shall see
later, these comparison processes form the basis of inter-group prejudice. The
net effect of assuming the shared attitudes of the group and the investment
of self-esteem within the group is a deindividuation of the person. Deindi-
viduation, proposed by Festinger (1954), supposes that the individual no
longer views themselves in individual terms and this results in anonymity
and a loss of individual identity. Clearly, the person does not lack identity,
instead they assume a social identity that they share with the other group
members.
Deindividuation, as an isolated process, has provided an explanation for
group members’ expression of behaviours that they would not normally dis-
play in an individual context (Reicher et al. 1995). In a group situation, each
group member’s relative anonymity within the group means that their indi-
vidual responsibility is reduced as they are no longer readily identifiable for
repercussions and there is a lack of the personal control that is normally exer-
cised in individual situations (Mann 1981). As a result, behaviour of group
members can become more disinhibited, extrovert, instinctive, and irrational.
A number of experimental studies have used uniforms and other means to
create an artificial group environment and deindividuate participants. These
studies found that deindividuated participants tended to display more divisive
behaviours than easily identifiable participants. For example, deindividuated
participants were more critical of their parents and were more likely to steal
when given the opportunity than identifiable participants (e.g. Diener 1976).
Most compelling and relevant to the study of crowd violence in sport are
Zimbardo’s (1970) studies that showed deinviduated participants to have a
greater propensity to act aggressively.
The processes of self-categorization, group identification, and self-

stereotyping, and deindividuation result in conformity to group norms and
group members assuming the prevailing attitudes of the other group members.
Under these circumstances, a social identity is created that is separate from
personal identity and personal self-esteem. As a result, ingroup members are
more positively evaluated than members of outgroups. Even though the
ingroup is viewed in a largely stereotypical way, their salient features are more
relevant to the ingroup members while outgroup members are viewed in a
relatively homogenous way. This creates an ‘us’ and a ‘them’ distinction. In
groups whose group membership is perceived to be under threat or leads them
to question their group membership, prejudice against the outgroup members
can result. This prejudice is likely to take the form of exaggerating differences
between the ingroup and the outgroup, such as stygmatizing the membership
of the outgroup through arbitrary surface features such as race and religion.
Such prejudice may be manifested in aggressive behaviours.
Although sports crowds very seldom resort to actual aggression, the threats
are often displays of solidarity more for the benefit of galvanizing the ingroup
members than to bring harm to the outgroup. The act of stygmatizing others
maintains the positive sense of self or social identity that the group members
glean from their membership of the ingroup. Interestingly, Tajfel and Turner
(1986) argue that group prejudice can occur even in groups that share the most
minimal of characteristics, as outlined by their minimal group experiments,
which may explain why often crowds of sports spectators often seem like
a coherent group with a coordinated set of behaviours. Indeed, studies of
crowd violence such as rioting and looting have shown surprisingly orderly
patterns of aggressive behaviours that are directed towards specific targets
representative of outgroups.
Despite the large body of literature on social identity theory examining col-
lective aggression and crowd violence in the context of riots between social
groups, there have been relatively few studies examining the role of social
identity theory on crowd violence among opposing supporters and between
sports crowds and police in a sport context. Stott and colleagues (Stott 2001)
have conducted one of the most in-depth and comprehensive analysis of col-
lective aggression adopting a social identity theory approach. Stott used
ethnographic and qualitative methods to develop an extended social identity
model that examined the effects of contextually defined social identity on the
display of aggressive behaviour in collective ‘disorder’ among soccer fans.
Reicher’s (1987) original theory of self-identity indicated, after Tajfel and
Turner (1986), that collective behaviour is determined by the social identity
acquired by group members that is, partly contextually defined (e.g. soccer
game context when confronted by opposing fans) and partly determined by
collective norms (e.g. shared attitudes towards outgroups) rather than the
personal identities and beliefs of the individuals in the group. In the extended
model, Stott argues that the effects of collective identity on group behaviours
and intergroup conflict is reciprocal rather than one-way. Therefore, not only
does collective identity determine the purpose and nature of sport crowd
behaviour, but the group behaviour affects and changes the situation. Given
that the collective identity is defined, in part by the context, changes in the
environment will therefore affect and change social identity.
In summary, social identity theory has provided a unique and comprehen-
sive explanation of the mechanisms and processes that underlie violent
behaviour in sport crowds. Adopting theoretical perspectives from Festinger
(1954) and self-esteem theorists, Tajfel and Turner (1986) highlighted the
importance of group-level individual difference constructs like social identity
that supersede personal constructs like attitudes and self-identity in social con-
texts. Importantly, self-esteem is considered a primary reason why prejudice
develops between groups. This is because group members’ self-esteem is tied in
with their group membership, and when the group is threatened by others,
this threat is viewed as a threat to self-esteem. In order to maintain a positive
sense of self-esteem, group members are likely to vigorously defend their group
in the interest of maintaining a positive sense of self through the group. This
can result in an escalation of aggression and violent displays toward outgroup
members and, without proper moderation, may result in scenes of confronta-
tion and violence between opposing sets of fans and/or police seen at soccer
venues throughout Europe.
Reversal theory
Kerr (1997) adapted Apter’s (1982) reversal theory to the study of sport aggres-
sion and crowd violence in soccer. Reversal theory is described in detail in
Chapter 6 with respect to motivation and emotional experiences in sport.
Recall that reversal theory outlines the metamotivational states – psycho-
logical evaluations of the direction of a person’s motives in that given context
– that underpin motivation to engage in a behaviour. The context is important
as it will determine which metamotivational state operates and can instigate a
‘reversal’, i.e. a switch from one metamotivational state to another. The
external contingencies that lead to these reversals are frustration and satiation
of goals. In Chapter 6, the telic and paratelic metamotivational states were
introduced. A telic state orientates a person toward being directed and serious
in the pursuit of their goals, while a paratelic state reflects a playful, activity-
oriented approach. The telic–paratelic motivational states are considered to be
bipolar, so in any given context, a person can be in either state. This is why it is
possible for a person to have a goal-directed motivational orientation at one
occasion or point in time, but to hold an opposite motivational orientation at
another according to the effect of the context to frustrate or satiate goals. Apter
introduced several other bipolar metamotivational states that operate along-
side the telic–paratelic states, conformity–negativism, mastery–sympathy, and
autic–alloic. The autic state represents a concern or focus on oneself while an
alloic state reflects a concern for others. The interpretation of arousal as pleas-
ant or unpleasant, or its hedonic tone, interacts with the metamotivational
states to give rise to specific emotional experiences. While metamotivational

states are changeable according to situations, personality factors can deter-
mine the dominant or prevalent metamotivational state and dictate how long
a person remains in that state.
Applied to soccer hooliganism and crowd violence, reversal theory makes
predictions regarding the events that lead to violence at soccer matches.
Members of ‘hooligan’ crowds are proposed to be paratelic dominant, that is,
they seek activity and excitement and, coupled with low arousal and an
unpleasant hedonic tone, are likely to become bored at football matches. In
this case their expected aroused state and pleasant hedonic tone from the
football has been frustrated and they have experienced reversal. This results in
a desire to seek satiation by pursuing arousal-giving, exciting actions else-
where in the context. According to Kerr, the excitement-seeking traits of the
soccer hooligan are similar to others who engage in thrill-seeking behaviour
such as bungee-jumping. Therefore, the unfulfilled satiation of anticipated
excitement in those fans who have a predominantly paratelic metamotiva-
tional state leads to compensatory behaviour and this manifests itself in
aggressive behaviours. Critics of the theory claim that Kerr does not offer any
explanation why certain people become hooligans in the first place and why
paratelic dominant fans do not seek other non-violent compensatory
behaviours.
The figurational approach

One of the most influential sociological theories of crowd violence was pro-
posed by Dunning and the ‘Leicester School’ of researchers (e.g. Dunning et al.
1988). Their theory was developed in the ethnographic study of soccer sup-
porter violence. As a starting point, the figurational approach to soccer
hooliganism adopts an individualist perspective and is at odds with the ‘group
mind’ and ‘submergence’ theories of collective behaviour that were pervasive
in early explanations of collective behaviour. The theory therefore focuses on
an individualist perspective because dispositions within the individual are
hypothesized to give rise to violent behaviour. Dunning and co-workers draw
on Elias’ (1978) theory of civilization that proposes that members of society
expect a ‘civilizing’ process to occur over time. This civilizing process is
hypothesized to occur in a top-down fashion from the privileged to the work-
ing classes. It is proposed that in this civilizing process, the unstructured and
unexpected periods of aggression displayed by groups in the past would be
replaced by ritualized, socially constructed types of aggression. Therefore,
individuals in society experience anxiety and anger at the prospect of the
expression of such aggression, and guilt and shame for experiencing aggressive
tendencies. Dunning et al. support Elias’ proposition, but argue that the civil-
izing process has not yet filtered down entirely to the working classes. How-
ever, Dunning et al. state that violent crowds in soccer are dominated by the
‘rough’ working class, a premise that is supported by statistics which indicate
that sports crowds comprise predominantly of partly skilled or unskilled

workers (Murphy 1990).
It is the disproportionately high presence of individuals of a ‘rough’ work-
ing-class background in the sports crowd that is thought to lead to displays of
collective aggression. Dunning et al. proposed that these elements are likely to
display aggression because of internalized cultural values that are inherently
masculine in nature and a propensity to resolve issues with aggression. These
values suggest that among the ‘rough’ working class, violent conduct is both
legitimized and respected as a means of addressing perceived injustice. As a
result, the figurational approach to soccer hooliganism explains why incidents
of hooliganism arise in the context of soccer fans, due to the civilization
process and presence of working-class elements, but focuses on individual dis-
positions rooted in cultural norms as causes of violent behaviour in group
contexts.
Conflict theory
Steeped in Marxist ideology, conflict theory suggests that violence in sports
contexts is connected to underlying social, political, and economic grievances
held by the working-class element of the crowd. Therefore, sport contexts in
which fans share these common resentments serve as a hotbed for the expres-
sion of aggression directed at the perceived perpetrators of the inequalities.
Leading proponents of this theory suggest that sport is an ideal conduit for the
expression of dissent through aggression because elements of modern profes-
sional sport are perceived as economically exploitative of the ‘average’ work-
ing-class fan (Levine and Vinten-Johansen 1981). Collective aggression is
therefore explained in terms of perceived economic and social injustices spe-
cific to the sport. Taylor (1971) and Clarke (1978) talk of crowd violence in
soccer as an attempt to wrestle the honest, working-class values underpinning
the game from middle-class elements who seek to claim it for themselves.
These working-class elements feel marginalized by an increasingly bourgeois
society controlled by the nouveau riche.
On the surface it seems that conflict theory and the figurational approach to
soccer hooliganism share common explanations in that sport violence is
viewed as emanating from the working class and their grievances. However,
they fundamentally differ in the proposed direction of the aggressive displays.
Conflict theory suggests aggression is a reflection of the grievances and
injustice felt by the working-class elements in sports crowds and an attempt to
show dissent, defiance, and redress the balance, while figurational theory
explains the violence as tendencies arising from cultural values that legitimize
violence as a means to resolve conflict. While conflict theory appeals, particu-
larly in light of the economic rise of elite professional sports leagues owned by
super-rich benefactors such as Premier League soccer in England (Weed 2001),
Taylor (1971) proposes that its premises are speculative and have yet to receive
empirical support. Furthermore, Weed (2001) views this approach as having
limited explanatory power as much sports crowd aggression occurs between

supporters of rival teams and not toward the authority figures of sport.
Influential factors in crowd aggression
There are numerous environmental and social factors that can affect the inci-
dence and public perception of crowd violence and collective aggression.
Some of these have been reviewed previously such as crowd size and drug
abuse. This section will briefly outline two of the important contributory
factors that may serve to moderate collective aggression, and also affect the
perception and evaluation of members of sports spectators by those outside
these groups.
The media
As mentioned previously, social learning theory predicts that the observation
of violence may have a pervasive influence on others (Bandura 1977, 1997).
Television and newspapers are prompt in their condemnation of collective
aggression when these events are portrayed in the media. However, the cover-
age is usually a thin, diluted, and one-sided view of the events, which can have
damaging effects on the sport and those not involved in the violence but
whose livelihood is tied in with their involvement in the sport. Moreover,
despite the bias in the coverage, such events are often given a huge profile
within the media, and such prominence, if not coupled with a sense of per-
spective or explanations of contingency, may result in such events being per-
ceived as the norm. As we shall see later, this may have the effect of damaging
the reputation of the sport or even create a hyperbolic response and moral
outrage. Further, it may have an effect on young players and fans in sport. If
crowd violence and violence between players is given such prominence and is
observed by young people over long periods, it may be viewed as the accepted
norm and legitimized in those contexts. The social learning of aggressive acts
may then be internalized and re-enacted by young players and fans, habits
which could perpetuate such behaviours. There is evidence to support this
such as the tendency for players to select their heroes on the basis of their
aggressive play rather than other performance indicators (Russell 1979). Fur-
thermore, research has shown that similar behaviours, including infringe-
ments, are viewed in the same sports team across generations, suggesting that
younger players mimic the behaviours of other older players and these are
passed on (Russell 1993).
Recently, Weed (2001) has indicated that media coverage of crowd violence
in sport tends to be portrayed disproportionately by the media, a fact that may
perpetuate hooligan stereotypes among the general public, but also by the
authorities such as the government and police forces involved. Weed uses the
example of media portrayal of crowd violence among English and German
fans at the 2000 European Championship soccer tournament in Charleroi,

Belgium. The English tabloid press tended to portray the violence as caused
by English soccer fans and on a wider scale, while broadsheet and television
coverage provided a greater sense of perspective: very little violence or hand-
to-hand fighting was observed with the crowds mostly reacting to a small
hard-core minority of hooligans that tried to drum up support for the violence.
It was argued that past reputation plus the culture of English soccer fans in
Europe may have resulted in the judgmental treatment of the evidence. To
explain this limited, small-scale violence, Weed cites Marsh and Harré’s (1978)
meso-level approach in which individuals in a group gain status through the
display of aggressive behaviour, but rely on the support of the other crowd
members to prevent any real harm coming to them. Overall, the research sug-
gests that some elements of the media produced biased and sensationalist
portrayals of the crowd violence in Charleroi. This misrepresented view had
the potential to evoke an unwarranted moral outrage among the general
public and at the same time sold a lot of newspapers.
The pervading sense of moral outrage that may be incurred by the exagger-
ation of the scale of events in hooligan contexts such as those outlined by
Weed (2001), may filter through to policy-makers, resulting in increasingly
stringent and heavy-handed measures towards even the most innocuous of
non-violent behaviours, such as singing, chanting, and shouting. Such meas-
ures are less likely to bring the potential for violence in check than to provoke
increased violence among fans and police, where fans are seen as protecting
their rights to perform non-violent forms of support for their team (Stott
2001). Ward (2002) suggests that governments and other institutions like
newspapers can use such events for positive public relations to restore public
faith in their desire to deal with unwanted social elements and to keep their
citizens safe. It is therefore important that authorities as well as the public view
the portrayal of collective aggression in sport from a number of sources and
adopt a considered approach to the role of the media and its tendency to
misrepresent such incidents and manipulate public opinion.
Alcohol abuse
Misuse of alcohol is commonly implicated in a great deal of social violence,
particularly aggressive behaviours in the home and in public places. In the case
of crowd violence and collective aggression, excessive use of alcohol is often
considered a contributory factor. This is particularly true of fans of particular
sports such as soccer, but again such a reputation may be one that is perpetu-
ated by the media. In the social psychological literature meta-analytic studies
have suggested that there is a link between alcohol consumption and aggres-
sion (Bushman and Cooper 1990), and explanations for this lie in the disinhi-
bition effect, that is, under the influence of alcohol intrapersonal controls that
may mitigate the expression of behaviours such as aggression are suspended.
However, the causal link between aggression and use of alcohol may be
moderated by contextual factors such as threat and encouragement. In terms

of threat, Taylor and Gammon (1976) found that participants intoxicated
with alcohol were prepared to give an electric shock of much higher intensity
to a confederate learner when they were told prior to the experiment that the
learner was prepared to give the highest levels of shocks when their roles
were reversed. In terms of encouragement, Taylor and Sears (1988) showed
that goading by the experimenter persuaded participants intoxicated with
alcohol to give much higher shocks to a confederate compared with control
participants. Without threat or goading, the participants behaved in a man-
ner consistent with the responses of non-intoxicated participants. These
studies suggest that the social context is important when considering the
effects of alcohol on aggressive behaviour, but if those social conditions are
present, then crowd members may be more likely to be aggressive. Alcohol
should not be made a scapegoat for collective aggression but this research
suggests that it has the potential to exacerbate behaviour, should the situ-
ational contingencies arise. With respect to solutions, it seems that calming
statements can reduce aggressive tendencies in intoxicated people, and so
telling people to calm down may be a simple and effective course of action
(Taylor and Gammon 1976). However, this may have less effect in a crowd
situation, especially if there are others goading and baiting the intoxicated
persons.
Suggested reading
Stott, C. (2001) ‘Hooligans’ abroad? Inter-group dynamics, social identity and participa-
tion in collective ‘disorder’ at the 1998 World Cup finals, British Journal of Social
Psychology, 40: 359–84. A detailed account of a social identity theory approach to
football hooliganism.
Ward, R.E. (2002) Fan violence: Social problem or moral panic? Aggression and Violent
Behavior, 7: 453–75. A brief yet comprehensive review of the theoretical approaches to
crowd violence in sport.
Weed, M. (2001) Ing-ger-land at Euro 2000: how ‘handbags at 20 paces’ was portrayed
as a full-scale riot, International Review for the Sociology of Sport, 36: 407–24. An
interesting insight into the role of media portrayal and sports crowd violence.
Summary
• Aggression is any behaviour or verbal comment that has intent to injure or

harm another person.
• Theoretical explanations of aggression among sports performers come from
the frustration–aggression hypothesis, social learning theory, and personal-
ity theories. Modern views of aggression have combined hypotheses from
social learning, frustration–aggression, and personality theories to explain
aggressive behaviours among athletes.
• Aggression is affected by drug and alcohol abuse, moral norms, and

physiological arousal.
• Collective aggression is unified violent behaviour with the intent to harm
other individuals or groups, and crowd violence is one form of collective
aggression.
• Hooliganism is a term that descries a set of aggressive tendencies in sports
fans, but can also refer to extroverted, boisterous, rowdy, and outrageous
non-violent behaviours.
• Theoretical explanations of crowd violence and collective aggression
include psychological social psychology (e.g. social identity theory and
reversal theory) and sociological social psychology (e.g. the figurational
approach and conflict theory) traditions.
• Two of the most pervasive influential factors on collective aggression are
the media and alcohol consumption.
9
Conclusion
Applied social psychology research aimed at understanding the processes,

mechanisms, and behaviours in the exercise and sport domains have tended to
treat each domain separately. This chapter will focus on some of the common
ground shared by the social psychological research in the two domains. The
chapter will identify commonalities in the research covered in the two parts
of this book and draw conclusions about these common aspects in terms of
themes, methods, and theories, and contrasts. The themes common to exer-
cise and sport research covered in this book are the prediction of behaviour,
the prediction and role of affect and other outcomes, and the prediction and
role of social influences. Common methodological approaches include cross-
sectional studies, longitudinal, cross-lagged panel designs, experimental
designs, and qualitative enquiry in exercise and sport. Theories of intention
and motivation are common across both disciplines, and these will be com-
pared and contrasted. Finally, the contrasts between the two approaches, par-
ticularly within the specific context of physical education will be covered and
recommendations for practice reviewed.
Themes
The prediction of behaviour

One of the recurrent themes in this book is the prediction of behaviour. This is
not surprising given that the overarching aim of psychology is the explanation
of human behaviour and one of the primary aims of social psychology is the
understanding of behaviour in social contexts. In the first part of this book,
social psychology was applied to the understanding of exercise behaviour, and
in particular, leisure time exercise associated with health. The main focus was
understanding the social psychological antecedents that underpinned exercise
behaviour particularly the social psychological constructs such as attitudes,
216 Conclusion
social pressures, self-efficacy and perceived behavioural control, anticipated

affect and regret, self-esteem, perceived susceptibility, and, of course, desires
and intentions. Central to the purpose of understanding exercise behaviour is
the development of social psychological models that not only identified what
constructs were important in explaining exercise behaviour, but also why they
were important. Therefore social psychological models that examined the
network of relationships that gave rise to exercise behaviour were covered,
such as the health belief model, the theory of planned behaviour, and the
model of goal-directed behaviour. In addition, the chapter also concerned
itself with the antecedent variables of exercise behaviour and how the various
antecedent and determinant variables were formed. Integrated models like the
transcontextual model aimed to explain the formation of variables like atti-
tudes and subjective norms on the basis of motivational styles such as identi-
fied external regulation from self-determination theory that were said to
arise from fundamental psychological needs. Together, these belief-based
constructs formed the distal bases for the proximal predictors of exercise
behaviour.
The second part of this book, in contrast, dealt with a very different set of
behaviours, namely sports performance. Generally, sports performance was
measured in terms of absolute success such as runs, points, or goals scored or
awarded, times recorded, distances thrown, and so on. However, on occasion,
sports performance was measured in terms of improvement relative to per-
sonal best performance. The important aspect here was that a valid, objective,
and reliable means of establishing performance was recorded. In order
to explain sports performance, the research focused on the motivational
(Chapter 5), emotional (Chapter 6), and social (Chapters 7 and 8) influences
on sports performance. The research examined a number of theoretical models
that aimed to understand sports performance from these different perspec-
tives, and identified key antecedent variables that were motivational (e.g.
achievement goal orientations and intrinsic motivation), emotional (e.g. self-
confidence and facilitative state anxiety), and social (e.g. group cohesion) in
nature. However, unlike the attempts to unify the antecedent variables in
cogent, comprehensive, and parsimonious models such as the transcontextual
model in the exercise domain, there are few models that have attempted such
integration in the domain of sport performance. One of the reasons for this
might be the diversity in performance parameters and methods responsible
for sports performance, but the other may be a lack of a clear framework to
incorporate these influences. Some of these models have attempted such a
union, such as Hanin’s (1995, 2000) IZOF model and the application of
Apter’s (1982) reversal theory by Kerr (1997). However, these theories tend to
focus on the affective and emotional aspects and less on the cognitive aspects
relating to sports performance. In addition, these theories lack parsimony and
the ability to have their hypotheses tested simultaneously. Future research
would do well to integrate findings from the affective, social cognitive and
motivational, and social influence on sports performance using an integrated
Conclusion 217
or multi-theory approach. Such a model may provide a more complete

explanation of sports performance. A good starting point may be more com-
plete tests of Hanin’s and Apter’s models, but researchers should also turn
their attention to integrate the findings of mainstream social psychological
models applied to sport such as Jones’s (1995) control model of facilitative
and debilitative anxiety and Deci and Ryan’s (1985) self-determination
theory.
The prediction and role of affect and other outcomes

