Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Personality and situational coping: exploring the relationship
between neuroticism and coping in the context of university
examinations and two laboratory-stressors.
Mark Boyes
B.A. (1st Class Hons.) (Psychology)
School of Psychology
University of Western Australia
This thesis is presented as partial requirement for the degree of Doctor of
Philosophy/Master of Psychology (Educational and Developmental)
2009
II
Abstract
A substantial body of research has revealed that neuroticism and extraversion
are associated with coping responses. Neuroticism is positively correlated with
emotion-focused and avoidant coping strategies, whilst extraversion is negatively
correlated with these strategies. However, findings have been limited by a reliance
on dispositional and retrospective coping measures, which may overestimate the
relationship between personality and coping strategy use in a given situation. It has
been argued that examining the relationship between personality and coping, in the
context of situation-specific stressors would allow firmer conclusions to be made.
The aim of this thesis was to explore relationships between personality and coping
with situation-specific stressors, in which limitations of dispositional/retrospective
coping measurement are minimised.
In Study 1, participants were assigned to either a dispositional condition
(completed a dispositional coping measure) or a situational condition (completed a
coping measure with regard to approaching university examinations). Results
offered support for the contention that correlations between personality and coping
may be overestimated when coping is measured dispositionally. In the situational
condition only neuroticism was systematically related to coping. Based on this,
neuroticism became the focus of the remainder of the thesis. Limitations of using
examinations as a stressor were noted, and it was concluded that relationships
between neuroticism and coping should be explored in the context of laboratory-
stressors.
III
Studies 2- 4 explored the utility of a ball-throwing game (Cyberball), during
which the participant is included or excluded, as a laboratory-stressor to examine
relationships between neuroticism and coping. These experiments confirmed that 1)
Cyberball is a promising laboratory-stressor for exploring individual differences in
coping, and 2) neuroticism is consistently associated with emotion-focused coping
during Cyberball. This neuroticism-linked difference in emotion-focused coping
emerged regardless of whether participants were completely excluded, partially
excluded, or completely included during the game. A relationship between
neuroticism and avoidance coping was only obtained when ostracism was
ambiguous in nature. Additionally, a neuroticism-linked difference in appraisal was
only obtained in the ambiguous Cyberball condition.
Study 5 aimed to 1) replicate the previously reported neuroticism-related
difference in emotion-focused coping using a cognitive laboratory-stressor (an
anagram-solving task in which anagram difficulty and level of participant control
were manipulated), and 2) determine whether neuroticism-linked differences in
avoidance coping could also be obtained in the context of the anagram-solving task.
Results revealed that neuroticism was associated with emotion-focused and
avoidance coping, regardless of whether participants were assigned to the Mild
Stress or High Stress conditions. Additionally, findings offered preliminary
evidence that stressor appraisals may mediate the relationship between neuroticism
and both emotion-focused and avoidance coping.
The fact that in Studies 2-5 neuroticism-linked differences in emotion-
focused coping were obtained regardless of stress condition raised the possibility
IV
that neuroticism and emotion-focused coping measures were confounded by item
overlap; that is some emotion-focused coping items may be directly assessing
neuroticism. In Study 6, Confirmatory Factor Analysis was conducted on responses
to neuroticism and emotion-focused coping items collected in Studies 3-5.
Regardless of the level of stress the participants were exposed to, the predicted two
factor model provided good fit. Additionally, no evidence of item cross-loading was
obtained suggesting that the relationships obtained in the previous studies are not
simply due to content overlap between items on neuroticism and emotion-focused
coping measures.
It was concluded that neuroticism is associated with emotion-focused and
avoidance coping in the context of a real-world situation-specific stressor (university
examinations), as well as in the context of two laboratory-stressors. These findings
overcome the reliance on dispositional and retrospective coping measures that has
limited much previous research. Importantly, these findings do not appear to be due
to item confounds between neuroticism and emotion-focused coping measures,
although future research should examine this possibility in more detail. Future
research should also examine the possibility that relationships between neuroticism
emotion-focused, and avoidance coping are mediated by stressor appraisals.
V
Acknowledgements
Firstly, I would like to thank my supervisor, Davina French, for her guidance
over the course of this PhD; her support and patience has been invaluable. Simply
put I could not have asked for a better supervisor and mentor. Also, heartfelt thanks
go to my family and Lorna. I am extremely grateful for your words of
encouragement as well as the emotional, psychological, and financial support you
have provided over the years I have been studying. A huge thank-you to Martyn
Churcher for programming a version of the Cyberball game allowing the level of
ostracism to be manipulated; the regular sing-along sessions were also fun. Thanks
to Colin MacLeod for reviewing this thesis prior to its submission, and for inviting
me along to the regular Cognition and Emotion Lab group meetings. Your
enthusiasm is infectious and really helped get the ball rolling with regard to writing
up my own research. Thanks also to my office mate, Nic Badcock, and the rest of
the Child Study Centre crew for the continued friendship and many Monty Python
references. Finally, thanks to the Lunch Ladies. There is a strong possibility I
learned more from our lunch time conversations (at least with regard to ‘real-world’
validity) than I did over the course of my PhD. At least now I know that there are
some chicken fillets that just shouldn’t be barbecued.