In addition to behaviour, one of the major themes of this text is the prediction
of outcomes such as emotion and self-esteem. Such outcomes are considered
adaptive and desirable alongside the explanation of behaviour exercise and
sport domains. One of the reasons why such outcomes are desirable is the link
that such outcomes have with behaviour itself, and also the importance of
salient outcomes to positive well-being and general psychological as well as
physical health. Indeed, well-being itself can be considered a salient outcome
of both exercise and sport, and is cited as a reason for participation in exercise
and sport (Ashford et al. 1993). In addition, affective variables are integral to
models in the prediction of behaviour. Therefore, such affective variables can
mediate and moderate the relationships between cognition and exercise and
sport behaviour.
In the exercise domain, affective processes play an important role in the
prediction of behaviour such as affective attitudes in the theory of planned
behaviour (Ajzen 1985) and desires, that represent the affective aspects of
intention, in the model of goal-directed behaviour (Perugini and Conner
2000). Indeed, in the latter case, the affective component of the model serves
to mediate the effect that intention, a cognitive construct, has on behaviour.
Thus, emotion is an important antecedent of behaviour in such models. How-
ever, research has also indicated that motivation also leads to concomitant
positive affective states. For example, high levels of intrinsic motivation in
self-determination theory often lead to increased satisfaction when engaging
in a given behaviour, while external regulation and regulatory styles on the
perceived locus of causality reflect close to external regulation are accom-
panied by maladaptive affective states like feelings of guilt and shame (Deci
and Ryan 1985). Thus, the motivational processes that underlie behaviour can
also affect emotional states in exercisers.
In sport, the prediction of affective states and emotion plays an important
role in explaining the mechanisms underlying sport performance. However,
the focus is not on the prediction of positive affective states per se, but on the
control of negative affective states such as anxiety. For example, Jones’s
(1995) control model of facilitative and debilitative anxiety suggests that
performance is best predicted from cognitive anxiety direction and self-
confidence levels. Catastrophe theory views the level of cognitive anxiety
as controlling the debilitating effect that high levels of arousal has on
218 Conclusion
performance (L. Hardy 1990). These examples illustrate that it is not emotion
per se, but the interpretive aspects of emotion that affect sports performance.
In addition to affective states, behaviour and motivation in both the exercise
and sport domains are also linked to self-esteem in exercisers. Self-esteem
tends to be viewed as a positive outcome of exercise, and research suggests that
self-esteem may be enhanced by the demonstration of competence in the
physical domain, such as Sonstroem and Morgan’s (1989) model of self-
esteem. Self-esteem in athletes is also regarded as an important outcome.
Athletes tend to report higher levels of physical self-esteem than the general
population and those who participate in recreational exercise (Marsh et al.
1995). This is generally attributable to the higher levels of competence in
sport and athletic situations. Importantly, self-esteem is also a predictor
of sports performance and exercise behaviour, and as a result the prediction of
self-esteem is an adaptive outcome in this regard.
The prediction and role of social influences

Social influences are also an important theme that straddles the theories in
both the exercise and sport domains. In the exercise domain, all theories of
intention have some element of social influence (see Chapter 2). The theory
of planned behaviour, for example, has subjective norms as a direct predictor
of exercise intentions, but has also been expanded to include other aspects of
social influence such as social support (Courneya et al. 2000) and descriptive
norm (Rivis and Sheeran 2003). Other theories such as self-determination
theory incorporate the need for relatedness with others that has a pervasive
effect on a person’s motivation to engage in exercise and sport. Indeed,
research has suggested that a primary motive for exercise participation is for
social benefits and affiliation (Ashford et al. 1993). Thus, while personal
constructs such as intentions, attitudes, autonomy, and competence may
have strong influences on exercise motivation and behaviour, motives linked
with social participation also make an important contribution.
Social factors are also core to athletes’ participation and performance. The
motives for affiliation are similar among exercisers for recreation and competi-
tive athletes, but theories relating social factors to sports performance often
raise different questions, such as the impact of performing in a socially sup-
portive environment (Colman and Carron 2000) or in front of a crowd of
sports fans (Varca 1980). An example of the importance of social factors is
illustrated in the impact of the strength of social bonds, or group cohesion, on
sports performers in sports (Carron and Hausenblas 1998). The effect that an
audience has on sports performance is also noteworthy, and the social facilita-
tion effect has illustrated that the social environment can be both supportive
(Strauss 2002) and detrimental to performance (Baumeister and Steinhilber
1984). In particular, it seems that home teams that have high levels of cohe-
sion and are familiar with their teammates’ style of play tend to provide
excellent results in terms of performance and persistence (see Chapter 7). In
Conclusion 219
summary, social factors cannot be ignored in the examination of motivation

and behaviour in exercise and sport psychology.
Methods
Cross-sectional studies
Perhaps the most often adopted methodology in the exercise and sport discip-
lines is the usage of cross-sectional, questionnaire-based studies to confirm
measurement validity, test theoretical hypotheses, and establish patterns of
prediction such as mediation and moderation. In exercise psychology, the
majority of the theories of intention lend themselves to being tested using this
approach. There is large number of studies examining relationship among the
key antecedent variables in the health belief model, protection motivation
theory, the theories of reasoned action and planned behaviour, the transcon-
textual model, and the model of goal-directed behaviour, a conclusion cor-
roborated by the number of meta-analyses conducted in these areas (e.g.
Hausenblas et al. 1997; Hagger et al. 2002b). This is also true for sport. Meta-
analytic and narrative review studies have used evidence from cross-sectional
studies of the key relationships among variables in theories of emotion like
multidimensional anxiety theory (Kliene 1990; Craft et al. 2003), theories of
motivation such as achievement goal theory (Ntoumanis and Biddle 1999)
and attribution theory (Biddle et al. 2001), and theories of group influence
such as group cohesion (Carron and Hausenblas 1998). In addition, there is a
great deal of converging evidence to be gleaned from a large body of cross-
sectional research on theories such as self-efficacy theory and social cognitive
theory (Feltz and Chase 1998; McAuley and Blissmer 2002). The cross-
sectional or prospective approach is attractive because it represents an eco-
nomical manner to test theoretical hypotheses within a given population and
contemporary multivariate statistical procedures such as factor analysis
provide robust tests of such relationships while accounting for artefacts like
measurement error.
Such studies are also efficacious in revealing important mechanisms like
mediation and moderation. For example, the theory of planned behaviour in
the exercise domain consistently illustrates the mediating role of intentions
on the relationships between attitudes and exercise behaviour. Analogously, in
a sports context, such cross-sectional studies have supported the mediation of
the self-efficacy–performance relationship by goal setting (Theodorakis 1995).
Furthermore, moderators of important relationships in theories of exercise
behaviour such as the moderation of the relationship between attitudes
towards success and intentions to try by expectations of success in the theory
of trying, and the moderation of the self-efficacy–performance relationship
by goal setting in a sport context have been indicated (Theodorakis 1995).
The latter example being one where the same variable is implicated in the
moderation and mediation of the same relationship. These results indicate
220 Conclusion
the importance of the cross-sectional approach in explaining proposed

mechanisms within the theories.
Longitudinal, cross-lagged panel designs

Cross-sectional studies are limited in terms of their ability to infer causality.
One of the reasons for this is that they measure the social psychological
independent and dependent variables simultaneously and therefore any
theoretical directional relationship cannot be tested unequivocally when
alternative hypotheses such as the opposite relationship or reciprocity cannot
be ruled out and are tenable in an empirical sense (see Introduction). Further-
more, the researcher cannot control for the stability of the constructs. A cross-
lagged panel design is more powerful because the cross-lagged relationships
permit alternative hypotheses to be tested and control for the stability of
the constructs. Examples of these models are rife in the exercise and sport
literature. In an exercise context, such designs have been used to test the
hierarchical nature of Fox and Corbin’s self-esteem model (Kowalski et al.
2003) and Vallerand’s hierarchical model of motivation (Guay et al. 2003). In
sport, the exact same model has been proposed to test the effect of group
cohesion on sport performance (Slater and Sewell 1994). These designs are
powerful and have been able to assist researchers in resolving the nature of a
set of relationships observed in cross-sectional studies.
Experimental designs
In comparison to studies in other areas of social psychology, there is a relative
dearth of experimental studies examining the effects of key independent vari-
ables on exercise and sport behaviour. Perhaps one of the reasons for this is
that exercise and sport behaviour is difficult to assess unequivocally in a
laboratory environment, and those that do frequently involve novel tasks that
do not necessarily reflect real-life sport situations. In other words, they lack
ecological validity. However, some experimental social psychologists would
view this as an advantage of experimental approach because it allows the
unequivocal focus on the mechanism alone while any extraneous variables
with the potential to disrupt the effects are negated. Nevertheless several
ingenious social psychological experimental designs have been adopted to
examine key processes in exercise and sport. For example, in the exercise
domain, experimental field studies have examined the effects of interventions
to alter the intention–behaviour relationship in favour of doing more exercise
(Milne et al. 2002; Prestwich et al. 2003). Such field studies are useful because
they involve more naturalistic, ecologically valid settings, suggesting the
effects of such manipulations are powerful in context that are more akin with
real life than the laboratory. In the domain of sport, the provision of false
performance feedback to athletes performing novel sports tasks in a laboratory
setting has been used to manipulate self-efficacy and such experiments have
Conclusion 221
provided evidence for the effect of self-efficacy on sports performance and

persistence of effort (e.g. Weinberg 1986). In sum, the experimental approach
has provided important data to support cross-sectional studies on intention
and motivation in exercise and sport respectively and is an important method
to confirm both direction of effects and mechanisms.
Qualitative approaches
Some studies have adopted qualitative approaches to the study of exercise and
sport phenomena. Such approaches have been found to co-exist and comple-
ment other quantitative approaches. For example, the theories of reasoned
action and planned behaviour have frequently relied upon the content analy-
sis of open-ended interviews and questionnaires to inform the development of
belief-based questionnaires for exercise behaviour (Ajzen 1991). In sport, Holt
and Sparkes (2001) used a qualitative ethnographic study to identify the fac-
tors that influenced group cohesion in sports teams over a competitive season.
This study was innovative in that it used a number of different means such as
participant observation, formal and informal interviews, documentary
sources, field diaries, and reflexive journals to generate the rich set of data for
subsequent content and interpretive analysis. There are other examples in
sport such as the numerous approaches to the study of crowd violence and
football hooliganism in sport (Ward 2002). Importantly, it needs to be recog-
nized that qualitative inquiry does not necessarily imply a non-theoretical
approach. Some studies adopting this method do, but other qualitative inquir-
ies can adopt a theoretical basis that has equal rigour with quantitative
inquiry. Therefore, although less prevalent, the adoption of qualitative
approaches provides rich, informative data and has also been shown to
augment and extend the meaning of data collected by quantitative means.
Theories
As illustrated earlier, a common theme running through this text is that of

intention and motivation, and it is therefore not surprising that social psycho-
logical inquiry into the exercise and sport domains has adopted similar theor-
etical approaches. For example, theories of intention such as the theory of
planned behaviour have been adopted to examine the social psychological
influences on exercise behaviour and some aspects of sport, such as training
adherence. The theory of planned behaviour and its variants and extensions
have been a primary theoretical framework which is both flexible and parsi-
monious in accounting for exercise intentions and behaviour (Hausenblas et
al. 1997; Hagger et al. 2002b). However, there are also numerous studies that
have applied the theory of planned behaviour to account for behaviours rele-
vant to facilitating sports performance. For example, studies have found that
the theory variables can account for significant variance in sports training
222 Conclusion
participation (e.g. Theodorakis et al. 1991b) suggesting that coaches and those
involved with promoting continued adherence to training regimes in athletes
would do well to adopt the intervention techniques offered by planned
behaviour theorists. Indeed, few differences have been noted between the pre-
dictors of intention in regular exercisers and athletes attending training pro-
grammes. This suggests, as Ajzen (1985) states, that the theory of planned
behaviour is a general theory of intention and can be applied across many
contexts. Ajzen does concede that the relative contribution that each theory
variable makes in the prediction of intention may differ, a hypothesis that has
been supported empirically. However, since exercise and sport are similar con-
texts and the behaviours involved are similar in terms of the pressures and
constraints involved such as the amount of relative control a person perceives
they can have over the behaviour, their antecedents are likely to exhibit
congruence.
In addition to theories of intention, self-determination theory has been
adopted to explain the effect of the motivational styles and the contextual
contingencies under which an individual operates on intrinsic motivation
and behaviour. In exercise, the primary focus of this work has been in the
adoption of motivational styles delineated by the perceived locus of causal-
ity and organismic integration sub-theory to predict exercise behaviour.
Intrinsic motivation and intrinsic forms of external regulation such as
identification have been found to influence the proximal psychological
determinants of exercise behaviour such as attitudes and intentions
(Chatzisarantis et al. 2002; Hagger et al. 2002a; Standage et al. 2003), as well as
exercise behaviour itself (Chatzisarantis et al. 1997; Chatzisarantis et al.
2002). In addition, contexts that support intrinsic motivation and identifica-
tion have been shown to give rise to intrinsic motivation and have a posi-
tive, indirect effect on leisure-time exercise behaviour (Chatzisarantis et al.
2002; Hagger et al. 2003b). In addition, self-determination theory, particu-
larly intrinsic motivation and an autonomy-supported motivational climate,
has been shown to predict intentions to participate in sports training
(Escarti and Gutierrez 2001) and actual sports participation (Robinson and
Carron 1982). Such evidence corroborates the adoption of the theory of
planned behaviour to account for the underlying influences on exercise and
sports training participation (Theodorakis et al. 1991b). However, self-
determination theory has also been adopted to explain motivation to par-
ticipate and persist in sports tasks. For example, intrinsic motivation has
been shown to be related to motivational variables like goal orientations
(Seifriz et al. 1992; Duda et al. 1995; Newton and Duda 1999). In addition,
intrinsic motivation has been shown to be associated with sport perform-
ance (Ryan et al. 1984). Autonomy support has also been linked with posi-
tive affect intrinsic motivation in athletes (Gagné et al. 2003). Together, this
evidence suggests that the fostering of intrinsic motivation or at least an
identified regulation by highlighting the importance of the goal to exercisers
and athletes, providing non-contingent positive feedback, and providing
Conclusion 223
activities and practices that enhance competence will positively influence

exercise behaviour.
One of the overarching themes among social psychological theories in exer-
cise and sport is the notion of generality of constructs. Throughout this book,
the distinction has been made between generalized, stable, and distal pre-
dictors of motivation and behaviour in theories of sport and exercise and spe-
cific, changeable, and proximal influences. Examples of state-like constructs
covered in this text are attitudes and intentions (Ajzen 1985, 1991) and com-
petitive state anxiety (Jones et al. 1993), while trait-like influences include
global intrinsic motivation (Vallerand 1997) and personality constructs
(Rhodes et al. 2002a). Theories that incorporate both types of construct often
arrange them hierarchically. Models that make this distinction explicitly
are the physical self-esteem models (e.g. Fox and Corbin 1989; Marsh and
Redmayne 1994) presented in Chapter 4 and Vallerand’s (1997) hierarchical
model of motivation presented in Chapters 3 and 5. These models have pro-
vided some evidence to suggest that there are proximal and distal forms of
motivation, some of which are changeable and alterable and others that are
less open to change and may serve as inter-individual covariates to be
accounted for in intervention studies aimed at promoting exercise behaviour
and sports performance. However, there is some recent evidence that such
arrangements may misrepresent the actual structure of these constructs. For
example, using cross-lagged panel designs Kowalski et al. (2003) and Guay et
al. (2003) have questioned the hierarchical nature of the physical self-esteem
and Vallerand’s (1997) model of motivation respectively. These studies were,
however, conducted over long time periods and more short-term panel designs
may shed more light on such arrangements.
Contrasts
In the introduction to this chapter, the contrasts between social psychological

approaches to exercise and sport have been made all too apparent and the
reasons for this, in some cases, are obvious. The objectives of sport and com-
petition are generally perceived to be at odds with those of exercise for health
reasons. This because sport, particularly high-level competitive sport, places
considerable motivational and physical demands on the participants. For
those whom participation in exercise is primarily for weight loss and the
benefits outlined in Chapter 1, such high demands are not only unrealistic,
but also likely to undermine motivation to participate and persist. However, it
must also be recognized that not all sport involves the competition and pres-
sure as found in elite sport, and for some individuals whose goals are to lose
weight or gain fitness, sport of the appropriate level, intensity, and type can be
instrumental in maintaining interest and persistence in exercise. In essence,
this relates to self-determination theory in that people with autonomous
motives are more likely to persist, and given that interest is one of the
224 Conclusion
characteristic features of autonomous forms of motivation, interesting

activities like sports that satisfy psychological needs are likely to foster
autonomous motives. Given that the role of sport, a specific subtype of phys-
ical activity, can be implicated in the process of fostering greater levels of
exercise behaviour among sedentary individuals because of the inherent inter-
est it may hold for such individuals, those involved with the promotion of
active lifestyles must be aware of the role of sport in leisure-time exercise
promotion. They must also, however, be aware of the conflicting messages
that may arise from the promotion of competitive sport to those who are not
interested in high-level competition.
This has been illustrated by the policy of promotion of sport and health-
related exercise in schools. Schools are recognized by health professionals,
sports promoters, and national governing bodies for sport as an important
existing network to promote exercise and sport among young people. In par-
ticular, health professionals are increasingly recognizing school physical edu-
cation as an opportunity to encourage an active lifestyle and regular exercise
for health benefits to young people, because they have a captive audience.
However, government organizations responsible for the development of sport
excellence also see schools as a hotbed of talent to be harnessed for future
success at national and international level. In addition, education policy-
makers often cite sport in physical education as an important conduit to pro-
mote the ideals of moral development and a productive work ethic in young
people. Such a view assumes that lessons learnt in sport participation can
generate positive behavioural traits in other domains such as academic work
and social maturity.
In a recent paper, Hagger and Weed (2000) have suggested that these three
roles for physical education can result in policies and curricula that lack
direction and convey conflicting messages for young people. Drawing on
policy documents put forward by UK Department of Education for the
National Curriculum in physical education and the UK government’s overall
strategy for sport, Hagger and Weed ask whether the ideals for sport promo-
tion such as talent identification and excellence and moral development can
co-exist with the promotion of health-related activity in school physical edu-
cation. They argue that the application of a model of sport performance in
physical education often ignores and undermines the competence and desire
for participation among children who do not display sport talent of the
highest level. They conclude that the challenge for physical education is to
provide alternative routes within physical education that permit the talented
to flourish while providing fun, interesting, and exciting opportunities for
the majority to participate in sport and exercise for a healthy and active
lifestyle. Such opportunities may then track into increased exercise outside
normal school hours and perhaps lead to persistence in later life (Hagger et
al. 2003b).
Conclusion 225
Interventions and practice
Drawing on the theories of motivation presented in this book, this section

highlights the commonality in the practical guidelines offered to exercisers
and athletes on the basis of research. The focus of this book has been on
applying social psychological research to exercise and sport behaviour, and, as
a result, practical guidelines across both disciplines are theory-based. Given
that there is considerable commonality in the theories and methods used to
explain phenomena in both the exercise and sport disciplines, it follows that
there will be some commonality in the strategies adopted by practitioners to
intervene. This section will briefly outline the practical recommendations
highlighted from the theoretical investigations in this book and demonstrate
that they share many of the techniques involved.
Exercise psychology practical guidelines

In an exercise context, research suggests that interventions to alter exercise
behaviour should focus on three specific aspects: (1) the antecedents of
intentions and motivation; (2) the factors that convert intentions into
behaviour; (3) and the underlying beliefs and motives that give rise to these
antecedents. Given the pervasive effects that attitudes and perceived
behavioural control, particularly self-efficacy, have on exercise behaviour from
various theoretical models (e.g. theory of planned behaviour), practitioners
would do well to target these antecedent variables (Hardeman et al. 2002). In
particular, interventions need to provide information about the advantages of
exercise, how people can overcome salient barriers, and the means available to
them to participate in regular exercise. Practitioners should also avoid an
approach that uses scare tactics or a controlling, authoritarian approach. This
is because intervention approaches such as motivational interviewing and
research on self-determination theory that focus on supporting people’s psy-
chological needs suggest that a controlling approach will undermine intrinsic
motivation and inhibit the formation of intentions to exercise. Such
approaches advocate the creation of a motivational climate that supports
intrinsic and identified motives for action. These autonomy-supportive
environments can be fostered by providing individuals with sensible, person-
ally relevant goals that are suggested by the individual, providing infor-
mational feedback, and encouraging the choice and selection of activities that
are interesting and satisfying.
Sports psychology practical guidelines

In sport, a number of practical guidelines for practitioners such as coaches,
trainers, and sports psychologists working with athletes have been adopted.
These are based on empirical studies examining the antecedents of sports
226 Conclusion
performance from theoretical perspectives such as social cognitive theory,

achievement goal theory, and self-determination theory. The focus of these
guidelines is on pragmatic recommendations that aim to increase athletes’
levels of motivation and adaptive outcomes such as self-esteem and com-
petence. Coaches should avoid making comparisons with others, not equate
performance outcomes with self-esteem, and handle and correct errors con-
structively and in an autonomy-supportive way. In addition, creating a task-
oriented motivational climate by providing challenges and handling failures,
setbacks, and mistakes constructively and avoiding language that undermines
autonomous motivation (e.g. using words like ‘should’ and ‘must’), being
positive in approach, and communicating and listening to athletes are corner-
stones of recommendations that theory states will generate interest, persist-
ence, and positive affective and psychological outcomes. In team sports,
research suggests that the coaches can foster greater team cohesion by being
relationship-oriented and trying to satisfy innate needs for relatedness. Spe-
cific behaviours that can assist in this are the forging of better relationships
with athletes by helping to solve problems, being empathic, and knowing the
team members as individuals. These comprehensive, theory-based guidelines
can result in improving intrinsic motivation, promoting a mastery-involved
goal orientation, and increase self-esteem in athletes.
Summary
• The application of social psychology to social processes and behaviour

in exercise and sport adopts similar perspectives in both disciplines.
These similarities can be classified in terms of themes, methods, theories,
contrasts, and practical guidelines.
• Common themes include the prediction of behaviour, the role and
prediction of affective outcomes, and the role and prediction of social
influences.
• Methods adopted across both disciplines include cross-sectional studies,
longitudinal, cross-lagged panel designs, experimental designs, and
qualitative approaches.
• The theories common to the exercise and sport disciplines include theories
of intention and motivation, and, in particular, self-determination theory.
• In terms of contrasts, the purposes of exercise for health and sport can
sometimes co-exist in that sports participants also have motives that are
health-related, but these may also conflict as in cases where competition
may undermine intrinsic motivation.
• Common to the interventions are the importance of an autonomy-
supportive context and the setting of appropriate goals.
Glossary
Achievement goal theory: Proposes that motivation is determined by an individual’s

global orientations regarding their perceived ability. Two orientations influence
motivation: a task orientation that represents a perception of ability as self-referenced,
and an ego orientation in which ability is perceived as other-referenced.
Activity disorder: Psychological condition in which individuals develop disordered
patterns of exercise characterized by extreme levels of physical activity; tends to be
accompanied by compromised psychological well-being and self-esteem.
Affect: Umbrella term referring to all types of ‘feeling states’. See emotion and mood.
Affective attitudes: Attitudes based on affective (emotional) beliefs about attitude
objects. See attitude.
Aggression: A set of behaviours that are likely to, or have the potential to, cause harm to
others, and are intended to cause harm.
Anticipated regret: Feelings and emotions that an individual would expect to
experience as a consequence of rejecting alternative behavioural decisions.
Anxiety: A negative affective state characterized by an inability to cope with the stress
placed on an individual by environmental demands.
Applied research: Empirical studies that examine the effects of social psychological
models and interventions in ecologically valid (everyday) contexts.
Arousal: A heightened sense of psychological and physiological awareness and
readiness.
Attitude: A general orientation, positive or negative, toward an attitude object. See
affective attitudes and cognitive attitudes.
Attitude strength: The strength of the association between a set of beliefs about an
attitude object and a target behaviour.
Attitudinal ambivalence: The holding of two potentially conflicting attitudinal beliefs
simultaneously.
Attribution theory: Weiner’s (Weiner et al. 1972) theory proposes that future sport
engagement and salient outcomes are determined by people’s attribution of the cause
of their behaviour to various sources that are perceived to be internal, stable, or
controllable.
Behavioural belief: Beliefs that engaging in a target behaviour will result in certain
salient outcomes. Along with outcome evaluation, this forms an expectancy-value model
of beliefs that underpins the attitude component of the theory of planned behaviour.
228 Glossary
Belief: A learned personal orientation regarding a person, object, or behaviour. See

attitude.
Catastrophe theory: Hardy’s (1990) theory of anxiety in sport that predicts non-linear
changes in the relationship between anxiety and performance with different levels of
arousal.
Choke: Term used to describe a catastrophic drop in sport performance in high-pressure
competition.
Cognitive attitudes: Attitudes based on instrumental beliefs about the utility of an
attitude object. See attitude and affective attitudes.
Cognitive evaluation theory: Sub-theory of self-determination theory that hypoth-
esizes that environmental contingencies such as rewards have a pervasive effect on
level of intrinsic motivation.
Collective aggression: Aggressive and violent behaviour of a group of people with
intent to cause injury or harm to another individual or group.
Collective efficacy: Beliefs in the ability of a group or team to produce certain desirable
outcomes such as win a point or execute an effective defence.
Conflict theory: Theory that suggests that violence in sports contexts reflects under-
lying social, political, and economic grievances held by the working-class element of
the sport crowd.
Continuation intention: A stated plan to engage in a behaviour like exercise even after
salient behavioural outcomes have been achieved.
Control beliefs: Beliefs that external factors will facilitate or inhibit behavioural
engagement. Along with perceived power, this forms the expectancy-value model of the
perceived behavioural control component of the theory of planned behaviour.
Correspondence rule: Boundary condition of the theory of planned behaviour dictating
that social cognitive predictors like attitudes and intentions must agree in terms of
action, target, context, and time.
Crowd violence: A form of collective aggression and violence between crowds of sports
spectators. See collective aggression.
Descriptive norms: Beliefs regarding the extent to which significant others engage in a
target behaviour. Normative construct included in augmented versions of the theory of
planned behaviour.
Desire: Reflects the motivational content of attitudes. See model of goal-directed behaviour.
Emotion: A specific feeling state often thought to have physiological (somatic) and
cognitive (interpretive) components.
Evaluation apprehension: An individual’s belief that an audience or crowd of
spectators are critically evaluating his/her performance of a behaviour.
Exercise: Formal form of physical activity encompassing all type of energy-expending
movement usually for health reasons.
Expectancy-value model: Model in which the effect of a social cognitive construct on a
behavioural or outcome variable (e.g. intention) is determined by the product of the
expectancy (e.g. behavioural belief) and the personal value attached to that belief (e.g.
outcome evaluation). Ajzen (1985) proposed that this model reflects the nature of the
belief systems that underpin the theory of planned behaviour and theory of reasoned
action constructs.
Figurational approach: Theory of crowd violence that proposes that aggression among
sport supporters arises because of violence is an acceptable means to resolve conflict
among the working classes. However, the violence seen among sports fans is ritualistic
and ordered and reflects a civilizing process.
Glossary 229
Formative research: Empirical research that focuses on the origins of social cognitive
constructs and behaviour.
Group cohesion: The social forces that maintain the attraction between members of a
group and the resistance of the group to disruptive forces.
Group dynamics: The study of the effect of group membership on group members’
perceptions and behaviour.
Group norm: The accepted roles and ways of behaving endorsed by group members.
Health belief model: A social cognitive theory of health behaviour that hypothesizes
that an individual’s readiness or intention to perform a health behaviour is function
of his/her perceived vulnerability to a health condition and the probable severity of
that condition.
Hierarchical model: Refers to a hierarchical organization of a psychological construct
from different levels of generality and implies both bottom-up and top-down patterns
of influence. Vallerand’s (1997) hierarchical models of self-determined motivation
and Fox and Corbin’s (1989) hierarchical model of physical self-perceptions are
examples.
Home advantage: Phenomenon in which a sport team has a significantly better chance
of winning when playing at their home ground or venue, although some argue this
may be indicative of an ‘away disadvantage’.
Hooliganism: A brand of sport violence usually reserved for conflict among soccer fans
in Europe, although some argue that this term also refers to non-violent extrovert and
outrageous behaviours.
Hostile aggression: Aggression for which the primary purpose is injurious to others and
coupled with high emotional arousal. Should be contrasted to instrumental aggression.
Implementation intention: Volitional strategy in which the link between intention and
behaviour is enhanced by the creation of cues (e.g. when and where to perform the
behaviour) that permit the automatic enactment of the behaviour once they are
encountered.
Individualized zones of optimal functioning (IZOF): Hanin’s (2000) eclectic theory
that has been primarily described as a theory of anxiety and hypothesizes that athletes
have an individual level of anxiety at which they will exhibit optimal performance.
Instrumental aggression: Aggressive acts that may result in injury, but the intent to
injure is peripheral to the performance advantage the aggressive act yields.
Intention: Stated plans to act and the most proximal predictor of behaviour in the
theory of planned behaviour. See theory of planned behaviour.
Intention stability: Consistency in intentions over time, thought to moderate the inten-
tion–behaviour relationship in the theory of planned behaviour.
Intervention: A strategy implemented to change a number of (usually) psychological
variables and examine their effect of perceptions and behaviour.
Intrinsic motivation: Motivation to participate in social behaviours with no discernible
external reinforcement. Intrinsic motivation is often characterized as behaviours that
are done out of enjoyment, fun, interest, a sense of choice, and personal autonomy.
Inverted-U hypothesis: Hypothesis that athletes have an ‘optimal’ level of arousal at
which performance is maximized, while lower or higher levels of arousal lead to
progressively inferior performance.
Meta-analysis: Statistical technique that aims to find the average effect size (relation-
ship or difference) between key variables from tests of the effect across a number of
empirical studies while correcting for sampling (e.g. sample size) and measurement
(e.g. reliability) error.
230 Glossary
Model of goal-directed behaviour: Social cognitive theory of volitional behaviour that

hypothesizes desires to be the most proximal determinant of goal intentions. In the
model, intentions are viewed as intentions to perform a target behaviour to achieve a
specific goal and are the most proximal predictor of behaviour. Attitudes, subjective
norms, perceived behavioural control, and anticipated positive and negative affect all
influence intentions via the mediation of desires.
Mood: A feeling state that is longer lasting and less intense than emotion, often viewed
from a ‘profile’ perspective.
Motivation: A social cognitive construct that activates, energizes, or drives behaviour
and determines its intensity, direction, and persistence of a behaviour.
Motivation to comply: Individuals’ beliefs that they will generally comply with the
wishes of their significant others. With normative beliefs, this construct forms the
expectancy-value belief system that underpins the subjective norm component of
the theory of planned behaviour.
Motivational climate: The environmental context operating in achievement settings.
A mastery motivational climate tends to engender a task-oriented motivational
orientation while performance motivational climate tends to foster an ego-oriented
motivational orientation.
Motivational interviewing: Intervention technique that aims to change exercise
behaviour by focusing on client-centred reasons for change, reducing ambivalence,
increasing self-efficacy, and reducing resistance to change.
Normative Beliefs: Individuals’ beliefs that significant others would want them to par-
ticipate in the target behaviour. With motivation to comply, this construct forms the
expectancy-value belief system that underpins the subjective norm component of the
theory of planned behaviour.
Organismic integration theory: A sub-theory of self-determination theory that
hypothesizes that externally regulated behaviours can be integrated into an indi-
vidual’s repertoire of behaviours that service personally salient goals or outcomes. This
integration is met with a shift in the perceived locus of causality from external to
internal.
Outcome evaluation: Individuals’ beliefs that certain behavioural outcomes are
positive or negative. With behavioural belief, this construct forms the expectancy-
value belief system that underpins the attitude component of the theory of planned
behaviour.
Past behaviour: Frequency or recency of past behaviour is considered to reflect the
influence of unmeasured constructs in social cognitive models like the theory of
planned behaviour and the model of goal-directed behaviour. Controversially, it is
also said to reflect more automatic routes to behavioural engagement.
Perceived autonomy support: An individual’s evaluation of the extent to which their
environment supports their intrinsic motivation.
Perceived behavioural control: Social cognitive construct that reflects an individual’s
degree of subjective control over a target behaviour in the theory of planned behaviour.
Hypothesized to affect behaviour directly and indirectly via the mediation of
intention.
Perceived locus of causality: Continuum that reflects the source of the reason for doing
behaviour in a given context. Intrinsic motivation and external regulation lie at the
extremes of the continuum with identified regulation and introjected regulation
reflecting intermediate levels of autonomous and controlling regulation respectively.
See organismic integration theory.
Glossary 231
Perceived power: The extent that external factors affect behavioural engagement.
Along with control beliefs, this forms the expectancy-value model of the perceived
behavioural control component of the theory of planned behaviour.
Personality: Trait or trait-like constructs that are general and stable in nature and are
hypothesized to affect a variety of behaviours across a number of contexts.
Physical activity: Umbrella term that encompasses all energy-expending body
movement.
Physical self-esteem: An individual’s evaluation of their self in physical contexts.
Protection motivation theory: A social cognitive theory that hypothesizes that
engagement in health behaviour is function of threat and coping appraisals.
Psychological social psychology: Approach to social psychological investigation that
adopts an empirical, hypothesis-testing framework based on the principles of falsifica-
tion and converging evidence.
Reversal theory: Theory of motivation and emotion that states that an individual’s
behaviour is determined by a series of metamotivational states. These states dictate
whether the individual experiences a positive or negative affective state or hedonic
tone and the level of arousal associated with the behaviour.
Role ambiguity: The degree of clarity that an individual has regarding his role in a sport
team or group.
Role conflict: The degree to which an individual perceives they are unable to meet the
demands of their established or assigned role in a sport team or group.
Role efficacy: An individual’s estimate of their ability to perform the behaviours
expected to successfully fulfil their role in the sport team or group.
Self-determination theory: Influential theory of motivation that hypothesizes that
human motivation is determined by the goal of satisfying the basic psychological
needs of autonomy (see intrinsic motivation), competence, and relatedness. Self-
determination theory concerns itself with the environmental conditions that give rise
to intrinsically motivated behaviour (see cognitive evaluation theory) and the way in
which extrinsically motivated behaviours (see perceived locus of causality) become inte-
grated into an individual’s repertoire of behaviours that satisfy psychological needs
(see organismic integration theory).
Self-efficacy: Situation-specific self-confidence.
Social cognition: Approach to social psychology that assumes individuals are rational
decision-makers who process information from their social environment and past
experience prior to making decisions to act.
Social cognitive theory: Bandura’s (1977) theory that proposed that behaviour was a
function of environmental reinforcements (presence of models) and internal factors
(such as observational learning).
Social facilitation: The study of the effects of audiences and coaction on individual
performance, particularly audiences and other competitors on sport performance.
Social identity theory: Tajfel and Turner’s (1986) theory that hypothesizes individual
group members’ behaviour to be affected by their own categorization of themselves as
stereotypical group members and their associated attributes.
Social learning theory: Branch of Bandura’s (1986) social cognitive theory that focuses on
how individuals’ observation of others’ behaviour can reinforce their own behaviours.
Social loafing: Phenomenon in which an individual in a group exhibits a decrement in
their personal performance due to the perception that they are unable to contribute to
the group effort and their individual contribution is not being evaluated.
Sociological social psychology: Approach to social psychology that focuses on theory
232 Glossary
building rather than theory testing through phenomenological accounts from

individuals regarding behaviour.
Sport: Competitive endeavours that usually have a system of governing rules. Many
sports have a physical element to them that may yield benefits similar to exercise.
Subjective norm: Social cognitive construct that reflects an individual’s perceived belief
that significant others want them to engage in a target behaviour like exercise. In the
theory of planned behaviour, subjective norm is hypothesized to influence behaviour via
the mediation of intentions.
Theory of planned behaviour: Often cited theory of intentional behaviour that
hypothesizes that intention is the most proximal predictor of volitional behaviour and
mediates the influence of a set of personal (see attitude), social (see subjective norms),
and control-related (see perceived behavioural control) beliefs on behaviour. Ajzen
(1985) hypothesizes that perceived behavioural control also influences behaviour
directly.
Theory of reasoned action: Precursor to the theory of planned behaviour. Hypothesizes
that volitional behaviour is influenced by the indirect effect of attitudes and subjective
norms via the mediation of intentions (Ajzen & Fishbein 1980).
Trans-contextual model: Multi-theory model that hypothesizes that an adolescent’s
perceived autonomy support in a physical education context influences his/her attitudes,
perceived behavioural control, and intention from the theory of planned behaviour and
exercise behaviour via the mediation of autonomous motives in physical education
and leisure-time contexts.
Volitional theories: Theories that aim to examine the processes by which intentions are
translated into actions.
Bibliography
Abraham, C., Clift, S. and Grabowski, P. (1999) Cognitive predictors of adherence to