The work in this thesis was supported by an Australian Postgraduate Award
and research funding from the School of Psychology at the University of Western
Australia.
VI
Publications Arising From this Thesis
This thesis contains published work and/or work prepared for publication which has
been co-authored. The bibliographic details of the work and where it appears is set
out below.
A paper based on Chapters 2, 5, and 7 has been published in Personality and
Individual Differences and is appended to this thesis.
Boyes, M. E., & French, D. J. (2009). Having a Cyberball: Using a ball-throwing
game as an experimental social stressor to examine the relationship between
personality and coping. Personality and Individual Differences, 47, 396-401.
Percentage contribution of: Candidate = 90% and DJF = 10%
All authors have given permission for work to be included in this thesis.
____________________ _____________________
Mark E. Boyes (Candidate) Davina J. French (Supervisor)
VII
A paper based on Chapter 2 and Chapter 9 has been prepared for submission to
Personality and Individual Differences.
Boyes, M. E. & French, D. J. Neuroticism and coping in the context of an anagram-
solving task. Personality and Individual Differences. Submitted.
Percentage contribution of: Candidate = 90% and DJF = 10%
All authors have given permission for work to be included in this thesis.
____________________ _____________________
Mark E. Boyes (Candidate) Davina J. French (Supervisor)
VIII
Contents
Abstract ....................................................................................................................... ii Acknowledgements ..................................................................................................... v Publications Arising From this Thesis ....................................................................... vi Contents ................................................................................................................... viii List of Tables............................................................................................................. xii List of Figures .......................................................................................................... xiv CHAPTER 1: A Brief Review of the Coping Literature .......................................... 15
Defining Coping .................................................................................................... 16 Conceptualising the Structure of Coping .............................................................. 18
The Problem-Focused/Emotion-Focused Coping Distinction .......................... 19 The Primary/Secondary Control Distinction ..................................................... 20 Approach/Avoidance Distinction ...................................................................... 20 Convergence of Classification Systems ............................................................ 21
Why Study Coping? .............................................................................................. 22 Effectiveness of Coping Strategies ....................................................................... 24 The Goodness of Fit Hypothesis ........................................................................... 27 Summary ............................................................................................................... 30
CHAPTER 2: Reviewing the Relationship between Personality and Coping .......... 32 Personality and Coping ......................................................................................... 33
The Five-Factor Model of Personality .............................................................. 36 Why Might General Personality Traits be Associated with Coping? ............... 37 Relationships between the Big Five Factors, Stress Appraisals, and Coping ... 38 Limitations of Previous Research ..................................................................... 41
Summary ............................................................................................................... 43 CHAPTER 3: Study 1 – Does Dispositional Coping Measurement Overestimate the Relationship between Personality and Coping? ........................................................ 45
Aims and Hypotheses of Study 1 .......................................................................... 45 Method .................................................................................................................. 47
Participants ........................................................................................................ 47 Materials ............................................................................................................ 47 Procedure........................................................................................................... 49
Results ................................................................................................................... 50 Personality, Stress, and Stressor Appraisal ....................................................... 50 Correlations between Personality and Dispositional Coping ............................ 51 Correlations between Personality and Situational Coping ................................ 52 Comparing the size of correlations between Personality and Coping in the Dispositional and Situational Groups ............................................................... 52
Discussion ............................................................................................................. 54 Personality, Stress, and Stressor Appraisal ....................................................... 54 Personality and Coping ..................................................................................... 55 Neuroticism and Situational Coping ................................................................. 58
Conclusion ............................................................................................................ 61 CHAPTER 4: Social Ostracism and Williams’ Ball-Throwing Paradigm ............... 63
IX
The Negative Outcomes of Social Ostracism ....................................................... 64 Williams’ Need-Threat Model of Ostracism......................................................... 66 The Ball-Throwing Paradigm and Cyberball ........................................................ 69 The Ball-Throwing Paradigm and Coping Research ............................................ 72 Summary ............................................................................................................... 75
CHAPTER 5: Study 2 – Cyberball as an Experimental Stressor .............................. 76 Aims and Hypotheses of Study 2 .......................................................................... 76 Method .................................................................................................................. 78
Participants ........................................................................................................ 78 Materials ............................................................................................................ 78 Measures ........................................................................................................... 79 Procedure........................................................................................................... 82
Results ................................................................................................................... 83 Cyberball Appraisal .......................................................................................... 84 Mood and Self-Esteem ...................................................................................... 85 Coping ............................................................................................................... 86 Mediational Analyses ........................................................................................ 87
Discussion ............................................................................................................. 87 CHAPTER 6: Study 3 – Increasing the Range of the Response Scales on Outcome Measures ................................................................................................................... 91
Aims and Hypotheses of Study 3 .......................................................................... 91 Method .................................................................................................................. 93
Participants ........................................................................................................ 93 Measures ........................................................................................................... 93 Procedure........................................................................................................... 94
Results ................................................................................................................... 95 Cyberball Appraisal .......................................................................................... 95 Mood and Self-Esteem ...................................................................................... 96 Coping ............................................................................................................... 97 Mediational Analyses ........................................................................................ 99