malaria prophylaxis regimens on return from a malarious region: a prospective study.
Social Science and Medicine, 48: 1641–54.
Abraham, C. and Sheeran, P. (2003) Acting on intentions: the role of anticipated regret.
British Journal of Social Psychology, 42: 495–511.
Abraham, C. and Sheeran, P. (2004) Deciding to exercise: the role of anticipated regret.
British Journal of Health Psychology, 9: 269–78.
Ackard, D.M., Croll, J.K. and Kearney-Cooke, A. (2002) Dieting frequency among college
females: association with disordered eating, body image and related psychological
problems. Journal of Psychosomatic Research, 52: 129–36.
Adams, J.M., Miller, T.W. and Kraus, R.F. (2003) Exercise dependence: diagnostic and
therapeutic issues for patients in psychotherapy. Journal of Contemporary Psycho-
therapy, 33: 93–107.
Agnew, G.A. and Carron, A.V. (1994) Crowd effects and the home advantage. Inter-
national Journal of Sport Psychology, 25: 53–62.
Aiello, J.R. and Douthitt, E.A. (2001) Social facilitation from Triplett to electronic per-
formance monitoring. Group Dynamics, 5, 163–80.
Ajzen, I. (1985) From intentions to actions: a theory of planned behavior. In J. Kuhl and
J. Beckmann (eds), Action – control: From Cognition to Behavior (pp. 11–39). Heidelberg:
Springer.
Ajzen, I. (1991) The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50: 179–211.
Ajzen, I. (2002a) Perceived behavioral control, self-efficacy, locus of control, and the
theory of planned behavior. Journal of Applied Social Psychology, 32: 1–20.
Ajzen, I. (2002b) Residual effects of past on later behavior: habituation and reasoned
action perspectives. Personality and Social Psychology Review, 6: 107–22.
Ajzen, I. and Fishbein, M. (1980) Understanding Attitudes and Predicting Social Behavior.
Englewood Cliffs, NJ: Prentice Hall.
Allen, J.B. (2003) Social motivation in youth sport. Journal of Sport and Exercise
Psychology, 25: 551–67.
Allport, F.H. (1920) The influence of the group upon association and thought. Lancet, 60:
159–82.
234 Bibliography
American College of Sports Medicine. (1998) ACSM Position stand: exercise and physical
activity for older adults. Medicine and Science in Sports and Exercise, 30: 992–1008.
American Heart Association. (1999) AHA/ACC scientific statement: assessment of
cardiovascular risk by use of multiple-risk-factor assessment equations. Circulation,
100: 1481–92.
American Psychiatric Association. (1994) Diagnostic and Statistical Manual of Mental
Disorders (revised 4th edn). Washington, DC: American Psychiatric Association.
American Psychiatric Association Work Group on Eating Disorders. (2000) Practice
guideline for the treatment of patients with eating disorders (revision) American
Journal of Psychiatry, 157 (Suppl.): S1 – S39.
Ames, C. (1992) Classrooms: goals, structures, and student motivation. Journal of
Educational Psychology, 84: 261–71.
Andersen, A.E. (1995) Eating disorders in males. In K.D. Brownell and C.G. Fairburn
(eds), Eating Disorders and Obesity: A Comprehensive Handbook (pp. 177–87). New York:
Guilford Press.
Apter, M. (1982) The Experience of Motivation: The Theory of Psychological Reversals. New
York: Academic Press.
Armitage, C. and Arden, M.A. (2002) Exploring discontinuity patterns in the trans-
theoretical model: an application of the theory of planned behaviour. British Journal of
Health Psychology, 7: 89–103.
Armitage, C.J. and Conner, M. (1999a) Predictive validity of the theory of planned
behaviour: the role of questionnaire format and social desirability. Journal of Com-
munity and Applied Social Psychology, 9: 261–72.
Armitage, C.J. and Conner, M. (1999b) The theory of planned behavior: assessment of
predictive validity and ‘perceived control’. British Journal of Social Psychology, 38:
35–54.
Armitage, C.J. and Conner, M. (2000) Attitudinal ambivalence: a test of three key
hypotheses. Personality and Social Psychology Bulletin, 26: 1421–32.
Armitage, C.J. and Conner, M. (2001) Efficacy of the theory of planned behaviour: a
meta-analytic review. British Journal of Social Psychology, 40: 471–99.
Armitage, C.J., Povey, R. and Arden, M.A. (2003) Evidence for discontinuity patterns
across the stages of change: a role for attitudinal ambivalence. Psychology and Health,
18: 373–86.
Asçi, F.H., Asçi, A. and Zorba, E. (1999) Cross-cultural validity and reliability of the
physical self-perception profile. International Journal of Sport Psychology, 30,:399–406.
Ashford, B., Biddle, S. and Goudas, M. (1993) Participation in community sports centres:
motives and predictors of enjoyment. Journal of Sport Sciences, 11: 249–56.
Bagozzi, R.P., Gürhan-Canli, Z. and Priester, J.R. (2002) The Social Psychology of Consumer
Behaviour. Buckingham: Open University Press.
Bagozzi, R.P. and Yi, Y. (1989) The degree of intention formation as a moderator of the
attitude-behavior relationship. Social Psychology Quarterly, 52: 266–79.
Baker, C.W., Little, T.D. and Brownell, K.D. (2003) Predicting adolescent eating and
activity behaviors: the role of social norms and personal agency. Health Psychology, 22:
189–98.
Bandura, A. (1977) Self-efficacy: toward a unifying theory of behavioral change.
Psychological Review, 84: 191–215.
Bandura, A. (1994) Self-efficacy. In V.S. Ramachaudran (ed.), Encyclopedia of Human
Behavior (Vol. 4, pp. 71–81). New York: Academic Press.
Bandura, A. (1997) Self-efficacy: The Exercise of Control. New York: Freeman.
Bibliography 235
Bargh, J.A. (1994) The Four Horsemen of automaticity: awareness, efficiency, intention,
and control in social cognition. In R.S. Wyer and T.K. Srull (eds), Handbook of Social
Cognition (Vol. 2, pp. 1–40). Hillsdale, NJ: Erlbaum.
Baron, R.S. (1986) Distraction-conflict theory: progress and problems. Advances in
Experimental Social Psychology, 19: 1–36.
Baumeister, R.F. (1984) Choking under pressure: self-consciousness and paradoxical
effects of incentives on skillful performance. Journal of Personality and Social Psychology,
46: 610–20.
Baumeister, R.F. (1995) Disputing the effects of championship pressures and home
audiences. Journal of Personality and Social Psychology, 68: 644–8.
Baumeister, R.F. and Leary, M.R. (1995) The need to belong: desire for interpersonal
attachments as a fundamental human motivation. Psychological Bulletin, 117: 497–529.
Baumeister, R.F. and Steinhilber, A. (1984) Paradoxical effects of supportive audiences on
performance under pressure: the home field disadvantage in sports championships.
Journal of Personality and Social Psychology, 47: 85–93.
Beauchamp, M.R. and Bray, S.R. (2001) Role ambiguity and role conflict within inter-
dependent teams. Small Group Research, 32: 133–57.
Beauchamp, M.R., Bray, S.R., Eys, M.A. and Carron, A.V. (2002) Role ambiguity, role
efficacy, and role performance: multidimensional and mediational relationships
within interdependent sport teams. Group Dynamics: Theory, Research and Practice, 6:
229–42.
Beauchamp, M.R., Bray, S.R., Eys, M.A. and Carron, A.V. (2003) Effect of role ambiguity
on competitive state anxiety. Journal of Sport and Exercise Psychology, 25: 77–92.
Becker, A.E., Grinspoon, S.K., Klibanski, A. and Herzog, D.B. (1999) Eating disorders. New
England Journal of Medicine, 340: 1092–8.
Beedie, C.J., Terry, P.C. and Lane, A.M. (2000) The profile of mood states and athletic
performance: two meta – analyses. Journal of Applied Sport Psychology, 12: 49–68.
Bell, P.A. and Yee, LA. (1989) Skill level and audience effects on performance of a karate
drill. Journal of Social Psychology, 129: 191–200.
Bellew, E. and Thatcher, J. (2002) Metamotivational state reversals in competitive sport.
Social Behavior and Personality, 30: 613–23.
Bennett, J.C. (1991) The irrationality of catharsis theory of aggression as a justification
for educators’ support for interscholastic football. Perceptual and Motor Skills, 72:
415–18.
Berger, B.G. and Motl, R.W. (2000) Exercise and mood: a selective review and synthesis of
research employing the profile of mood states. Journal of Applied Sport Psychology, 12:
69–92.
Berglund, B. and Safstrom, H. (1994) Psychological monitoring and modulation of
training load of world-class canoeists. Medicine and Science in Sports and Exercise, 26:
1036–40.
Berkowitz, L. (1993) Aggression: Its Causes, Consequences, and Control. Philadelphia, PA:
Temple University Press.
Biasi, V. (1999) Personological studies on dancers: motivations, conflicts and defense
mechanisms. Empirical Studies of the Arts, 17: 171–86.
Biddle, S.J.H. (1997) Cognitive theories of motivation and the physical self. In K. R. Fox
(ed.), The Physical Self (pp. 59–82). Champaign, IL: Human Kinetics.
Biddle, S.J.H. (1999) Motivation and perceptions of control: tracing its development
and plotting its future in exercise and sport psychology. Journal of Sport and Exercise
236 Bibliography
Biddle, S.J.H., Akande, D., Armstrong, N., Ashcroft, M., Brooke, R. and Goudas, M.
(1996) The self-motivation inventory modified for children: evidence on psycho-
metric properties and its use in physical exercise. International Journal Sport Psychology,
27: 237–50.
Biddle, S.J.H., Hanrahan, S.J. and Sellars, C.N. (2001) Attributions: past, present, and
future. In R.N. Singer, H.A. Hausenblas and C.M. Janelle (eds), Handbook of Sport
Psychology (pp. 444–71). New York: Wiley.
Bond, C.F. and Titus, L.J. (1983) Social facilitation: a meta-analysis of 241 studies.
Psychological Bulletin, 94: 265–92.
Bond, K.A., Biddle, S.J.H. and Ntoumanis, N. (2001) Self-efficacy and causal attribution
in female golfers. International Journal of Sport Psychology, 32: 243–56.
Bornstein, R.F. (2001) A meta-analysis of the dependency eating-disorders relationship:
strength, specificity, and temporal stability. Journal of Psychopathology and Behavioral
Assessment, 23: 151–62.
Boyd, M.P., Weinmann, C. and Yin, Z. (2002) The relationship of physical self-
perceptions and goal orientations to intrinsic motivation for exercise. Journal of Sport
Behavior, 25: 1–18.
Boyd, M.P., Yin, Z., Ellis, D. and French, K. (1995) Perceived motivational climate,
socialization influences, and affective responses in Little League Baseball. Journal of
Sport and Exercise Psychology, 17 (Suppl.): S30.
Bozionelos, G. and Bennett, P. (1999) The theory of planned behaviour as predictor of
exercise: the moderating influence of beliefs and personality variables. Journal of
Health Psychology, 4: 517–29.
Branca, F. (1999) Physical activity, diet and skeletal health. Public Health Nutrition, 2
(Suppl.): S391 – S396.
Brawley, L.R. (1993) The practicality of using psychological theories for exercise and
health research and intervention. Journal of Applied Sport Psychology, 5: 99–115.
Brawley, L.R., Carron, A.V. and Widmeyer, W.M. (1988) Exploring the relationship
between cohesion and group resistance to disruption. Journal of Sport and Exercise
Bray, C.D. and Whaley, D.E. (2001) Team cohesion, effort, and objective individual
performance of high school basketball players. The Sport Psychologist, 15: 260–75.
Bray, R.M. and Sugarman, R. (1980) Social facilitation among interacting groups:
Evidence for the evaluation-apprehension hypothesis. Personality and Social Psychology
Bulletin, 6: 137–42.
Bray, S.R. (1999) The home advantage from an individual team perspective. Journal of
Applied Sport Psychology, 11: 116–25.
Bray, S.R. (2004) Collective efficacy, group goals, and group performance of a muscular
endurance task. Small Group Research, 35: 230–8.
Bray, S.R. and Brawley, L.R. (2002) Role efficacy, role clarity, and role performance
effectiveness. Small Group Research, 33: 233–53.
Bray, S.R., Brawley, L.R. and Carron, A.V. (2002) Efficacy for interdependent role
functions: evidence from the sport domain. Small Group Research, 33: 644–66.
Bray, S.R. and Martin, KA. (2003) The effect of competition location on individual
athlete performance and psychological states. Psychology of Sport and Exercise, 4:
117–23.
Breaux, C.A. and Moreno, J.K. (1994) Comparing anorectics and bulimics on measures of
depression, anxiety, and anger. Eating Disorders: The Journal of Treatment and Prevention,
2: 158–67.
Bibliography 237
Bredemeier, B.J.L. and Shields, D.L. (1986) Moral growth among athletes and non-
athletes: a comparative analysis. Journal of Genetic Psychology, 147: 7–18.
Brown, T.D., van Raalte, J.L., Brewer, B.W., Winter, C.R., Cornelius, A.E. and Andersen,
M.B. (2002) World cup soccer home advantage. Journal of Sport Behavior, 25: 134–44.
Bundred, P., Kitchiner, D. and Buchan, I. (2001) Prevalence of overweight and obese
children between 1989 and 1998: population-based series of cross-sectional studies.
British Medical Journal, 322: 326–8.
Burton, D. (1988) Do anxious swimmers swim slower? Reexamining the elusive anxiety-
performance relationship. Journal of Sport and Exercise Psychology, 10: 45–61.
Burton, D. (1998) Measuring competitive state anxiety. In J.L. Duda (ed.), Advances in
Sport and Exercise Psychology Measurement (pp. 129–48). Morgantown, WV: Fitness
Information Technology.
Bushman, B.J., Baumeister, R.F. and Stack, A.D. (1999) Catharsis, aggression, and per-
suasive influence: self-fulfilling or self-defeating prophecies? Journal of Personality and
Social Psychology, 76: 367–76.
Bushman, B.J. and Cooper, H.M. (1990) Effects of alcohol on human aggression: an
integrative research review. Psychological Bulletin, 107: 341–54.
Butler, R. (1987) Task-involving and ego-involving properties of evaluation: effects of
different feedback conditions on motivational perceptions, interest, and performance.
Journal of Educational Psychology, 79: 474–82.
Byers, T., Nestle, M., McTiernan, A., Doyle, C., Currie-Williams, A., Gansler, T. and Thun,
M. (2002) American Cancer Society Guidelines on Nutrition and Physical Activity for
cancer prevention: reducing the risk of cancer with healthy food choices and physical
activity. CA – Cancer Journal of Clinicians, 52: 92–119.
Cale, L. and Almond, L. (1992) Children’s physical activity levels: a review of studies
conducted on British children. Physical Education Review, 15: 111–18.
Cameron, J. and Pierce, W.D. (1994) Reinforcement, reward, and intrinsic motivation: a
meta-analysis. Review of Educational Research, 64: 363–423.
Carnahan, B.J., Shea, J.B. and Davis, G.S. (1990) Motivational cue effects on bench-press
performance and self-efficacy. Journal of Sport Behavior, 13: 240–54.
Carnes, M. and Sachs, M. (2002, October 9–12) Too much of a good thing. Paper pre-
sented at the Addiction Medicine Review Course 2002, Newport Beach, California.
Carron, A.V. (1982) Cohesiveness in sport groups: interpretations and considerations.
Journal of Sport Psychology, 4: 123–38.
Carron, A.V. and Hausenblas, H.A. (1998) Group Dynamics in Sport (2nd edn). Morgan-
town, WV: Fitness Information Technology.
Carron, A.V., Widmeyer, W.M. and Brawley, L.R. (1985) The development of an
instrument to assess cohesion in sport teams: the group environment questionnaire.
CDC/NCHS (2001) National Health and Nutrition Examination Survey. Atlanta, GA:
National Center for Disease Control and Prevention and National Center for Health
Statistics.
Centers for Disease Control and Prevention. (2002) 2002 BRFSS Summary Prevalence
Report. Atlanta, GA: US Department of Health and Human Services, Public Health
Service, Centers for Disease Control and Prevention, National Center for Chronic
Disease Prevention and Health Promotion.
Centers for Disease Control and Prevention. (2003) Physical activity levels among
children aged 9–13 years: United States, 2002. Mortality and Morbidity Weekly, 52: 785–8.
Chaiken, S. (1980) Heuristic versus systematic information processing and the use of
238 Bibliography
source versus message cues in persuasion. Journal of Personality and Social Psychology,
39: 752–66.
Chase, M.A., Lirgg, C.D. and Feltz, D.L. (1997) Do coaches’ efficacy expectations for their
teams predict team performance? The Sport Psychologist, 11: 8–23.
Chatzisarantis, N.L.D. and Biddle, S.J.H. (1998) Functional significance of psychological
variables that are included in the theory of planned behaviour: a self-determination
theory approach to the study of attitudes, subjective norms, perceptions of control
and intentions. European Journal of Social Psychology, 28: 303–22.
Chatzisarantis, N.L.D., Biddle, S.J H. and Meek, G.A. (1997) A self-determination theory
approach to the study of intentions and the intention-behaviour relationship in
children’s physical activity. British Journal of Health Psychology, 2: 343–60.
Chatzisarantis, N.L.D., Hagger, M.S., Biddle, S.J.H. and Karageorghis, C. (2002) The
cognitive processes by which perceived locus of causality predicts participation in
physical activity. Journal of Health Psychology, 7: 685–99.
Chatzisarantis, N.L.D., Hagger, M.S., Biddle, S.J.H. and Smith, B. (in press a) The stability
of the attitude-intention relationship in the context of physical activity. Journal of
Sport Sciences.
Chatzisarantis, N.L.D., Hagger, M.S., Biddle, S.J.H., Smith, B.M. and Sage, L. (in press b)
The influences of perceived autonomy support on physical activity within the theory
of planned behaviour. Journal of Sport and Exercise Psychology.
Chatzisarantis, N.L.D., Hagger, M.S., Biddle, S.J.H., Smith, B. and Wang, J.C.K. (2003) A
meta-analysis of perceived locus of causality in exercise, sport, and physical education
contexts. Journal of Sport and Exercise Psychology, 25: 284–306.
Chatzisarantis, N.L.D., Hagger, M., Smith, B. and Phoenix, C. (in press c) The influences
of continuation intentions on the execution of social behaviour within the theory of
planned behaviour. British Journal of Social Psychology.
Chatzisarantis, N.L.D., Hagger, M.S., Smith, B. and Sage, L.D. (in press d) The influences
of intrinsic motivation on execution of social behaviour within the theory of planned
behaviour. European Journal of Social Psychology.
Clarke, J. (1978) Football and the working class fans: tradition and change. In R. Ingham
(ed.), Football Hooliganism: The Wider Context. London: Interaction.
Coakley, J.J. (1981) The sociological perspective: alternate causations of violence in
sport. Arena, 5: 44–56.
Cockerill, I.M., Neville, A.M. and Lyons, N. (1991) Modelling mood states in athletic
performance. Journal of Sports Sciences, 9: 205–12.
Cockerill, I.M. and Riddington, M.E. (1996) Exercise dependence and associated
disorders: a review. Counselling Psychology Quarterly, 9: 119–29.
Colman, M.M. and Carron, A.V. (2000) The group norm for productivity in individual
sport teams. Journal of Sport and Exercise Psychology, 22 (Suppl.): S27 – S28.
Colman, M.M. and Carron, A.V. (2001) The nature of norms in individual sport teams.
Small Group Research, 32: 206–22.
Connelly, D. (1988) Increasing intensity of play of nonassertive athletes. The Sport Psych-
ologist, 2: 255–65.
Conner, M. and Abraham, C. (2001) Conscientiousness and the theory of planned
behavior: toward a more complete model of the antecedents of intentions and
behavior. Personality and Social Psychology Bulletin, 27: 1547–61.
Conner, M. and Armitage, C.J. (1998) Extending the Theory of Planned Behavior:
a review and avenues for further research. Journal of Applied Social Psychology, 28:
1429–64.
Bibliography 239
Conner, M., Povey, R., Sparks, P., James, R. and Shepherd, R. (2003) Moderating role of
attitudinal ambivalence within the theory of planned behaviour. British Journal of
Conner, M., Sheeran, P., Norman, P. and Armitage, C. (2000) Temporal stability as a
moderator of relationships in the theory of planned behaviour. British Journal of Social
Conner, M., Sherlock, K. and Orbell, S. (1998) Psychosocial determinants of ecstasy use
in young people in the UK. British Journal of Social Psychology, 3: 295–317.
Cottrell, N.B. (1972) Social facilitation. In C.G. McClintock (ed.), Experimental Social
Psychology (pp. 185–236). New York: Holt, Rinehart and Winston.
Courneya, K.S. (1995) Understanding readiness for regular physical activity in older
individuals: an application of the Theory of Planned Behavior. Health Psychology, 14:
80–7.
Courneya, K.S. and Carron, A.V. (1991) Effects of travel and length of home stand/road
trip on the home advantage. Journal of Sport and Exercise Psychology, 13: 42–9.
Courneya, K.S. and Friedenreich, C.M. (1997) Relationship between exercise pattern
across the cancer experience and current quality of life in colorectal cancer survivors.
Journal of Alternative and Complementary Medicine, 3: 215–26.
Courneya, K.S., Friedenreich, C.M., Sela, R.A., Quinney, H.A. and Rhodes, R.E. (2002)
Correlates of adherence and contamination in a randomized controlled trial of exer-
cise in cancer survivors: an application of the theory of planned behavior and the five
factor model of personality. Annals of Behavioral Medicine, 24: 257–68.
Courneya, K.S. and McAuley, E. (1994) Factors affecting the intention-physical
activity relationship: intention versus expectation and scale correspondence. Research
Quarterly for Exercise and Sport, 65: 280–5.
Courneya, K.S., Plotnikoff, R.C., Hotz, S.B. and Birkett, N.J. (2000) Social support and the
theory of planned behavior in the exercise domain. American Journal of Health
Cox, R.H., Martens, M.P. and Russell, W.D. (2003) Measuring anxiety in athletics: The
Revised Competitive State Anxiety Inventory-2. Journal of Sport and Exercise Psychology,
25: 519–33.
Craft, L.L., Magyar, T.M., Becker, B.J. and Feltz, D.L. (2003) The relationship between
the Competitive State Anxiety Inventory-2 and sport performance: a meta-analysis.
Journal of Sport and Exercise Psychology, 25: 44–65.
Crocker, J. and Luhtanen, R.K. (2003) Level of self-esteem and contingencies of self-
worth: unique effects on academic, social, and financial problems in college students.
Personality and Social Psychology Bulletin, 29: 710–12.
Crocker, P.R.E., Alderman, R.B. and Smith, F.M.R. (1988) Cognitive-affective stress man-
agement training with high performance youth volleyball players: effects on affect,
cognition, and performance. Journal of Sport and Exercise Psychology, 10: 448–60.
Dashiell, J.F. (1930) An experimental analysis of some group effects. Journal of Abnormal
and Social Psychology, 25: 190–9.
Davidson, R. (1992) Prochaska and DiClemente’s model of change: a case study? British
Journal of Addiction, 87: 1–2.
Davis, C. (1999) Excessive exercise and anorexia nervosa: addictive and compulsive
behaviors. Psychiatric Annals, 29: 221–4.
Deci, E.L., Betley, G., Kahle, J., Abrams, L. and Porac, J. (1981) When trying to win:
competition and intrinsic motivation. Personality and Social Psychology Bulletin, 7:
79–83.
240 Bibliography
Deci, E.L., Eghrari, H., Patrick, B.C. and Leone, D.R. (1994) Facilitating internalization:
the self-determination theory perspective. Journal of Personality, 62: 119–42.
Deci, E.L., Koestner, R. and Ryan, R.M. (1999a) A meta-analytic review of experiments
examining the effects of extrinsic rewards on intrinsic motivation. Psychological
Bulletin, 125: 627–68.
Deci, E.L., Koestner, R. and Ryan, R.M. (1999b) The undermining effect is a reality after
all: extrinsic rewards, task interest, and self-determination: Reply to Eisenberger, Pierce
and Cameron (1999) and Lepper, Henderlong and Gingras (1999). Psychological
Bulletin, 125: 692–700.
Deci, E.L. and Ryan, R.M. (1985) Intrinsic Motivation and Self-determination in Human
Behavior. New York: Plenum Press.
Deci, E.L. and Ryan, R.M. (2000) The ‘what’ and ‘why’ of goal pursuits: human needs and
the self-determination of behavior. Psychological Inquiry, 11: 227–68.
Department of Health (1996) Strategy Statement on Physical Activity. London: Department
of Health.
Devine, D.J. (1999) Effects of cognitive ability, task knowledge, information sharing, and
conflict on group decision-making effectiveness. Small Group Research, 30: 608–34.
Diener, E. (1976) Effects of prior destructive behavior, anonymity, and group presence
on deindividuation and aggression. Journal of Personality and Social Psychology, 33:
497–507.
Dollard, J., Doob, L., Miller, N., Mowrer, O.W. and Sears, R.R. (1939) Frustration and
Aggression. New Haven, CT: Yale University Press.
Duda, J.L. (1993) Goals: a social cognitive approach to the study of achievement motiv-
ation in sport. In R.N. Singer, M. Murphey and L.K. Tennant (eds), Handbook of
Research on Sport Psychology (pp. 421–36). New York: Macmillan.
Duda, J.L. (2001) Achievement goal research in sport: pushing the boundaries and
clarifying some misunderstandings. In G.C. Roberts (ed.), Advances in Motivation in
Sport and Exercise (pp. 129–82). Champaign, IL: Human Kinetics.
Duda, J.L., Chi, L., Newton, M.L., Walling, M.D. and Catley, D. (1995) Task and ego
orientation and intrinsic motivation in sport. International Journal of Sport Psychology,
26: 40–63.
Duda, J.L. and Nicholls, J.G. (1992) Dimensions of achievement motivation in
schoolwork and sport. Journal of Educational Psychology, 84: 290–9.
Duncan, T. and McAuley, E. (1987) Efficacy expectations and perceptions of causality in
motor performance. Journal of Sport Psychology, 9: 385–93.
Dunning, E., Murphy, P. and Williams, J. (1988) The Roots of Football Hooliganism.
London: Routledge.
Dweck, C.S. (1992) The study of goals in psychology. Psychological Science, 3: 165–7.
Eagly, A.H. and Chaiken, S. (1993) The Psychology of Attitudes. San Diego, CA: Harcourt,
Brace and Jovanovich.
Edwards, T. and Hardy, L. (1996) The interactive effects of intensity and direction of
cognitive and somatic anxiety and self-confidence upon performance. Journal of Sport
and Exercise Psychology, 18: 296–312.
Edwards, T., Kingston, K., Hardy, L. and Gould, D. (2002) A qualitative analysis of
catastrophic performances and the associated thoughts, feelings, and emotions. The
Sport Psychologist, 16: 1–19.
Ekman, P. (1992) Facial expression of emotion: new findings, new questions. Psycho-
logical Science, 3: 34–8.
Elias, N. (1978) The Civilising Process. Oxford: Blackwell.
Bibliography 241
Escarti, A. and Gutierrez, M. (2001) Influence of motivational climate in physical

education on the intention to practice physical activity or sport. European Journal of
Sport Science, 1: 1–12.
Everett, J.J., Smith, R.E. and Williams, K.D. (1992) Effects of team cohesion and identifi-
ability on social loafing in relay swimming performance. International Journal of Sport
Eys, M.A., Carron, A.V., Bray, S.R. and Beauchamp, M.R. (2003) Role ambiguity and
athlete satisfaction. Journal of Sports Sciences, 21: 391–401.
Eys, M.A., Hardy, J., Carron, A.V. and Beauchamp, M.R. (2003) The relationship between
task cohesion and competitive state anxiety. Journal of Sport and Exercise Psychology, 25:
66–76.
Fazio, R.H. (1990) Multiple processes by which attitudes guide behavior: the MODE
model as an integrative framework. In M.P. Zanna (ed.), Advances in Experimental Social
Psychology (Vol. 23, pp. 75–109). San Diego, CA: Academic Press.
Feltz, D.L. and Chase, M.A. (1998) The measurement of self-efficacy and confidence
in sport. In J.L. Duda (ed.), Advances in Sport and Exercise Psychology Measurement
(pp. 65–80). Morgantown, WV: Fitness Information Technology.
Feltz, D.L. and Lirgg, C.D. (1998) Perceived team and player efficacy in hockey. Journal of
Applied Psychology, 83: 557–64.
Fernquist, R.M. (2000) An aggregate analysis of professional sports, suicide, and homi-
cide rates: 30 U.S. metropolitan areas, 1971–1990. Aggression and Violent Behavior, 5:
329–41.
Festinger, L. (1954) A theory of social comparison processes. Human Relations, 7: 117–40.
Festinger, L. and Carlsmith, J.M. (1958) Cognitive consequences of forced compliance.
Journal of Abnormal and Social Psychology, 58: 203–10.
Findlay, L.C. and Ste-Marie, D.M. (2004) A reputation bias in figure skating judging.
Finlay, K.A., Trafimow, D. and Villarreal, A. (2002) Predicting exercise and health
behavioral intentions: attitudes, subjective norms, and other behavioral determin-
ants. Journal of Applied Social Psychology, 32: 342–58.
Flegal, K.M. (1999) The obesity epidemic in children and adults: current evidence and
research issues. Medicine and Science in Sports and Exercise, 31 (Suppl.): S509–S514.
Fonseca, A. and Balague, G. (1996) Measuring goal orientations in youth competitive
soccer: a comparison of TEOSQ and POSQ measures. Journal of Applied Sport Psychology,
8 (Suppl.): S143.
Forgas, J.P., Brennan, G., Howe, S., Kane, J.F. and Sweet, S. (1980) Audience effects on
squash players’ performance. Journal of Social Psychology, 111: 41–7.
Fox, K.R. (1999) The influence of physical activity on mental well-being. Public Health
Nutrition, 2 (Suppl.): 411–18.
Fox, K.R. (2000) The effects of exercise on self-perceptions and self-esteem. In S.J.H.
Biddle, K.R. Fox and S.H. Boutcher (eds), Physical Activity and Psychological Well-being
(Vol. 13, pp. 81–118). London: Routledge.
Fox, K.R. and Biddle, S. (1989) The child’s perspective in P.E. Part IV: psychology and
professional issues. British Journal of Physical Education, 20: 35–38.
Fox, K.R. and Corbin, C. (1989) Physical self-perception profile: development and
preliminary validation. Journal of Sport and Exercise Psychology, 11: 408–30.
Frijda, N.H. (1994) Varieties of emotions: emotions and episodes, moods and
sentiments. In P. Ekman and R.J. Davidson (eds), The Nature of Emotion: Fundamental
Questions (pp. 59–67). New York: Oxford University Press.
242 Bibliography
Gabennesch, H. (1988) When promises fail: a theory of temporal fluctuations in suicide.

Social Forces, 67: 129–45.
Gagné, M., Ryan, R.M. and Bargmann, K. (2003) Autonomy support and need satisfac-
tion in the motivation and well-being of gymnasts. Journal of Applied Sport Psychology,
15: 372–90.
Gardner, D.E., Shields, D.L.L., Bredemeier, B.J.L. and Bostrom, A. (1996) The relationship
between perceived coaching behaviors and team cohesion among baseball and
softball players. The Sport Psychologist, 10: 367–81.
Gaudreau, P., Blondin, J.P. and Lapierre, A.M. (2002) Athletes’ coping during a competi-
tion: relationship of coping strategies with positive affect, negative affect, and
performance-goal discrepancy. Psychology of Sport and Exercise, 3: 125–50.
Gayton, W.F. and Langevin, G. (1992) Home advantage: does it exist in individual
sports? Perceptual and Motor Skills, 74: 706.
George, T.R. (1994) Self-confidence and baseball performance: A causal examination of
self-efficacy theory. Journal of Sport and Exercise Psychology, 16: 381–99.
Gernigon, C. and Delloye, J.B. (2003) Self-efficacy, causal attribution, and track athletic
performance following unexpected success or failure among elite sprinters. The Sport
Psychologist, 17: 55–76.
Gill, D.L. (1984) Individual and group performance in sport. In J.M. Silva and R.S. Wein-
berg (eds), Psychological Foundations of Sport (pp. 315–28). Champaign, IL: Human
Kinetics.
Gill, D.L. and Deeter, T.E. (1988) Development of the sport orientation questionnaire.
Research Quarterly for Exercise and Sport, 59: 191–202.
Godin, G., Valois, R., Jobin, J. and Ross, A. (1991) Prediction of intention to exercise of
individuals who have suffered from coronary heart disease. Journal of Clinical Psych-
ology, 47: 762–72.
Golden, N.H. (2002) A review of the female athlete triad (amenorrhea, osteoporosis and
disordered eating). International Journal of Adolescent Medicine and Health, 14: 9–17.
Gollwitzer, P.M. (1999) Implementation intentions: strong effects of simple plans. Amer-
ican Psychologist, 54: 493–503.
Goudas, M., Biddle, S.J.H. and Fox, K.R. (1994) Perceived locus of causality, goal orienta-
tions, and perceived competence in school physical education classes. British Journal of
Gould, D., Greenleaf, C. and Krane, V. (2002) Arousal-anxiety and sport behavior. In T.S.
Horn (ed.), Advances in Sport Psychology (pp. 207–41). Champaign, Il: Human Kinetics.
Gould, D., Petlichkoff, L., Hodge, K. and Simons, J. (1990) Evaluating the effectiveness of
a psychological skills educational workshop. The Sport Psychologist, 4: 249–60.
Gould, D., Petlichkoff, L. and Weinberg, R.S. (1984) Antecedents of, temporal changes in,
and relationships between CSAI-2 components. Journal of Sport Psychology, 6: 289–304.
Gould, D. and Tuffey, S. (1996) Zones of optimal functioning research: a review and
critique. Anxiety, Stress and Coping, 9: 53–68.
Gould, D., Tuffey, S., Hardy, L. and Lockbaum, M. (1993) Multidimensional state anxiety
and middle distance running performance: an exploratory analysis of Hanin’s (1980)
zones of optimal functioning hypothesis. Journal of Applied Sport Psychology, 5: 85–95.
Gould, D. and Weiss, M.R. (1981) The effects of model similarity and model talk on
self-efficacy and muscular endurance. Journal of Sport Psychology, 3: 17–29.
Greenlees, I., Graydon, J. and Maynard, I.W. (2000) The impact of individual efficacy
beliefs on group goal selection and group goal commitment. Journal of Sport Sciences,
18: 451–59.
Bibliography 243
Greer, D.L. (1983) Spectator booing and the home advantage: a study of social influence
in the basketball arena. Social Psychology Quarterly, 46: 252–61.
Grieve, F.G., Whelan, J.P. and Meyers, A.W. (2000) An experimental examination of the
cohesion-performance relationship in an interactive team sport. Journal of Applied
Sport Psychology, 12: 219–35.
Griffin, J. and Harris, J. (1996) Coaches’ attitudes, knowledge, experiences, and recom-
mendations regarding weight control. The Sport Psychologist, 10: 180–94.
Griffiths, M. (1999) Exercise addiction: a case study. Addiction Research, 5 161–8.
Grove, J.R. and Pargman, D. (1986) Relationships among success/failure, attributions,
and performance expectancies in competitive situations. In L.V. Velden and J.H.
Humphrey (eds), Psychology and Sociology of Sport: Current Selected Research (Vol. 1,
pp. 85–95). New York: AMS Press.
Guay, F., Boggiano, A.K. and Vallerand, R.J. (2001) Autonomy support, motivation,
and perceived competence: conceptual and empirical linkages. Personality and Social
Psychology Bulletin, 27: 643–50.
Guay, F., Mageau, G.A. and Vallerand, R.J. (2003) On the hierarchical structure of self-
determined motivation: a test of top-down, bottom-up, reciprocal, and horizontal
effects. Personality and Social Psychology Bulletin, 29: 992–1004.
Hagbloom, S.J., Warnick, R., Warnick, J.E., Jones, V.K., Yarbrough, G.L., Russell, T.M.,
Borecky, C.M., McGahhey, R., Powell, J.L., Beavers, J. and Monte, E. (2002) The 100
most eminent psychologists of the 20th century. Review of General Psychology, 6:
139–52.
Hagger, M.S., Ashford, B. and Stambulova, N. (1998) Russian and British children’s
physical self-perceptions and physical activity participation. Pediatric Exercise Science,
10: 137–52.
Hagger, M.S., Biddle, S.J.H., Chow, E.W., Stambulova, N. and Kavussanu, M. (2003a)
Physical self-perceptions in adolescence: generalizability of a hierarchical multi-
dimensional model across three cultures. Journal of Cross-Cultural Psychology, 34:
611–28.
Hagger, M.S., Biddle, S.J.H. and Wang, C.K.J. (in press a) Physical self-perceptions in
adolescence: generalizability of a multidimensional, hierarchical model across gender
and grade. Educational and Psychological Measurement.
Hagger, M.S. and Chatzisarantis, N.L.D. (in press) First- and higher-order models of
attitudes, normative influence, and perceived behavioural control in the Theory of
Planned Behaviour. British Journal of Social Psychology.
Hagger, M.S., Chatzisarantis, N.L.D., Barkoukis, V., Wang, C.K.J. and Baranowski, J. (in
press b) Perceived autonomy support in physical education and leisure-time physical
activity: a cross-cultural evaluation of the trans-contextual model. Journal of
Educational Psychology.
Hagger, M.S., Chatzisarantis, N. and Biddle, S.J.H. (2002a) The influence of autonomous
and controlling motives on physical activity intentions within the Theory of Planned
Behaviour. British Journal of Health Psychology, 7: 283–97.
Hagger, M.S., Chatzisarantis, N. and Biddle, S.J.H. (2002b) A meta-analytic review of the
theories of reasoned action and planned behavior in physical activity: predictive valid-
ity and the contribution of additional variables. Journal of Sport and Exercise Psychology,
24 3–32.
Hagger, M.S., Chatzisarantis, N., Biddle, S.J.H. and Orbell, S. (2001) Antecedents of
children’s physical activity intentions and behaviour: predictive validity and
longitudinal effects. Psychology and Health, 16 391–407.
244 Bibliography
Hagger, M.S., Chatzisarantis, N., Culverhouse, T. and Biddle, S.J.H. (2003b) The processes
by which perceived autonomy support in physical education promotes leisure-time
physical activity intentions and behavior: a trans-contextual model. Journal of
Hagger, M.S., Lindwall, M. and Asçi, F.H. (2004) A cross-cultural evaluation of a multi-
dimensional and hierarchical model of physical self-perceptions in three national
samples. Journal of Applied Social Psychology, 34: 1075–1107.
Hagger, M.S. and Weed, M.E. (2000) Developing physical activity in children: The
relationship between P.E. and sport. Paper presented at the Annual Congress of the
European College of Sports Science, University of Jyaväskylä, Jyaväskylä, Finland,
19–24 July.
Hall, E.G. and Bunker, L.K. (1979) Locus of control as a mediator of social facilitation
effects during motor skill learning. Journal of Sport Psychology, 1: 332–5.
Hall, H.K. and Kerr, A.W. (1998) Predicting achievement anxiety: a social-cognitive
perspective. Journal of Sport and Exercise Psychology, 20: 98–111.
Halliburton, A.L. and Weiss, M.R. (2002) Sources of competence information and
perceived motivational climate among adolescent female gymnasts varying in skill
level. Journal of Sport and Exercise Psychology, 24: 396–418.
Hanin, Y.L. (1995) Individual zones of optimal functioning (IZOF) model: an idiographic
approach to performance anxiety. In K. Henschen and W. Straub (eds), Sport Psych-
ology: An Analysis of Athlete Behavior (pp. 103–99). Longmeadow, NY: Mouvement.
Hanin, Y.L. (2000) Successful and poor performance and emotions. In Y.L. Hanin (ed.),
Emotions in Sport (pp. 157–88). Champaign, IL: Human Kinetics.
Hanton, S. and Jones, G. (1999a) The acquisition and development of cognitive skills
and strategies: I. Making the butterflies fly in formation. The Sport Psychologist, 13:
1–21.
Hanton, S. and Jones, G. (1999b) The effects of a multimodal intervention program on
performers: II. Training the butterflies to fly in formation. The Sport Psychologist, 13:
22–41.
Hardeman, W., Johnston, M., Johnston, D.W., Bonetti, D., Wareham, N.J. and Kinmonth,
A.L. (2002) Application of the theory of planned behaviour change interventions: a
systematic review. Psychology and Health, 17: 123–58.
Hardy, C.J. (1990) Social loafing: motivational losses in collective performance.
International Journal of Sport Psychology, 21: 305–27.
Hardy, C.J. and Crace, R.K. (1991) The effects of task structure and teammate
competence on social loafing. Journal of Sport and Exercise Psychology, 13: 372–81.
Hardy, C.J. and Latané, B. (1988) Social loafing in cheerleaders: effects of team member-
ship and competition. Journal of Sport and Exercise Psychology, 1988, 10: 109–14.
Hardy, L. (1990) A catastrophe model of performance in sport. In J.G. Jones and L. Hardy
(eds), Stress and Performance in Sport (pp. 81–106). New York: Wiley.
Hardy, L. (1996) Testing the predictions of the Cusp Catastrophe Model of anxiety and
performance. The Sport Psychologist, 10: 140–56.
Hardy, L. and Parfitt, C.G. (1991) A catastrophe model of anxiety and performance.
British Journal of Psychology, 82: 163–78.
Harland, J., White, M., Drinkwater, C., Chinn, D., Farr, L. and Howel, D. (1999) The
Newcastle exercise project: a randomised controlled trial of methods to promote
physical activity in primary care. British Medical Journal, 319: 828–32.
Harris, N.B. (1992) Sex, race, and experiences of aggression. Aggressive Behavior, 18:
201–17.
Bibliography 245
Harrison, J., Mullen, P. and Green, L. (1992) A meta analysis of studies of the health
belief model with adults. Health Education Research, 7,:107–16.
Harter, S. (1988) Manual for the Self-Perception Profile for Adolescents. Denver, CO:
University of Denver.
Harter, S. (1996) Historical roots and contemporary issues involving self-concept. In B.A.
Bracken (ed.), Handbook of Self-concept: Developmental, Social, and Clinical Consider-
ations (pp. 1–37). New York: Wiley.
Harwood, C.G. (2002) Assessing achievement goals in sport: caveats for consultants and
a case for contextualization. Journal of Applied Sport Psychology, 14: 106–19.
Hausenblas, H.A., Carron, A.V. and Mack, D.E. (1997) Application of the Theories of
Reasoned Action and Planned Behavior to exercise behavior: a meta analysis. Journal of
Sport and Exercise Psychology, 19: 36–41.
Hausenblas, H.A. and Giacobbi, P.R., Jr. (2004) Relationship between exercise depend-
ence symptoms and personality. Personality and Individual Differences, 36: 1265–73.
Hausenblas, H.A. and Symons-Downs, D. (2002a) Exercise dependence: a systematic
review. Psychology of Sport and Exercise, 3: 89–123.
Hausenblas, H.A. and Symons-Downs, D. (2002b) How much is too much? The
development and validation of the exercise dependence scale. Psychology and Health,
17: 387–404.
Heider, F. (1958) The Psychology of Interpersonal Relations. New York: John Wiley.
Heuze, J.P. and Brunel, P.C. (2003) Social loafing in a competitive context. International
Hinrichsen, H., Wright, F., Waller, G. and Meyer, C. (2003) Social anxiety and coping
strategies in the eating disorders. Eating Behaviors, 4: 117–26.
Hodge, K. and Petlichkoff, L. (2000) Goal profiles in sport motivation: a cluster analysis.
Hodgkins, S. and Orbell, S. (1998) Can protection motivation theory predict behaviour?
A longitudinal test exploring the role of previous behaviour. Psychology and Health, 13:
237–50.
Holt, N.L. and Sparkes, A.C. (2001) An ethnographic study of cohesiveness in a college
soccer team over a season. The Sport Psychologist, 15: 237–59.
Huddleston, S., Doody, S.G. and Ruder, M.K. (1985) The effect of prior knowledge of the
social loafing phenomenon on performance in a group. International Journal of Sport
Hulley, A.J. and Hill, A.J. (2001) Eating disorders and health in elite women distance
runners. International Journal of Eating Disorders, 30: 312–17.
Institute of European Food Studies (1999) Pan EU Survey on Consumer Attitudes to Physical
Activity, Body Weight and Health. Brussels: European Commission Directorate V/F.3.
Institute of Psychiatry (2004) Illness Prevalence in Society: Is There an Increase of Eating
Disorder Cases in General Population? London: Institute of Psychiatry.
Iso-Ahola, S. (1977) Immediate attributional effects of success and failure in the field:
Testing some laboratory hypotheses. European Journal of Social Psychology, 7: 275–96.
Iyengar, S.S. and DeVoe, S.E. (2003) Rethinking the value of choice: considering cultural
mediators of intrinsic motivation. In V. Murphy-Berman and J.J. Berman (eds),
Cross-cultural Differences in Perspectives on the Self (Vol. 49, pp. 129–74). Lincoln, NE:
University of Nebraska Press.
Jackson, J.M. and Williams, K.D. (1985) Social loafing on difficult tasks: working collect-
ively can improve performance. Journal of Personality and Social Psychology, 49: 937–42.
Joint Surveys Unit (1999) Health Survey for England, 1998. London: The Stationery Office.
246 Bibliography
Jokela, M. and Hanin, Y.L. (1999) Does the individual zones of optimal functioning
model discriminate between successful and less successful athletes? A meta-analysis.
Journal of Sports Sciences, 17: 873–87.
Jones, G. (1995) More than just a game: research developments and issues in competiti-
tive anxiety in sport. British Journal of Psychology, 86: 449–78.
Jones, G., Hanton, S. and Swain, A. (1994) Intensity and interpretation of anxiety
symptoms in elite and nonelite performers. Personality and Individual Differences, 17:
657–63.
Jones, G. and Swain, A. (1995) Predispositions to experience debilitative and facilitative
anxiety in elite and nonelite performers. The Sport Psychologist, 9: 201–11.
Jones, G., Swain, A. and Hardy, L. (1993) Intensity and direction dimensions of com-
petitive state anxiety and relationships with performance. Journal of Sport Sciences, 11:
525–32.
Jones, G., Swain, A. and Harwood, C. (1996) Positive and negative affect as predictors of
competitive anxiety. Personality and Individual Differences, 20: 109–14.
Jones, J.G. (1990) A cognitive perspective on the processes underlying the relationship
between stress and performance in sport. In J.G. Jones and L. Hardy (eds), Stress and
Performance in Sport (pp. 17–42). New York: Wiley.
Jones, J.G. and Hanton, S. (1996) Interpretation of competitive anxiety symptoms and
goal attainment expectancies. Journal of Sport and Exercise Psychology, 18: 144–57.
Jones, J.G. and Hardy, L. (1990) The academic study of stress in sport. In J.G. Jones and
L. Hardy (eds), Stress and Performance in Sport (pp. 3–16). New York: Wiley.
Jones, M.V., Bray, S.R. and Bolton, L. (2001) Game location and officiating bias in
English club cricket. Perceptual and Motor Skills, 93: 359–62.
Jones, M.V., Mace, R.D., Bray, S.R., MacRae, A.W. and Stockbridge, C. (2002) The impact
of motivational imagery on the emotional state and self-efficacy levels of novice
climbers. Journal of Sport Behavior, 25: 57–73.
Kalodner, C.R. and DeLucia-Waack, J.L. (2003) Theory and research on eating disorders
and disturbances in women: suggestions for practice. In M. Kopala and M.A. Keitel
(eds), Handbook of Counseling Women (pp. 506–34). Thousand Oaks, CA: Sage.
Kane, T.D., Marks, M.A., Zaccaro, S.J. and Blair, V. (1996) Self-efficacy, personal goals,
and wrestlers’ self-regulation. Journal of Sport and Exercise Psychology, 18: 36–48.
Kavussanu, M., Crews, D.J. and Gill, D.L. (1998) The effects of single versus multiple
measures of biofeedback on basketball free throw shooting performance. International
Kavussanu, M. and Roberts, G.C. (1996) Motivation in physical activity contexts: The
relationship of perceived motivational climate to intrinsic motivation and self-
efficacy. Journal of Sport and Exercise Psychology, 18: 264–80.
Kavussanu, M. and Roberts, G.C. (2001) Moral functioning in sport: An achievement
goal perspective. Journal of Sport and Exercise Psychology, 23: 37–54.
Kerr, J.H. (1997) Motivation and Emotion in Sport: Reversal Theory. Hove: Psychology Press.
Kerr, J.H., Fujiyama, H. and Campano, J. (2002) Emotion and stress in serious and
hedonistic leisure sport activities. Journal of Leisure Research, 34: 272–89.
Kerr, J.H. and Tacon, P. (2000) Environmental events and induction of metamotivational
reversals. Perceptual and Motor Skills, 91: 337–8.
Keski-Rahkonen, A. (2001) Exercise dependence: a myth or a real issue? European Eating
Disorders Review, 9: 279–83.
Kliene, D. (1990) Anxiety and sport performance: a meta-analysis. Anxiety Research, 2:
113–31.
Bibliography 247
Koestner, R., Lekes, N., Powers, T.A. and Chicoine, E. (2002) Attaining personal goals:
self-concordance plus implementation intentions equals success. Journal of Personality
Koestner, R., Ryan, R.M., Bernieri, F. and Holt, K. (1984) Setting limits on children’s
behavior: the differential effects of controlling versus informational styles on intrinsic
motivation and creativity. Journal of Personality, 52: 233–48.
Koka, A. and Hein, V. (in press) The effect of perceived teacher’s feedback on intrinsic
motivation in physical education. Psychology of Sport and Exercise.
Kowalski, N.P., Crocker, P.R.E. and Kowalski, K.C. (2001) Physical self and physical
anxiety relationships in college women: does social physique anxiety moderate
effects? Research Quarterly for Exercise and Sport, 72: 55–62.
Kowalski, K.C., Crocker, P.R.E., Kowalski, N.P., Chad, K.E. and Humbert, M.L. (2003)
Examining the physical self in adolescent girls over time: further evidence against the
hierarchical model. Journal of Sport and Exercise Psychology, 25: 5–18.
Kozar, B. (1973) The effects of a supportive and nonsupportive audience upon learning a
gross motor skill. International Journal of Sport Psychology, 3: 27–38.
Landers, D.M. (1980) The arousal-performance relationship revisited. Research Quarterly
for Exercise and Sport, 51: 77–90.
Lane, A.M. and Lane, H.J. (2002) Predictive effectiveness of mood measures. Perceptual
and Motor Skills, 94: 785–91.
Latané, B., Williams, K. and Harkins, S. (1979) Many hands make light the work: the
causes and consequences of social loafing. Journal of Personality and Social Psychology,
37: 822–32.
Legrand, F.D. and LeScanff, C. (2003) Tension-stress, effort-stress and mood profiling
with an elite javelin performer. Psychology of Sport and Exercise, 4: 429–36.
Lehman, D.R. and Reifman, A. (1987) Spectator influence on basketball officiating.
Journal of Social Psychology, 127: 673–5.
Lehoux, P.M., Steiger, H. and Jabalpurlawa, S. (2000) State/trait distinctions in bulimic
syndromes. International Journal of Eating Disorders, 27: 36–42.
Levine, P. and Vinten-Johansen, P. (1981) The historical perspective: violence and sport.
Arena Review, Journal of Sport and Social Issues, 5: 583–95.
Lichacz, F.M. and Partington, J.T. (1996) Collective efficacy and true group performance,
27: 146–58.
Lindner, K.J. and Kerr, J.H. (2001) Predictability of sport participation motivation from
metamotivational dominances and orientations. Personality and Individual Differences,
30: 759–73.
Liu, J.L.Y., Maniadakis, N., Gray, A. and Rayner, M. (2002) The economic burden of
coronary heart disease in the UK. Heart, 88: 597–603.
Lohasz, P.G. and Leith, L.M. (1997) The effect of three mental preparation strategies on
the performance of a complex response time task. International Journal of Sport
Loomes, G. and Sugden, R. (1982) Regret theory: an alternative theory of rational choice
under uncertainty. Economic Journal, 92: 805–24.
Loughead, T.M., Carron, A.V., Bray, S.R. and Kim, A.J. (2003) Facility familiarity and
the home advantage in professional sports. International Journal of Sport and Exercise
Lowe, R., Eves, F. and Carroll, D. (2002) The influence of affective and instrumental
beliefs on exercise intentions and behavior: a longitudinal analysis. Journal of Applied
248 Bibliography
Madison, J.K. and Ruma, S.L. (2003) Exercise and athletic involvement as moderators of
severity in adolescents with eating disorders. Journal of Applied Sport Psychology, 15:
213–22.
Magyar, T.M., Feltz, D.L. and Simpson, I.P. (2004) Individual and crew level determinants
of collective efficacy in rowing. Journal of Sport and Exercise Psychology, 26: 136–53.
Mann, L. (1981) The baiting crowd in episodes of threatened suicide. Journal of Personal-
ity and Social Psychology, 41: 703–9.
Markland, D. (2004) Motivational interviewing and self-determination theory.
Retrieved August 1, 2004, from University of Wales, Bangor, School of Sport, Health,
and Exercise Sciences Exercise Psychology Web site: http://www.bangor.ac.uk/
%7Epes004/exercise psych/misdt.htm: University of Wales, Bangor.
Markland, D. and Tobin, V. (2004) A modification to the Behavioural Regulation in
Exercise Questionnaire to include an assessment of amotivation. Journal of Sport and
Exercise Psychology, 26: 191–6.
Marsh, H.W. (1989) Age and sex effects in multiple dimensions of self-concept:
Preadolescence to early adulthood. Journal of Educational Psychology, 81: 417–30.
Marsh, H.W. (1990) A multidimensional, hierarchical model of self-concept: theoretical
and empirical justification. Educational Psychology Review, 2: 77–172.
Marsh, H.W., Marco, I.T. and Asçi, F.H. (2002) Cross-cultural validity of the physical
self-description questionnaire: comparison of factor structures in Australia, Spain and
Turkey. Research Quarterly for Exercise and Sport, 73: 257–70.
Marsh, H.W. and O’Niell, R. (1984) Self-description questionnaire III (SDQ III): the
construct validity of multidimensional self-concept ratings by late-adolescents.
Journal of Educational Measurement, 21: 153–74.
Marsh, H.W., Perry, C., Horsely, C. and Roche, L. (1995) Multidimensional self-concept
of elite athletes: how do they differ from the general population? Journal of Sport and
Exercise Psychology, 17: 70–83.
Marsh, H.W. and Redmayne, R.S. (1994) A multidimensional physical self-concept and
its relations to multiple components of physical fitness. Journal of Sport and Exercise
Marsh, H.W., Richards, G.E., Johnson, S., Roche, S. and Tremayne, P. (1994) Physical self
description questionnaire: psychometric properties and a multitrait-multimethod
analysis of relations to existing instruments. Journal of Sport and Exercise Psychology, 16:
270–305.
Marsh, H.W. and Shavelson, R. (1985) Self-concept: its multifaceted hierarchical
structure. Educational Psychologist, 20: 107–23.
Marsh, H.W. and Yeung, A.S. (1998) Top-down, bottom-up, and horizontal models: the
direction of causality in multidimensional, hierarchical self-concept models. Journal of
Personality and Social Psychology, 75: 509–27.
Marsh, P. and Harré, R. (1978) The world of football hooliganism. Human Nature, 1: 62–9.
Marshall, S.J. and Biddle, S.J.H. (2001) The transtheoretical model of behavior change: a
meta-analysis of applications to physical activity and exercise. Annals of Behavioral
Medicine, 23: 229–46.
Martens, R., Vealey, R.S. and Burton, D. (1990) Competitive Anxiety in Sport. Champaign,
IL: Human Kinetics.
Martin, J.J. and Gill, D.L. (1991) The relationship among competitive orientation,
sport-confidence, self-efficacy, anxiety, and performance. Journal of Sport and Exercise
Martin, J.J. and Gill, D.L. (1995) The relationships of competitive orientations and
Bibliography 249
self-efficacy to goal importance, thoughts, and performance in high school distance

runners. Journal of Applied Sport Psychology, 7: 50–62.
Martin, K.A. and Hausenblas, H.A. (1998) Psychological commitment to exercise
and eating disorder symptomology among female aerobic instructors. The Sport
Psychologist, 12: 180–90.
Maslow, A. (1943) A theory of human motivation. Psychological Reports, 50: 370–96.
Matheson, H., Mathes, S. and Murray, M. (1996) Group cohesion of female inter-
collegiate coacting and interacting team across a competitive season. International
Matheson, H., Mathes, S. and Murray, M. (1997) The effect of winning and losing on
female interactive and coactive team cohesion. Journal of Sport Behavior, 20: 284–98.
Matthews, K.A., Krantz, D.S., Dembroski, T.M. and MacDougall, J.M. (1982) Unique and
common variance in Structured Interview and Jenkins Activity Survey measures of
Type A behavior pattern. Journal of Personality and Social Psychology, 42: 303–13.
McAuley, E. (1985) Modeling and self-efficacy: a test of Bandura’s model. Journal of Sport
McAuley, E. and Blissmer, B. (2002) Self-efficacy and attributional processes in physical
activity. In T.S. Horn (ed.), Advances in Sport Psychology (pp. 185–206). Champaign, IL:
Human Kinetics.
McAuley, E. and Tammen, V.V. (1989) The effects of subjective and objective competitive
outcomes on intrinsic motivation. Journal of Sport and Exercise Psychology, 11: 84–93.
McAuliffe, B.J., Jetten, J., Hornsey, M.J. and Hogg, M.A. (2003) Individualist and
collectivist norms: when it’s ok to go your own way. European Journal of Social
McCrae, R.R. and Costa, P.T. (1996) Toward a new generation of personality theories:
theoretical contexts for the five-factor model. In J.S. Wiggins (ed.), The Five-factor
Model of Personality: Theoretical Perspectives (pp. 51–87). New York: Guilford Press.
McCutcheon, L.E. (1984) The home advantage in high school athletics. Journal of Sport
McGuire, E.J., Courneya, K.S., Widmeyer, W.N. and Carron, A.V. (1992) Aggression as a
potential mediator of the home advantage in professional ice hockey. Journal of Sport
McNair, D., Lorr, M. and Droppleman, L. (1971) Profile of Mood States. San Diego:
Education and Industrial Testing Service.
McNamara, J.J., Molot, M.A., Stremple, J.F. and Cutting, R.T. (1971) Coronary artery
disease in combat casualties in Vietnam. Journal of the American Medical Association,
216(7): 1185–7.
Meiland, J. W. (1970) The Nature of Intention. London: Methuen.
Mellalieu, S.D., Hanton, S. and Jones, G. (2003) Emotional labeling and competitive
anxiety in preparation and competition. The Sport Psychologist, 17: 157–74.
Meyers, A.W., Schleser, R. and Okwumabua, T.M. (1982) A cognitive behavioral inter-
vention for improving basketball performance. Research Quarterly for Exercise and Sport,
53: 344–7.
Miller, C.T. and Downey, K.T. (1999) A meta-analysis of heavy weight and self-esteem.
Personality and Social Psychology Review, 3: 68–84.
Miller, D.L. (2001) Reexamining teamwork KSAs and team performance. Small Group
Research, 32: 745–66.
Miller, J.T. and McAuley, E. (1987) Effects of a goal-setting training program on basket-
ball free-throw self-efficacy and performance. The Sport Psychologist, 1: 103–13.
250 Bibliography
Miller, W.R. (1999) Toward a theory of motivational interviewing. Motivational Interview-

ing Newsletter: Updates, Education and Training, 6: 2–4.
Milne, S., Sheeran, P. and Orbell, S. (2000) Prediction and intervention in health-related
behavior: a meta-analytic review of protection motivation theory. Journal of Applied
Milne, S.E., Orbell, S. and Sheeran, P. (2002) Combining motivational and volitional
interventions to promote exercise participation: protection motivation theory and
implementation intentions. British Journal of Health Psychology, 7: 163–84.
Morgan, W.P. (1980) Test of a champion: the iceberg profile. Psychology Today, 14: 92–108.
Morgan, W.P., O’Connor, P.J.O., Ellickson, K.A. and Bradley, P.W. (1988) Personality
structure, mood states, and performance in elite male distance runners. International
Mullen, B. and Copper, C. (1994) The relation between group cohesiveness and
performance: an integration. Psychological Bulletin, 115: 210–27.
Mullen, B. and Riordan, C. (1988) Self-serving attributions for performance in natural-
istic settings: a meta-analytic review. Journal of Applied Social Psychology, 18: 3–22.
Murphy, P. (1990) Why there are no equivalents of soccer hooliganism in the United
States. In P. Murphy, J. Williams and E. Dunning (eds), Football on Trial: Spectator
Violence and Development in the Football World (pp. 194–212). London: Routledge.
National Centre for Social Research (1999) Health Survey for England. London:
Department of Health.
National Health Service Health Development Agency (1996) Active For Life. London:
Health Development Agency.
National Institute of Mental Health (1993) Eating Disorders (NIH Publication No. 94–3477)
Bethesda, MD: Office of Communications and Public Liaison, National Institute of
Mental Health (NIMH).
Neave, N. and Wolfson, S. (2003) Testosterone, territoriality, and the ‘home advantage’.
Physiology and Behavior, 78, 269–75.
Nederhof, A. (1989) Self-involvement, intention certainty and attitude-intention
consistency. British Journal of Social Psychology, 28: 123–33.
Nevill, A.M., Balmer, N.J. and Williams, A.M. (2002) The influence of crowd noise and
experience upon refereeing decisions in football. Psychology of Sport and Exercise, 3:
261–72.
Newton, M.L. and Duda, J.L. (1995) Relations of goal orientations and expectations on
multidimensional state anxiety. Perceptual and Motor Skills, 81: 1107–12.
Newton, M.L. and Duda, J.L. (1999) The interaction of motivational climate,
dispositional goal orientations, and perceived ability in predicting indices of motiv-
ation. International Journal of Sport Psychology, 30: 63–82.
Nicholls, J.G. (1989) The Competitive Ethos and Democratic Education. Cambridge, MA:
Harvard University Press.
Notani, A.S. (1998) Moderators of perceived behavioral control’s predictiveness in the
theory of planned behavior: a meta-analysis. Journal of Consumer Psychology, 7: 247–71.
Ntoumanis, N. (2001) A self-determination approach to the understanding of
motivation in physical education. British Journal of Educational Psychology, 71: 225–42.
Ntoumanis, N. and Biddle, S. (1998) The relationship between competitive anxiety,
achievement goals and motivational climates. Research Quarterly for Exercise and Sport,
69: 176–87.
Ntoumanis, N. and Biddle, S.J.H. (1999) A review of motivational climate in physical
activity. Journal of Sports Sciences, 17: 543–665.
Bibliography 251
Ntoumanis, N., Biddle, S.J.H. and Haddock, G. (1999) The mediating role of coping
strategies on the relationship between achievement motivation and affect in sport.
Anxiety, Stress and Coping, 12: 299–327.
Ntoumanis, N., Pensgaard, M., Martin, C. and Pipe, K. (in press) An idiographic analysis
of amotivation in compulsory school physical education. Journal of Sport and Exercise
Psychology.
Ogden, J., Veale, D. and Summers, Z. (1997) The development and validation of the
Exercise Dependence Questionnaire. Addiction Research, 5: 343–55.
Ommundsen, Y. and Pedersen, B.H. (1999) The role of achievement goal orientations
and perceived ability upon somatic and cognitive indices of sport competition trait
anxiety: a study of young athletes. Scandinavian Journal of Medicine and Science in Sports,
9: 333–43.
Orbell, S. (2003) Personality systems interactions theory and the theory of planned
behaviour: evidence that self-regulatory volitional components enhance enactment
of studying behaviour. British Journal of Social Psychology, 42: 95–112.
Orbell, S., Hodgkins, S. and Sheeran, P. (1997) Implementation intentions and the
Theory of Planned Behavior. Personality and Social Psychology Bulletin, 23: 945–54.
Orbell, S. and Sheeran, P. (1998) ‘Inclined abstainers’: a problem for predicting health
related behaviour. British Journal of Social Psychology, 37: 151–65.
Oxendine, J.B. (1970) Emotional arousal and motor performance. Quest, 13: 23–32.
Oyserman, D. (2004) Self-concept and identity. In M.B. Brewer and M. Hewstone (eds),
Self and Social Identity (pp. 5–24). Oxford: Blackwell.
Pace, A. and Carron, A.V. (1992) Travel and the home advantage. Canadian Journal of
Sport Sciences, 17: 60–4.
Paffenbarger, R.S. and Hale, W.E. (1975) Work activity and coronary heart mortality. New
England Journal of Medicine, 292: 545–50.
Parker, D. and Bar-Or, O. (1991) Juvenile obesity: the importance of exercise. The
Physician and Sports Medicine, 19: 113–16.
Parker, D., Manstead, A.S.R. and Stradling, S.G. (1995) Extending the theory of planned
behaviour: the role of personal norm. British Journal of Social Psychology, 34: 127–37.
Pate, R.R., Pratt, M., Blair, S.N., Haskell, W.L., Macera, C.A., Bouchard, C., Buchner, D.,
Ettiger, W., Heath, G.W., King, A.C., Kriska, A., Leon, A.S., Marcus, B., Morris, J.,
Paffenbarger, R.S., Patrick, K., Pollock, M.L., Rippe, J.M., Sallis, J. and Wilmore, J.H.
(1995) Physical activity and public health: a recommendation from the Centers for
Disease Control and Prevention and the American College of Sports Medicine. Journal
of the American Medical Association, 273: 402–7.
Paulus, P.B., Shannon, J.C., Wilson, D.L. and Boone, T.D. (1972) The effect of spectator
presence on gymnastic performance in a field situation. Psychonomic Science, 29: 88–90.
Pelletier, L.G., Fortier, M.S., Vallerand, R.J., Tuson, K.M., Briere, N.M. and Blais, M.R.
(1995) Toward a new measure of intrinsic motivation, extrinsic motivation and
amotivation in sport: the sport motivation scale (SMS). Journal of Sport and Exercise
Pensgaard, A.M. (1999) The dynamics of motivation and perceptions of control when
competing in the Olympic Games. Perceptual and Motor Skills, 89: 116–25.
Pensgaard, A.M. and Roberts, G.C. (1995) Perceived motivational climate and sources of
stress for winter Olympic athletes. Journal of Applied Sport Psychology, 7(Suppl.): S9.
Perugini, M. and Bagozzi, R.P. (2001) The role of desires and anticipated emotions in
goal-directed behaviours: broadening and deepening the theory of planned behavior.
British Journal of Social Psychology, 40: 79–98.
252 Bibliography
Perugini, M. and Conner, M. (2000) Predicting and understanding behavioral volitions:

the interplay between goals and behaviors. European Journal of Social Psychology, 30:
705–31.
Petersen, S., Peto, V. and Rayner, M. (2004) Coronary Heart Disease Statistics. London:
British Heart Foundation.
Petherick, C.M. and Weigand, D.A. (2002) The relationship of dispositional goal orienta-
tions and perceived motivational climates on indices of motivation in male and
female swimmers. International Journal of Sport Psychology, 33: 218–37.
Petty, R.E. and Cacioppo, J. (1986) Communication and Persuasion: Central and Peripheral
Routes to Attitude Change. New York: Springer-Verlag.
Pierce, E.F., Eastman, N.W., Tripathi, H.L. and Olson, K.G. (1993) B-Endorphin response
to endurance exercise: relationship to exercise dependence. Perceptual and Motor Skills,
77: 767–70.
Pope, H.G. and Katz, D.L. (1994) Psychiatric and medical effects of anabolic-androgenic
steroid use: a controlled study of 160 athletes. Archives of General Psychiatry, 51: 375–82.
Povey, R., Conner, M., Sparks, P., James, R. and Shepherd, R. (2000) Application of the
Theory of Planned Behaviour to two dietary behaviours: roles of perceived control and
self-efficacy. British Journal of Health Psychology, 5: 121–39.
Prapavessis, H. (2000) The POMS and sports performance: a review. Journal of Applied
Prapavessis, H. and Carron, A.V. (1997) Sacrifice, cohesion, and conformity to norms in
sport teams. Group Dynamics, 1: 231–40.
Prestwich, A., Lawton, R. and Conner, M. (2003) The use of implementation intentions
and the decision balance sheet in promoting exercise behaviour. Psychology and Health,
10: 707–21.
Prichard, R.D., Campbell, K.M. and Campbell, D.J. (1977) Effects of extrinsic financial
rewards on intrinsic motivation. Journal of Applied Psychology, 62: 9–15.
Prochaska, J.O. and DiClemente, C.C. (1982) Trans-theoretical theory: Towards a more
integrated model of change. Journal of Consultative Clinical Psychology, 19: 276–88.
Prochaska, J.O., Velicer, W.F., Rossie, J.S., Goldstein, M.G., Marcus, B.H., Rakowski, W.,
Fiore, C., Harlow, L.L., Redding, C.A., Rosenbloom, D.A. and Rossi, S.R. (1994) Stages of
change and decisional balance for 12 problem behaviors. Health Psychology, 13: 39–46.
Quine, L., Rutter, D.R. and Arnold, L. (1998) Predicting and understanding safety helmet
use among schoolboy cyclists: a comparison of the theory of planned behaviour and
the health belief model. Psychology and Health, 13: 251–69.
Quine, L., Rutter, D. and Arnold, L. (2001) Persuading school-age cyclists to use safety
helmets: effectiveness of an intervention based on the Theory of Planned Behaviour.
British Journal of Health Psychology, 6: 327–45.
Rainey, D.W. and Schweickert, G.J. (1988) An exploratory study of team cohesion before
and after a spring trip. The Sport Psychologist, 2: 314–17.
Raitakari, O.T., Porkka, K.V., Taimela, S., Telama, R., Rasanen, L. and Viikari, J.S. (1994)
Effects of persistent physical activity and inactivity on coronary risk factors in
children and young adults. American Journal of Epidemiology, 140: 195–205.
Rajecki, D.W., Nerenz, D.R., Freedenberg, T.G. and McCarthy, P.J. (1979) Components
of aggression in chickens and conceptualizations of aggression in general. Journal of
Personality and Social Psychology, 37: 1902–14.
Raudsepp, L., Liblik, R. and Hannus, A. (2002) Children’s and adolescents’ physical
self-perceptions as related to moderate to vigorous physical activity and fitness.
Pediatric Exercise Science, 14: 97–106.
Bibliography 253
Reeve, J. (2002) Self-determination theory applied to educational settings. In E.L. Deci

and R.M. Ryan (eds), Handbook of Self-determination Research (pp. 183–203). Rochester,
NY: University of Rochester Press.
Reeve, J., Bolt, E. and Cai, Y. (1999) Autonomy-supportive teachers: how they teach and
motivate students. Journal of Educational Psychology, 91: 537–48.
Reeve, J. and Deci, E.L. (1996) Elements of the competitive situation that affect intrinsic
motivation. Personality and Social Psychology Bulletin, 22: 24–33.
Reeve, J., Olson, B.C. and Cole, S.G. (1985) Motivation and performance: two con-
sequences of winning and losing a competition. Motivation and Emotion, 9: 291–8.
Reicher, S.D. (1987) Crowd behaviour and social action. In J.C. Turner, M.A. Hogg, P.J.
Oakes, S.D. Reicher and M.S. Wetherell (eds), Rediscovering the Social Group: A Self-
categorisation Theory (pp. 171–202). Oxford: Blackwell.
Reicher, S.D., Spears, R. and Postmes, T. (1995) A social identity model of deindividua-
tion phenomena. In M. Hewstone and W. Stroebe (eds), European Review of Social
Psychology (Vol. 6). Chichester: Wiley.
Rhodes, R.E., Courneya, K.S. and Jones, L.W. (2002a) Personality, the theory of planned
behavior and exercise: a unique role for extroversion’s activity facet. Journal of Applied
Rhodes, R.E., Jones, L.W. and Courneya, K.S. (2002b) Extending the theory of planned
behavior in the exercise domain: a comparison of social support and subjective norm.
Research Quarterly for Exercise and Sport, 73: 193–9.
Richards, G.E. (1988) Physical Self-concept Scale. Sydney: Australian Outward Bound
Foundation.
Rivis, A. and Sheeran, P. (2003) Descriptive norms as an additional predictor in the
theory of planned behaviour: a meta-analysis. Current Psychology, 22: 218–33.
Roberts, G.C. and Pascuzzi, D. (1979) Causal attributions in sport: some theoretical
implications. Journal of Sport Psychology, 1: 203–11.
Roberts, G.C., Treasure, D.C. and Balague, G. (1998) Achievement goals in sport: the
development and validation of the perception of success questionnaire. Journal of
Sports Sciences, 16: 337–47.
Robinson, T.T. and Carron, A.V. (1982) Personal and situational factors associated with
dropping out versus maintaining participation in competitive sport. Journal of Sport
Rodgers, W.M. and Brawley, L.R. (1993) Using both self-efficacy theory and the theory of
planned behavior to discriminate adherers and dropouts from structured programs.
Journal of Applied Sport Psychology, 5: 195–206.
Rogers, R.W. (1975) A protection motivation theory of fear appeals and attitude change.
Journal of Psychology, 91: 93–114.
Rogers, R.W. (1983) Cognitive and physiological processes in fear appraisals and attitude
change: a revised theory of protection motivation. In J.T. Cacioppo and R.E. Petty
(eds), Social Psychology: A Source Book (pp. 153–76). New York: Guilford Press.
Rollnick, S. and Miller, W.R. (1995) What is Motivational Interviewing? Behavioural and
Cognitive Psychotherapy, 23: 325–34.
Rosenberg, M. (1979) Conceiving the Self. New York: Basic Books.
Rosenstock, I.M. (1974) Historical origins of the health belief model. Health Education
Monographs, 2: 328–35.
Rowley, A.J., Landers, D.M., Kyllo, L.B. and Etnier, J.L. (1995) Does the Iceberg Profile
discriminate between successful and less successful athletes? A meta-analysis. Journal
of Sport and Exercise Psychology, 17: 185–99.
254 Bibliography
Rudisill, M. (1989) Influence of perceived competence and causal dimension orienta-

tions on expectations, persistence and performance during perceived failure. Research
Quarterly for Exercise and Sport, 60: 166–75.
Russell, G.W. (1979) Hero selection by Canadian ice hockey players: Skill or aggression?
Canadian Journal of Applied Sport Sciences, 4: 309–13.
Russell, G.W. (1983) Crowd size and density in relation to athletic aggression and
performance. Social Behavior and Personality, 11: 9–15.
Russell, G.W. (1993) The Social Psychology of Sport. New York: Springer-Verlag.
Ryan, R. (1992) Management of eating problems in athletic settings. In K.D. Brownell, J.
Rodin and J.H. Wilmore (eds), Eating, Body Weight, and Performance in Athletes:
Disorders of Modern Society (pp. 344–62). Philadelphia, PA: Lea and Febiger.
Ryan, R.M. and Connell, J.P. (1989) Perceived locus of causality and internalization:
examining reasons for acting in two domains. Journal of Personality and Social Psych-
ology, 57: 749–61.
Ryan, R.M. and Deci, E.L. (1989) Bridging the research traditions of task/ego
involvement and intrinsic/extrinsic motivation: comment on Butler (1987). Journal of
Ryan, R.M., Vallerand, R.J. and Deci, E.L. (1984) Intrinsic motivation in sport: a
cognitive evaluation theory interpretation. In W.F. Straub and J.M. Williams (eds),
Cognitive Sport Psychology (pp. 231–42). Lansing, NY: Sport Science Associates.
Sallis, J.F. and Patrick, K. (1994) Physical activity guidelines for adolescents: consensus
statement. Pediatric Exercise Science, 6: 302–14.
Salminen, S. (1993) The effect of the audience on the home advantage. Perceptual and
Motor Skills, 76: 1123–8.
Sanna, L.J. (1992) Self-efficacy theory: Implications for social facilitation and social
loafing. Journal of Personality and Social Psychology, 62: 774–86.
Sarason, S.B., Davidson, K.S., Lighthall, F.F., Waite, R.R. and Ruebush, B.K. (1960) Anxiety
in Elementary School Children. New York: Wiley.
Sarrazin, P., Biddle, S., Famose, J. P., Cury, F., Fox, K. and Durand, M. (1996) Goal orienta-
tions and conceptions of the nature of sport ability in children: a social cognitive
approach. British Journal of Psychological Society, 35: 399–414.
Sarrazin, P., Vallerand, R.J., Guillet, E., Pelletier, L.G. and Cury, F. (2002) Motivation and
dropout in female handballers: a 21-month prospective study. European Journal of
Schacter, S. and Singer, J.E. (1962) Cognitive, social, and physiological determinants of
emotional state. Psychological Review, 69: 379–399.
Scheer, J.K. and Ansorge, C.J. (1979) Influence due to expectations of judges. Journal of
Schlenker, B.R., Phillips, S.T., Boniecki, K.A. and Schlenker, D.R. (1995) Championship
pressures: choking or triumphing in one’s own territory? Journal of Personality and
Schwartz, B. and Barsky, S.F. (1977) The home advantage. Social Forces, 55:
641–61.
Schwenkmezger, P. and Steffgen, G. (1989) Anxiety and motor performance. In B.
Kirkcaldy (ed.), Normalities and Abnormalities in Human Movement (Vol. 29, pp. 78–99).
Basle, Switzerland: S. Karger AG.
Sedikides, C. and Gregg, A.P. (2003) Portraits of the self. In M.A. Hogg and J. Cooper
(eds), Sage Handbook of Social Psychology (pp. 110–38). London: Sage.
Seifriz, J.J., Duda, J.L. and Chi, L. (1992) The relationship of perceived motivational
Bibliography 255
climate to intrinsic motivation and beliefs about success in basketball. Journal of Sport
Shavelson, R.J., Hubner, J.J. and Stanton, G.C. (1976) Self-concept: validation of con-
struct interpretations. Review of Educational Research, 46: 407–41.
Sheeran, P. (2002) Intention-behavior relations: a conceptual and empirical review. In
W. Stroebe and M. Hewstone (eds), European Review of Social Psychology (pp. 1–36).
London: Wiley.
Sheeran, P., Norman, P. and Orbell, S. (1999a) Evidence that intentions based on
attitudes better predict behaviour than intentions based on subjective norms.
European Journal of Social Psychology, 29: 403–6.
Sheeran, P. and Orbell, S. (1999) Implementation intentions and repeated behaviour:
augmenting the predictive validity of the theory of planned behaviour. European
Journal of Social Psychology, 29: 349–69.
Sheeran, P. and Orbell, S. (2000) Self schemas and the theory of planned behaviour.
European Journal of Social Psychology, 30: 533–50.
Sheeran, P., Orbell, S. and Trafimow, D. (1999b) Does the temporal stability of behavioral
intentions moderate intention-behavior and past behavior-future behavior relations?
Personality and Social Psychology Bulletin, 25: 721–30.
Sheeran, P. and Silverman, M. (2003) Evaluation of three interventions to promote
workplace health and safety: evidence for the utility of implementation intentions.
Social Science and Medicine, 56: 2153–63.
Sheldon, K.M., Elliot, A.J., Kim, Y. and Kasser, T. (2001) What is satisfying about
satisfying events? Testing 10 candidate psychological needs. Journal of Personality and
Sheppard, B.H., Hartwick, J. and Warshaw, P.R. (1988) The theory of reasoned action: a
meta-analysis of past research with recommendation and future research. Journal of
Consumer Research, 15: 325–43.
Sherman, S. and Fazio, R. (1983) Parallels between attitudes and traits as predictors of
behavior. Journal of Personality, 51: 308–45.
Short, S.E., Bruggeman, J.M., Engel, S.G., Marback, T.L., Wang, L J., Willadsen, A. and
Short, M.W. (2002) The effect of imagery function and imagery direction on
self-efficacy and performance on a golf-putting task. The Sport Psychologist, 16:
48–67.
Silva, J.M. (1979) Behavioral and situational factors affecting concentration and skill
performance. Journal of Sport Psychology, 1: 221–7.
Silva, J.M. (1980) Assertive and aggressive behavior in sport: a definitional clarification.
In C.H. Nadeau, W R. Halliwell, K.M. Newell and G.C. Roberts (eds), Psychology of
Motor Behavior and Sport, 1979 (pp. 199–208). Champaign, IL: Human Kinetics.
Silva, J.M. and Andrew, A. (1987) An analysis of game location and basketball perform-
ance in the Atlantic Coast Conference. International Journal of Sport Psychology, 18:
188–204.
Singer, R.N. and McCaughan, L. (1978) Motivational effects of attributions expectancy,
and achievement motivation during the learning of a novel motor task. Journal of
Motor Behavior, 10: 245–53.
Slater, M.R. and Sewell, D.F. (1994) An examination of the cohesion-performance
relationship in university hockey teams. Journal of Sport Sciences, 12: 423–31.
Smith, C.A., Kelly, N.H., Lazarus, R.S. and Pope, L.K. (1993) In search of the ‘Hot’
cognitions: attributes, appraisals, and their relation to emotion. Journal of Personality
256 Bibliography
Smith, L.E. and Crabbe, J. (1976) Experimenter role relative to social facilitation and
motor learning. International Journal of Sport Psychology, 7: 158–68.
Smith, R.A. and Biddle, S.J.H. (1999) Attitudes and exercise adherence: test of the Theor-
ies of Reasoned Action and Planned Behaviour. Journal of Sports Sciences, 17: 269–81.
Sonstroem, R.J., Harlow, L.L., Gemma, L.M. and Osborne, S. (1991) Test of structural
relationships within a proposed exercise and self-esteem model. Journal of Personality
Assessment, 56: 348–64.
Sonstroem, R.J., Harlow, L.L. and Josephs, L. (1994) Exercise and self-esteem: validity of
model expansion and exercise associations. Journal of Sport and Exercise Psychology, 16:
29–42.
Sonstroem, R.J., Harlow, L.L. and Salisbury, K.S. (1993) Path analysis of a self-esteem
model across a competitive swim season. Research Quarterly for Exercise and Sport, 64(3):
335–42.
Sonstroem, R.J. and Morgan, W.P. (1989) Exercise and self-esteem: rationale and model.
Medicine and Science in Sports and Exercise, 21: 329–37.
Spano, L. (2001) The relationship between exercise and anxiety, obsessive-
compulsiveness, and narcissism. Personality and Individual Differences, 30: 87–93.
Spielberger, C.D., Gorusch, R.L. and Lushene, R.E. (1970) Manual for the State Trait
Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press.
Spink, K.S. (1990) Collective efficacy in the sport setting. International Journal of Sport
Stajkovic, A.D. and Luthans, F. (1998) Self-efficacy and work-related performance: a
meta-analysis. Psychological Bulletin, 124: 240–61.
Standage, M., Duda, J.L. and Ntoumanis, N. (2003) A model of contextual motivation in
physical education: using constructs from self-determination and achievement goal
theories to predict physical activity intentions. Journal of Educational Psychology,
95: 97–110.
Starek, J. and McCullagh, P. (1999) The effect of self-modeling on the performance of
beginning swimmers. The Sport Psychologist, 13: 269–87.
Steenland, K. and Deddens, J.A. (1997) Effect of travel and rest on performance of
professional basketball players. Sleep: Journal of Sleep Research and Sleep Medicine, 20:
366–9.
Stephens, D.E. and Bredemeier, B.J.L. (1996) Moral atmosphere and judgments about
aggression in girls’ soccer: relationships among moral and motivational variables.
Stott, C. (2001) ‘Hooligans’ abroad? Inter-group dynamics, social identity and participa-
tion in collective ‘disorder’ at the 1998 World Cup finals. British Journal of Social Psych-
ology, 40: 359–84.
Strauss, B. (2002) Social facilitation in motor tasks: a review of research and theory.
Psychology of Sport and Exercise, 3: 237–56.
Sullivan, P. and Feltz, D.L. (2003) The preliminary development of the Scale for Effective
Communication in Team Sports (SECTS). Journal of Applied Social Psychology, 33:
1693–715.
Sundgot-Borgen, J. (2000) Eating disorders in athletes. In R.J. Maughan (ed.), Nutrition in
Sport (pp. 510–22). Oxford: Blackwell.
Sutton, S. (2000) Interpreting cross-sectional data on stages of change. Psychology and
Health, 15: 163–71.
Swain, A. and Jones, G. (1992) Relationships between sport achievement orientation and
competitive state anxiety. The Sport Psychologist, 6: 42–54.
Bibliography 257
Swain, A.B.J. and Harwood, C.G. (1996) Antecedents of state goals in age-group
swimmers: an interactionist perspective. Journal of Sport Sciences, 14: 111–24.
Tajfel, H. and Turner, J.C. (1986) The social identity theory of intergroup behaviour. In
S.G. Worchel and W. Austin (eds), Psychology of Intergroup Relations (2nd edn, pp. 7–24).
Chicago, IL: Nelson-Hall.
Taylor, I. (1971) Football mad: a speculative sociology of soccer hooliganism. In E.
Dunning (ed.), The Sociology of Sport. London: Routledge.
Taylor, S.P. and Gammon, C.B. (1976) Aggressive behavior of intoxicated subjects: the
effects of third-party intervention. Journal of Studies on Alcohol, 37: 917–30.
Taylor, S.P. and Sears, J.D. (1988) The effects of alcohol and persuasive social pressure on
human physical aggression. Aggressive Behavior, 14: 237–43.
Terry, D.J. and O’Leary, J.E. (1995) The Theory of Planned Behaviour: the effects of
perceived behavioural control and self-efficacy. British Journal of Social Psychology, 34:
199–220.
Terry, P.C., Lane, A.M. and Fogarty, G.J. (2003) Construct validity of the Profile of Mood
States: adolescents for use with adults. Psychology of Sport and Exercise, 4: 125–39.
Terry, P.C. and Slade, A. (1995) Discriminant effectiveness of psychological state
measures in predicting performance outcome in karate competition. Perceptual and
Motor Skills, 81: 275–86.
Terry, P.C. and Youngs, E.L. (1996) Discriminant effectiveness of psychological state
measures in predicting selection during field hockey trials. Perceptual and Motor Skills,
82: 371–7.
Theodorakis, Y. (1994) Planned behavior, attitude strength, role identity, and the
prediction of exercise behavior. The Sport Psychologist, 8: 149–65.
Theodorakis, Y. (1995) Effects of self-efficacy, satisfaction, and personal goals on
swimming performance. The Sport Psychologist, 9: 245–53.
Theodorakis, Y., Doganis, G., Bagiatis, K. and Gouthas, M. (1991a) Preliminary study
of the ability of Reasoned Action Model in predicting exercise behavior in young
children. Perceptual and Motor Skills, 72: 51–8.
Theodorakis, Y., Goudas, M., Bagiatis, K. and Doganis, G. (1991b) Reasoned action
theory and the prediction of training participation in young swimmers. British Journal
of Physical Education, 10: 10–13.
Thompson, R.A. and Sherman, R.T. (1999) ‘Good athlete’ traits and characteristics of
anorexia nervosa: are they similar? Eating Disorders: The Journal of Treatment and Preven-
tion, 7: 181–90.
Toro, J., Salamero, M. and Martinez, E. (1994) Assessment of sociocultural influences on
the aesthetic body shape model in anorexia nervosa. Acta Psychiatrica Scandinavica, 89:
147–51.
Trafimow, D. and Finlay, K.A. (1996) The importance of subjective norms for a minority
of people: between-subjects and within-subjects effects. Personality and Social
Psychology Bulletin, 22: 820–8.
Trafimow, D. and Sheeran, P. (1998) Some tests of the distinction between cognitive and
affective beliefs. Journal of Experimental Social Psychology, 34: 378–97.
Trafimow, D., Sheeran, P., Conner, M. and Finlay, K.A. (2002) Evidence that perceived
behavioral control is a multidimensional construct: perceived control and perceived
difficulty. Journal of Applied Social Psychology, 41: 101–21.
Trafimow, D., Triandis, H.C. and Goto, S.G. (1991) Some tests of the distinction between
the private self and the collective self. Personality and Social Psychology Bulletin, 60:
649–55.
258 Bibliography
Treasure, D.C. and Roberts, G.C. (1998) Relationship between female adolescents’
achievement goal orientations, perceptions of the motivational climate, belief about
success and sources of satisfaction in basketball. International Journal of Sport Psych-
ology, 29: 211–30.
Triandis, H.C. (1995) Individualism and Collectivism. Boulder, CO: Westview Press.
Triplett, N. (1898) The dynamogenic factors in pacemaking and competition. American
Journal of Psychology, 9: 507–33.
Tucker, L.W. and Parks, J.B. (2001) Effects of gender and sport type on intercollegiate
athletes’ perceptions of the legitimacy of aggressive behaviors in sport. Sociology of
Sport Journal, 18: 403–13.
Turman, P.D. (2003) Coaches and cohesion: The impact of coaching techniques on team
cohesion in the small group sport setting. Journal of Sport Behavior, 26: 86–103.
UK Health Education Authority and Sports Council. (1992) Allied Dunbar National Fitness
Survey. London: UK Sports Council.
Vallerand, R.J. (1997) Towards a hierarchical model of intrinsic and extrinsic motivation.
In M.P. Zanna (ed.), Advances in Experimental Social Psychology (pp. 271–359). New
York: Academic Press.
Vallerand, R.J., Gauvin, L.I. and Halliwell, W.R. (1986) Effects of zero-sum competition
on children’s perceived competence and intrinsic motivation. Journal of Social Psych-
ology, 126: 465–72.
Vallerand, R. J. and Reid, G. (1984) On the causal effects of perceived competence on
intrinsic motivation: a test of cognitive evaluation theory. Journal of Sport Psychology,
6: 94–102.
van Goozen, S., Frijda, N. and Van de Poll, N. (1994) Anger and aggression in women:
influence of sports choice and testosterone administration. Aggressive Behavior, 20:
213–22.
Vansteenkiste, M. and Deci, E.L. (2003) Competitively contingent rewards and intrinsic
motivation: can losers remain motivated? Motivation and Emotion, 27: 273–99.
Varca, P. (1980) An analysis of home and away game performance of male college
basketball teams. Journal of Sport Psychology, 2: 245–57.
Veale, D. (1995) Does primary exercise dependence really exist? In J. Annett, B. Cripps
and H. Steinberg (eds), Exercise Addiction: Motivation for Participation in Sport and
Exercise (pp. 1–5). Leicester: British Psychological Society.
Vealey, R.S. (1986) Conceptualization of sport-confidence and competitive orientation:
preliminary investigation and instrument validation. Journal of Sport Psychology, 8:
221–46.
Verplanken, B. and Faes, S. (1999) Good intentions, bad habits, and effects of forming
implementation intentions on healthy eating. European Journal of Social Psychology, 29:
591–604.
Verplanken, B. and Orbell, S. (2003) Reflections on past behavior: a self-report index of
habit strength. Journal of Applied Social Psychology, 33: 1313–30.
Vlachopoulos, S. and Biddle, S.J.H. (1997) Modeling the relation of goal orientations to
achievement-related affect in physical activity: does perceived ability matter? Journal
of Sport and Exercise Psychology, 19: 169–87.
Vlachopoulos, S., Biddle, S.J.H. and Fox, K.R. (1997) Determinants of emotion in
children’s physical activity: a test of goal perspectives and attribution theories.
Pediatric Exercise Science, 9: 65–79.
Voight, M. and Callaghan, J. (2001) A team building intervention program: application
and evaluation with two university soccer teams. Journal of Sport Behavior, 24: 420–31.
Bibliography 259
Walters, E.E. and Kendler, K.S. (1995) Anorexia nervosa and anorexic-like syndromes in a
population-based female twin sample. American Journal of Psychiatry, 152: 64–71.
Wann, D.L., Carlson, J.D., Holland, L.C., Jacob, B.E., Owens, D.A. and Wells, D.D. (1999)
Beliefs in symbolic catharsis: the important of involvement in aggressive sports. Social
Behavior and Personality, 27: 155–64.
Wannamethee, S.G. and Shaper, A.G. (2001) Physical activity and the prevention of
cardiovascular disease. Sports Medicine, 31: 101–14.
Ward, R.E. (2002) Fan violence: social problem or moral panic. Aggression and Violent
Webber, L.S., Baugh, J.G., Cresanta, J.L. and Berenson, G.S. (1983) Transition of
cardiovascular disease risk factors from adolescence to young adulthood: the Bogalusa
post-high school study. Circulation, 68(Suppl.)(3): 160.
Weed, M. (2000) The social dynamics of sports groups: arguments for a meso-level
analysis. In J. Avela, P.V. Komi and J. Komulainen (eds), Proceedings of the 5th Annual
Congress of the European College of Sport Science. Javaskyla, Finland: LIKES Research
Centre.
Weed, M. (2001) Ing-ger-land at Euro 2000: how ‘handbags at 20 paces’ was portrayed as
a full-scale riot. International Review for the Sociology of Sport, 36: 407–24.
Weinberg, R.S. (1986) Relationship between self-efficacy and cognitive strategies
in enhancing endurance performance. International Journal of Sport Psychology, 17:
280–92.
Weinberg, R.S., Yukelson, D. and Jackson, A. (1980) Effect of public and private efficacy
expectations on competitive performance. Journal of Sport Psychology, 2: 340–9.
Weiner, B., Heckhausen, H., Meyer, W.U. and Cook, R.E. (1972) Causal ascriptions and
achievement motivation: a conceptual analysis and reanalysis of locus of control.
Journal of Personality and Social Psychology, 21: 239–48.
Weiss, M.R., Wiese, D.M. and Klint, K.A. (1989) Head over heels with success: the rela-
tionship between self-efficacy and performance in competitive youth gymnastics.
White, R.W. (1959) Motivation reconsidered: the concept of competence. Psychological
Review, 66: 297–333.
White, S.A. (1998) Adolescent goal profiles, perceptions of the parent-initiated
motivational climate, and competitive trait anxiety. The Sport Psychologist, 12:
16–28.
Whitehead, J.R. (1995) A study of children’s physical self-perceptions using an adapted
physical self-perception profile questionnaire. Pediatric Exercise Science, 7: 132–51.
Whitehead, J.R. and Corbin, C.B. (1991) Youth fitness testing: the effect of percentile-
based evaluative feedback on intrinsic motivation. Research Quarterly for Exercise and
Sport, 62: 225–31.
Whitehead, J.R. and Corbin, C.B. (1997) Self-esteem in children and youth: the role of
sport and physical education. In K.R. Fox (ed.), The Physical Self. Champaign, IL:
Human Kinetics.
Widmeyer, W.N. (1990) Group composition in sport. International Journal of Sport
Widmeyer, W.N., Brawley, L.R. and Carron, A.V. (1990) The effects of group size in sport.
Widmeyer, W.N. and Ducharme, K. (1997) Team building through team goal setting.
Journal of Applied Sport Psychology, 9: 97–113.
Wild, T.C. and Enzle, M.E. (2002) Social contagion of motivational orientations. In E.L.
260 Bibliography
Deci and R.M. Ryan (eds), Handbook of Self-determination Research (pp. 141–57).
Rochester, NY: University of Rochester Press.
Williams, G.C., Gagne, M., Ryan, R.M. and Deci, E.L. (2002) Facilitating autonomous
motivation for smoking cessation. Health Psychology, 21: 40–50.
Williams, G.C., Rodin, G.C., Ryan, R.M., Grolnick, W.S. and Deci, E.L. (1998) Auto-
nomous regulation and long-term medication adherence in adult outpatients. Health
Williams, J.M. and Widmeyer, W.N. (1991) The cohesion-performance outcome
relationship in a coaching sport. Journal of Sport and Exercise Psychology, 13: 364–71.
Williams, K.E. and Bond, M.J. (2002) The roles of self-efficacy, outcome expectancies and
social support in the self-care behaviours of diabetics. Psychology, Health and Medicine,
7: 127–41.
Williams, L. (1994) Goal orientations and athlete’s preference for competence
information sources. Journal of Sport and Exercise Psychology, 16: 416–30.
Williams, L. and Gill, D.L. (1995) The role of perceived competence in the motivation of
physical activity. Journal of Sport and Exercise Psychology, 17: 363–78.
Wise, J.B. and Trunnell, E.P. (2001) The influence of sources of self-efficacy upon efficacy
strength. Journal of Sport and Exercise Psychology, 23: 268–80.
Wrangham, R. and Peterson, D. (1996) Demonic Males. New York: Houghton Mifflin.
Yan Lan, L. and Gill, D.L. (1984) The relationship among self-efficacy, stress responses,
and a cognitive feedback manipulation. Journal of Sport Psychology, 6: 227–38.
Yates, A.B. (1991) Compulsive Exercise and the Eating Disorders. New York: Brunner/Mazel.
Yates, W.R. (1999) Medical problems of the athlete with an eating disorder. In P.S. Mehler
and A.E. Andersen (eds), Eating Disorders: A Guide to Medical Care and Complications
(pp. 153–66). Baltimore, MD: Johns Hopkins University Press.
Yerkes, R.M. and Dodson, J.D. (1908) The relation of strength of stimulus to rapidity of
habit formation. Journal of Comparative Neurology and Psychology, 18: 459–82.
Yoo, J. (2003) Motivational climate and perceived competence in anxiety and tennis
performance. Perceptual and Motor Skills, 96: 403–13.
Zaccaro, S.J., Blair, V., Peterson, C. and Zazanis, M. (1995) Collective efficacy. In J.E.
Maddux (ed.), Self-efficacy, Adaptation, and Adjustment: Theory, Research, and Application
(pp. 305–28). New York: Plenum Press.
Zajonc, R.B. (1965) Social facilitation. Science, 149: 269–74.
Zeeman, E.C. (1976) Catastrophe theory. Scientific American, 234: 65–82.
Zillman, D., Johnson, R.C. and Day, K.D. (1974) Attribution of apparent arousal and
proficiency of recovery from sympathetic activation affecting excitation transfer to
aggressive behavior. Journal of Experimental Social Psychology, 10: 503–15.
Zimbardo, P. (1970) The human choice: individuation, reason, and order versus
individuation, impulse, and chaos. In W.J. Arnold and D. Levine (eds), Nebraska
Symposium on Motivation (Vol. 17, pp. 237–307). Lincoln, NE: University of Nebraska
Press.
Zuckerman, M., Kuhlman, D.M., Joireman, J., Teta, P. and Kraft, M. (1993) A comparison
of three structural models for personality: the Big Three, the Big Five, and the
Alternative Five. Journal of Personality and Social Psychology, 65: 757–68.
Author index
Abraham, C., 17, 34 Deci, E., 15, 46, 50–6, 58–9, 62, 67–8, 71,
Aiello, J., 177, 180–3 100, 118–21, 129, 161, 217
Ames, C., 62, 113, 116–17 Duda, J., 99, 113–15, 128, 143–4, 222
Ajzen, I., 13, 15, 20–9, 33, 44–6, 48–9, 100, Dunning, E., 205, 209–10
105, 217, 221–3 Dweck, C., 113
Apter, M., 156, 208, 216–17
Armitage, C., 23–4, 27, 31, 33, 38–40 Eagly, A., 13, 30, 131, 165, 200
Eys, M., 172, 175
Bagozzi, R., 2, 45, 132–3
Bandura, A., 23, 32, 71, 83, 100, 105–10, Fazio, R., 28–9, 31
128, 165–6, 174–5, 197, 205, 211 Feltz, D., 105, 108–11, 113, 128, 166–7,
Bargh, J., 28–9, 47 219
Baumeister, R., 130, 161, 184, 190–1, Festinger, L., 46, 164, 206, 208
218 Fox, K., 10, 77–9, 81–2, 86, 94, 220, 223
Beauchamp, M., 174–5, 194
Beedie, C., 138, 155–6 Gill, D., 143, 164
Biddle, S., 81, 103–4, 106, 115, 128, 219 Godin, G., 24, 28, 42
Brawley, L., 12–14, 24, 172 Gollwitzer, P., 15, 43, 46–7, 49, 68
Bray, S., 167, 184, 187 Gould, D., 111, 113, 133, 135–6, 142, 150,
Burton, D., 135–8, 140, 145, 147 152
Guay, F., 58, 125, 127, 220, 223
Cale, L., 11, 14
Carron, A., 161–2, 164, 168–70, 172–3, Hagger, M., 13, 20–1, 24–5, 28, 31–2, 35,
191–2, 218–19 41, 44–5, 58–60, 62, 68, 73, 79, 81–2,
Chaiken, S., 28–29, 60 85–6, 123, 127, 131, 219, 221–2, 224
Chatzisarantis, N., 21, 24, 28–9, 44, 46, 48, Hall, H., 112, 143
58–60, 62, 128, 222 Hanin, Y., 151–3, 155, 158–9, 216–17
Conner, M., 23, 27, 29–33, 35, 41, 43–4 Hardy, L., 141, 148–50, 159, 218
Cottrell, N., 176–8 Harter, S., 72, 76–7
Courneya, K., 9, 26, 29–30, 32–3, 35, 39, Harwood, C., 116, 146
45, 184, 188, 218 Hausenblas, H., 20–1, 25, 90–3, 219, 221
Craft, L., 136, 138, 158–9, 219 Heider, F., 100–1, 128
262 Author index
Iyengar, S., 124 Roberts, G., 103, 115

Rogers, R., 17, 49
Jones, G., 133, 139–40, 145–9, 158–9, 217, Rollnick, S., 63, 66
223 Rosenberg, M., 72, 74, 76, 84
Jones, M., 113, 184, 188–9, 191 Russell, G., 183, 185–6, 188, 193,, 197–8,
211
Kavussanu, M., 110, 115, 117 Ryan, R., 58, 89, 91, 114, 120, 222
Kerr, J., 157–9, 205, 208–9, 216,
Koestner, R., 47–9, 123 Sallis, J., 10–11, 81
Sarrazin, P., 59, 115, 117, 120, 127
Landers, D., 133, 140 Sedikides, C., 73
Schwartz, B., 184, 186–8, 191
Markland, D., 58, 63, 67–8 Shavelson, R., 74–7, 82–3
Marsh, H., 73–4, 76–83, 85, 218, 223 Sheeran, P., 24, 43–7, 49, 60
Marshall, S., 39, 41 Sheldon, K., 50–1, 55
McAuley, E., 54, 103, 109, 111, 121, 219 Sherman, S., 29
McCrae, R., 29 Silva, J., 184, 186, 194–6, 198
Martens, R., 133, 135–8, 142, 159 Sonstroem, R., 77–9, 81, 83–5, 218
Maslow, A., 100 Spielberger, C., 134–5
Meiland, J., 20, 46, 100 Standage, M., 59, 222
Miller, W., 63, 66–7 Sutton, S., 39–40
Milne, S., 17, 19, 47, 220
Morgan, W., 77, 81, 83–4, 151, 153–5, 218 Tajfel, H., 161, 164, 204–5, 207–8
Terry, P., 138, 155–6
Nicholls, J., 100, 113–14, 116, 129 Theodorakis, Y., 21, 30, 110, 112, 219, 222
Ntoumanis, N., 55, 58–9, 115, 117, 120, Trafimow, D., 30–1, 33, 35, 85
144, 219 Triplett, N., 176, 180
Orbell, S., 44, 46–7, 49 Vallerand, R., 60, 116, 120–1, 125–8, 142,
Oxendine, J., 140 220, 223
Varca, P., 184–5, 187, 218
Perugini, M., 36–7, 41, 131, 217
Petty, R., 26, 165 Weed, M., 203–5, 210–13
Prestwich, A., 47–9, 220 Weinberg, R., 110, 221
Prochaska, J., 38–9 Weiner, B., 101–4, 106, 128
Whitehead, J., 54, 81–2, 120
Quine, L., 17, 24–6 Widmeyer, W., 163, 166–7, 172–3
Williams, G., 58–9, 123
Reeve, J., 54, 62, 76, 99, 120–3
Reicher, S., 206–7 Yates, A., 88–94
Rhodes, R., 29–30, 32, 223
Rivis, A., 32, 218 Zajonc, R., 130, 160, 176–9, 187
Subject index
achievement goal theory definition of, 133–4

anxiety and, 143–5 distinction from arousal, 133
definition of, 100–1 facilitative and debilitative, 145–7
goal involvement, 115–16 self-confidence component, 134,
group processes, 167–8 136–9, 142–4, 146–7, 150, 156, 159
in sport, 113–15, 127–9 trait, 142–3
see also goal orientation; motivational trait-state distinction, 134–5
climate see also cognitive anxiety; somatic
affect anxiety, self-confidence
as a component of attitude, 24–5; 30–1, amotivation, 55, 125
35, 42 applied research, 13, 15
anticipated, 34–5 see also formative research
definition of, 131–2 arousal
aggression aggression and, 194, 196–7, 201–2
alcohol abuse and, 212–13 catastrophe theory and, 148–50
definition of, 193–6 definition of, 133–4
frustration-aggression hypothesis, distinction from affective components,
196–7, 213 131–2
home advantage and, 187 home advantage and, 187, 190
and hooliganism, 202–4 inverted-U hypothesis, 139–41
hostile and instrumental types, 194–6 in motivational theories, 100, 108
and the media, 211–12 reversal theory and, 156–7, 208–9
social learning theory and, 197–8 social facilitation and, 177–8
see also collective aggression attitude
anorexia nervosa, 86–7, 92, 95 affective, 25, 30–1, 132, 217
see also bulimia nervosa; eating ambivalence, 30, 40–2
disorders change, 24–7
anticipated regret, 33–5, 42 cognitive/instrumental, 30–1, 35, 42
see also theory of planned behaviour relationship with intention, 20–1,
anxiety 23–5
anxiety-performance hypothesis, 135, relationship with behaviour, 20
147, 159 strength, 30
264 Subject index
see also behavioural beliefs; expectancy see also intention; implementation

× value model; outcome evaluation; intention; theory of planned
theory of planned behaviour; theory behaviour
of reasoned action coronary heart disease, 8
attribution theory, 100–4 see also cardiovascular disease
automaticity, 29 correspondence
automatic processes, 28, 37, 47 boundary condition, 22, 105
autonomous motivation, 56, 223–4, 226 scale, 45, 69
see also intrinsic motivation; self- see also theory of planned behaviour;
determination theory theory of reasoned action
behavioural beliefs, 20–2, 25–7, 39 decisional balance, 38–9, 42, 48, 66

see also expectancy × value; outcome deliberative behavioural processes,
evaluation; theory of planned 28–9
behaviour; theory of reasoned action descriptive norms, 32–3, 42, 218
bulimia nervosa, 86–7, 92, 95 see also theory of planned behaviour
see also anorexia nervosa; eating desires, 36–7, 42, 217
disorders see also intention; model of
goal-directed behaviour
cardiovascular disease, 8–11, 14, 16–17 dieting behaviour, 17, 71, 86, 91–2
cognitive anxiety see also obesity
achievement goals and, 143–5 discontinuity patterns, 39–40, 42
in catastrophe theory, 148–50, 217
definition of, eating disorders, 72, 86–9, 95
134–6 see also anorexia nervosa; bulimia
intensity vs direction distinction, 145–7 nervosa
relation to inverted-u hypothesis, elaboration likelihood model, 165
139–41 emotion
cognitive behavioural therapy, 93–5 in attribution theory, 101, 104
cognitive dissonance theory, 46, 67, 164 cognition and, 131–3
cognitive evaluation theory definition of, 131–2
rewards, 52–4, 59, 118–22, 129 distinction from affect, 132
informational and controlling feedback, distinction from mood, 132, 153
52, 54, 59, 62, 67, 118–19, 121–3, links with self-esteem, 76
225 in self-determination theory, 56
in sport, 117–22 in social cognitive theory, 108
in exercise, 50–5, 69 stability, 76, 91
see also self-determination theory see also cognitive-motivational-
cognitive-motivational-relational theories relational theories of emotion
of emotion, 104, 133, 143–7 emotion-focused coping, 117, 147
collective aggression, 202–5, 207, evaluation apprehension, 177–9, 192
211–12 see also social facilitation
collective efficacy, 162, 165–8, 172–4, exercise behaviour
180–1, 192 addiction and dependency, 90–4
conflict theory, 205, 210–11, 214 changing, 19, 24–7, 34–6, 49
control beliefs, 23–4, 26–7 definition of, 7–8
see also expectancy × value; theory of distinction from physical activity
reasoned action and sport, 7–8
continuation intentions, 48–9 see also physical activity
Subject index 265
exercise and physical self-esteem model, hierarchical model

81, 83–5, 218 of physical self-esteem, 74–81, 84, 94,
see also physical self-esteem; 220, 223
hierarchical model of motivation, 125–8, 220, 223
expectancy × value model, 20–3 see also self-determination theory; self-
extrinsic motivation, 50–1, 55–6, 58–60, esteem
117, 120, 125–6 home advantage
external regulation, 55–6, 58, 70, 216–17, definition of, 183–4
222 factors affecting, 185–90
see also organismic integration theory; home disadvantage, 190
self-determination theory prevalence, 184
factor analysis, 84, 136, 219 identified regulation, 51, 55–6, 58, 67, 120,
feedback 222
controlling and informational, 53–5, 59, see also organismic integration theory;
62, 69, 112, 118, 118–23, 225 self-determination theory
false feedback in experiments, 109, imagery, 108, 112–13, 147
220 implementation intention, 46–9, 69
positive, 54, 67, 107, 204, 222 individual difference, 116, 150, 198,
see also cognitive evaluation theory; 203–5, 208
self-determination theory information processing, 28, 103, 203–5
figurational approach, 205, 209–10, 214 integrated regulation, 51, 55–6, 58, 69
formative research, 13–14, 17, 19, 21, see also organismic integration theory;
24–5, 27, 35 self-determination theory
see also applied research intention
distinction from desire, 36–7
goal formation, 16, 45, 49, 69
directed intention, 46–48 stability, 44, 69
setting, 94, 109, 112, 147, 173, 219 volitional and forced, 46
striving, 99, 125 see also implementation intention;
goal orientation, 113–17, 129, 141, 143–5, theory of planned behaviour; theory
216, 222, 226 of reasoned action
see also achievement goal theory internalization, 55–6, 67
group norms, 124, 161–2, 164–5, 168, see also organismic integration theory;
200–1, 207 self-determination theory
group cohesion interventions
definition of, 168–9 based on health belief model/protection
conceptual framework, 169–70 motivation theory, 19
interventions, 173 based on extensions to social cognitive
relationship with performance, theories, 34–6
170–2 based on theory of planned
relationships with outcomes, 172, behaviour/reasoned action, 24–7
218–21 based on volitional strategies, 49
role behaviour, 173–4 intrinsic motivation
in cognitive evaluation theory, 52–5
habit, 28–9 definition of, 50–1
see also past behaviour in exercise, 50–1
health belief model, 16–19, 24, 5, 41, 216, in hierarchical model, 125–8
219 in sport, 110, 114–15, 117–25,
266 Subject index
in trans-contextual model, 59–60 attribution theory and, 101–4

rewards and, 52–5 cognitive evaluation theory and, 52–3
see also cognitive evaluation theory; definition in exercise, 15–16
organismic integration theory; definition in sport, 100–1
rewards; self-determination theory emotion and, 130, 132
introjected regulation, 51, 55–6, 58, 91–2 group cohesion and, 168, 172
see also organismic integration theory; hierarchical model and, 125–8
self-determination theory organismic integration theory and,
IZOF (individual zones of optimal 55–6
functioning) model, 150–3, 159 physical self-esteem and, 71, 73, 77
protection motivation, 17–19
mediation, 17, 20, 27, 29–30, 36, 37, 42, relation to intentions, 20
58–60, 62, 72–3, 77, 83–5, 105, 107, relation to volitional theories, 27, 46–7
112–13, 120, 125–6, 141–4, 162, 166–9, reversal theory and, 156–8
174, 177, 181, 183, 192, 195, 199, self-determination theory and, 50–1,
217, 219 117–20
menu of strategies, 64–5, 67, 69 social cognitive theory and, 105–8
see also motivational interviewing social facilitation and, 179
meta-analysis strategies to enhance, 47–9
anxiety-performance relationship, 136, see also intention; cognitive evaluation
138–9 theory; desire; organismic integration
cognitive evaluation theory, 53, 119–20 theory; self-determination theory;
cohesion-performance relationships, theory of planned behaviour; theory
170–1 of reasoned action
descriptive norms, 32 motivational climate, 116–17, 127–8,
IZOF, 152 144–5, 167–8, 181, 225
mood states in sport, 154–6 motivational interviewing, 63–9, 225
physical self-esteem, 82 motivation to comply, 20–1
protection motivation theory, 17–18, 41 see also expectancy × value; normative
social facilitation, 178 beliefs; theory of planned behaviour;
theory of planned behaviour, 20–1, 23–4 theory of reasoned action
transtheoretical model, 39
model of goal directed behaviour, 36–7, 42, normative beliefs, 20–2, 24, 26–7, 42
216–17, 219 see also expectancy × value; motivation
moderation 24, 32–4, 40, 45, 73, 105, to comply; theory of planned
111–12, 138–9, 145, 147, 156, 163–4, behaviour; theory of reasoned action
167, 179–80, 183, 188, 190, 192, 197,
199, 211, 213, 217, 219 obesity, 7–10, 65
mood organismic integration theory, 50–1, 55–9,
activity disorders and, 91 69–70, 222
definition of, 131–2 outcome evaluation, 20, 48
eating disorders and, 87 see also behavioural beliefs; expectancy
iceberg profile, 153–4 × value theory of planned behaviour;
states in sport, 153–6 theory of reasoned action
moral norms, 33–4, 42, 214 outcome expectations, 108–9
see also theory of planned behaviour see also social cognitive theory
motivation
achievement goal theory and, 113–15 past behaviour, 28–9, 31, 35, 37, 42, 62
activity disorders and, 91 see also habit
Subject index 267
physical activity reversal theory

definition of, 7–8 definition of, 157
guidelines, 11, 14 relationship with arousal, 157–8, 159
physical inactivity, 1, 7, 9, 11–12 collective aggression, 205, 208–9, 214
see also exercise rewards, 52–4, 59, 118–22, 129, 130
physical self-esteem see also cognitive evaluation theory;
in cross-cultural research, 85–6 self-determination theory
in elite athletes, 83, 88, 95 role
multidimensional model, 77–81 ambiguity, 173–5, 192
in young people, 82 conflict, 174–5
see also hierarchical model efficacy, 173–5
physical self-perceptions, 77, 81–2, 85 performance, 174–6
see also hierarchical models; physical
self-esteem self-determination theory
perceived autonomy support, 58–62, 123, authoritarian and democratic styles,
127 123–5
see also organismic integration theory; causality orientations theory, 50
self-determination theory competition, 121–3, 129
perceived barriers, 16–17, 23–4, 225 definition of, 50–1, 117–18
perceived behavioural control and in exercise, 50–1, 55, 59–60, 62–3,
perceived controllability, 23–7, 32–4, 67–9
36–7, 39, 42, 60–2, 216, 225 in sport, 117–29
perceived power, 23 see also external regulation; extrinsic
see also control beliefs; expectancy × motivation; organismic integration
value; theory of planned behaviour theory; identified regulation; intrinsic
personality motivation; internalization;
activity disorders and, 90, 92 introjected regulation
collective aggression and, 198–9 self-efficacy
eating disorders and, 88–9 in achievement goal theory, 117, 129
state-trait distinction, 134–5 antecedents, 111–12
in theory of planned behaviour, 27, definition of, 105
29–30, 35, 42 in exercise and self-esteem model, 83–5
plans and planning, 20, 37, 47–8, 66–7, experimental designs, 109–10, 222–3
144 and goal setting/imagery, 112–13
see also intention; implementation in motivational interviewing, 64, 67, 69
intentions; theory of planned in protection motivation theory, 17–19
behaviour; theory of reasoned action in social cognitive theory, 105–11, 128
perceived locus of causality (PLOC), 52–3, task, 168–9, 174
57–62, 69, 217, 222 in theory of planned behaviour, 23, 25,
see also cognitive evaluation theory; 33, 42
organismic integration theory; self- in transtheoretical model, 38–9
determination theory see also feedback; outcome expectations;
protection motivation theory, 17–20, perceived behavioural control; social
24–5, 41, 219 cognitive theory; theory of planned
persuasion, 26, 35, 63, 106–8, 11, 129, 166 behaviour; theory of reasoned action;
POMS (profile of mood states), 153–5 transtheoretical model
see also mood; emotion social cognition, 15, 24, 39, 131–3, 160,
problem -focused coping and active 198
coping, 117, 144 see also information processing
268 Subject index
social cognitive theory, 105–7, 109–10, see also anxiety; cognitive anxiety;
112, 128, 219, 226 self-confidence
see also self-efficacy spontaneous behavioural processes, 29, 31
social facilitation sport
definition of, 176 definition of, 7–8
evaluation apprehension, 177–9 importance of, 1
integrative model, 181–3 psychologist/practitioner, 3, 131, 135,
role of cognition, 179 183
see also evaluation apprehension; social see also exercise; physical activity
loafing stages of change, 38–40, 65, 67,
social identity theory, 71, 164, 201, 204–8, see also transtheoretical model
214 subjective norms, 20–4, 25–7, 30–4, 36–7,
social learning theory, 106, 165, 197–8, 42–3, 60, 62, 216, 218
205, 211, 213
see also aggression; social cognitive theory of planned behaviour, 22–36,
theory 39–40, 42–4, 46–9, 60, 63, 131,
social loafing, 163, 166, 172, 179–81, 186, 216–19, 221–5
192 theory of reasoned action, 20–5, 42
social support, 32–3, 42, 172, 218 trans-contextual model, 59–62, 69, 216,
see also theory of planned behaviour 219
self-schema, 45–6, 69 transtheoretical model, 38–42
somatic anxiety see also stages of change
definition of, 134–6
intensity vs direction distinction, 145–7 volitional strategies, 47, 49
relation to inverted-u hypothesis,
139–41 within-participants analyses, 31
APPLYING SOCIAL PSYCHOLOGY
SERIES EDITOR: STEPHEN SUTTON
THE SOCIAL PSYCHOLOGY OF EXERCISE AND SPORT
The Social Psychology of Exercise and Sport

This new textbook examines the role that social psychology has in the
explanation of exercise and sport behaviour. It devotes considerable attention
to key social psychological issues within the two disciplines; health-related
exercise behaviour and the behaviour of competitive sport participants and
the spectators of elite sport. The
Rather than presenting a broad, superficial overview of diverse areas in
exercise and sport, the book focuses on a range of selected topics and provides
a comprehensive, in-depth and analytical coverage using social psychology as
a framework. It thoroughly examines how social psychological research and
Social Psychology
intervention has contributed to the understanding of key topics in exercise
and sport behaviour including:
• The social psychology of exercise and health of Exercise
• Social cognitive theories of exercise behaviour
• Exercise and the physical self
• Eating disorders in exercise and sport
• Emotion and mood in athletes and Sport
• Social psychology and motivation in sport
• Group processes in sport
• Aggression and crowd violence
The Social Psychology of Exercise and Sport is key reading for undergraduate
and postgraduate students on social or sport psychology courses and on
health-related or sports science courses. Illustrated throughout with practical
guidelines for researchers and practitioners, it is also a valuable resource for
professionals interested in understanding and changing the behaviour of
exercise participants and athletes.
Cover design: Kate Prentice
Martin Hagger is Lecturer in Social and Health Psychology at the University
of Essex. He is also a qualified Chartered Health Psychologist with the British
Psychological Society (BPS) and an accredited Sports Scientist (Psychology)
with the British Association of Sport and Exercise Sciences. He has diverse
research interests in the fields of health and social psychology.
Nikos Chatzisarantis is a lecturer in the School of Sport and Health Sciences

Hagger • Chatzisarantis
at the University of Exeter. His research interests span the fields of philosophy,
social psychology, and quantitative research methodology.
Martin Hagger and

ISBN 0-335-21618-8
9 780335 216185

The Social Psychology of Exercise and Sport 2005 (For Lydia

Uploaded by

Copyright:

Available Formats

The Social Psychology of Exercise and Sport 2005 (For Lydia

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Social Psychology of Exercise and Sport 2005 (For Lydia

Uploaded by

Copyright:

Available Formats

APPLYING SOCIAL PSYCHOLOGY

SERIES EDITOR: STEPHEN SUTTON

The Social Psychology of Exercise and Sport

Nikos Chatzisarantis is a lecturer in the School of Sport and Health Sciences

Martin Hagger and

Richard P. Bagozzi, Zeynep Gürhan-Canli and Joseph R. Priester: The Social

Martin Hagger and Nikos Chatzisarantis

Open University Press

First published 2005

Series editor’s foreword vii

Part I The social psychology of exercise 5

Part II The social psychology of sport 97

Social psychology is sometimes criticized for not being sufficiently ‘relevant’ to

work on performance in sport draws on the traditional social psychological

examination of how social psychological research and intervention have

behaviour. We have focused on the individual as the unit of analysis, as is the

an applied social psychological perspective have as many similarities as they

The social psychology of exercise

Social psychology, exercise, and health

What is physical activity?

as losing weight. Examples of these types of activity are jogging, cycling,

Evidence for the effect of physical activity on health

Physical activity and chronic disease in adults

can act as a preventive measure against osteoporosis – a chronic disease that

Physical activity and young people

Guidelines for physical activity

The physical inactivity epidemic

The role of social psychology

One way to promote exercise participation is to consult theories of social

explain more than 50 per cent of variance in physical activity behaviour

• Physical activity refers to all bodily movement that expends energy.

Social cognitive theories of exercise

Understanding regular participation in exercise requires an understanding of

Intentions and social cognitive models of exercise behaviour

Many important social psychological theories of human motivation incorpor-

The health belief model

Figure 2.1 The health belief model

Protection motivation theory

Designing interventions based on the health belief model and the

The theory of reasoned action

Figure 2.3 Theories of reasoned action and planned behaviours

Boundary conditions of the theory of reasoned action

The theory of planned behaviour

acknowledges that intentions explain behaviour only when it is under

control on behavioural intentions. In addition, it has been shown that the

Designing interventions based on the theories of reasoned action and

This persuasive appeal should be effective in changing young people’s

in changing attitudes, subjective norms, perception of control, and inten-

Sufficiency of the theory of planned behaviour

The theory of planned behaviour focuses mainly on the immediate determin-

Frequency of past behaviour and habit

on intention may also reflect recent performance of the behaviour. This is

that personality traits can facilitate non-intentional behaviour in the theory of

Attitude-based modifications of the theory of planned behaviour

Affective and cognitive attitudes

Normative extensions of the theory of planned behaviour

Normatively controlled individuals

perceive that behaviour is difficult to execute (Povey et al. 2000). Further, in

Control-related extensions of the theory of planned behaviour