Discussion ............................................................................................................. 99 CHAPTER 7: Study 4 – Introducing an Ambiguous Cyberball Condition ............ 104
Aims and Hypotheses of Study 4 ........................................................................ 104 Method ................................................................................................................ 106
Participants ...................................................................................................... 106 Materials .......................................................................................................... 107 Mood, Self-Esteem, Appraisal, and Coping.................................................... 109 Procedure......................................................................................................... 110
Results ................................................................................................................. 110 Cyberball Appraisal ........................................................................................ 111 Mood and Self-Esteem .................................................................................... 112 Coping ............................................................................................................. 114 Mediational Analyses ...................................................................................... 115
Discussion ........................................................................................................... 117 CHAPTER 8: Summary of Cyberball Studies ........................................................ 122
Aim 1: The Cyberball Game, Appraisals, Mood and Self-Esteem ..................... 123
X
Aim 2: Cyberball as a Laboratory-Stressor ......................................................... 125 Aim 3: Neuroticism, Appraisal, and Coping in the Context of Cyberball .......... 126 Conclusions from the Cyberball Studies ............................................................. 128
CHAPTER 9: Study 5 – Neuroticism and Coping in the Context of an Anagram-solving Task ............................................................................................................ 130
Aims and Hypotheses of Study 5 ........................................................................ 130 Method ................................................................................................................ 134
Participants ...................................................................................................... 134 Materials .......................................................................................................... 134 Other Measures ............................................................................................... 136 Procedure......................................................................................................... 137
Results ................................................................................................................. 137 Number of Anagrams Solved Correctly .......................................................... 138 Appraisal ......................................................................................................... 139 Mood and Self-Esteem .................................................................................... 140 Coping ............................................................................................................. 141 Mediational Analyses ...................................................................................... 143 Mild Stress Condition ..................................................................................... 145 High Stress Condition ..................................................................................... 146
Discussion ........................................................................................................... 148 The Anagram-Solving Task as an Experimental Stressor............................... 149 Neuroticism, Appraisal, and Coping in the Context of the Anagram-Solving Task ................................................................................................................. 149
Conclusion .......................................................................................................... 154 CHAPTER 10: Study 6 – Neuroticism and Emotion-Focused Coping Measures: Content Overlap and Item Cross-Loading? ............................................................ 155
Aims of Study 6 .................................................................................................. 155 Method ................................................................................................................ 157
Participants ...................................................................................................... 157 Procedure......................................................................................................... 158 Analysis ........................................................................................................... 158
Results ................................................................................................................. 160 Independent CFAs conducted in the High and Mild Stress samples .............. 160 Multi-group CFA and measurement invariance .............................................. 161
Discussion ........................................................................................................... 163 Conclusion .......................................................................................................... 165
CHAPTER 11: General Discussion ........................................................................ 167 Conclusion .......................................................................................................... 178
REFERENCES ........................................................................................................ 181 APPENDICES ........................................................................................................ 209 Appendix A: Exam Appraisal Scale ....................................................................... 210 Appendix B: COPE (Dispositional Version) .......................................................... 211 Appendix C: COPE (Situational Version) .............................................................. 214 Appendix D: Neuroticism Scale.............................................................................. 217 Appendix E: Cyberball Appraisal Scale ................................................................. 218 Appendix F: CITS (Situational Version) ................................................................ 219
XI
Appendix G: Reliabilities of Measures in Studies 3-5 ............................................ 221 Appendix H: Anagram Appraisal Scale .................................................................. 222 Appendix I: Publications ......................................................................................... 223
XII
List of Tables
Table 1. Correlations between neuroticism (N), extraversion (E), current stress level (in the dispositional condition) and stress appraisal (in the situational condition) obtained in Study 1 Table 2. Correlations (and comparisons of these correlations) between neuroticism (N), extraversion (E), and coping strategy use in the dispositional (DC) and situational (SC) condition obtained in Study 1 Table 3. Differences in appraisal (exclusion), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being included or excluded during the Cyberball game in Study 2 Table 4. Differences in appraisal (exclusion), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 2 Table 5: Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being included or excluded during the Cyberball game in Study 3 Table 6. Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 3 Table 7. Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being assigned to the excluded or ambiguous condition during the Cyberball game in Study 4 Table 8 . Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 4 Table 9. Partial correlations (controlling for gender) between neuroticism, appraisal (excluded), appraisal (control), task-focused (T-F), emotion-focused (E-F), and avoidance coping in the excluded Cyberball condition in Study 4 Table 10. Partial correlations (controlling for gender) between neuroticism, appraisal (excluded), appraisal (control), task-focused (T-F), emotion-focused (E-F) and avoidance coping in the ambiguous Cyberball condition in Study 4
XIII
Table 11. Anagrams and solutions (with median solve times taken from Tresselt & Mayzner, 1966) used in the mild and high stress conditions of Study 5 Table 12. Mean number of anagrams solved (and standard deviations) as a function of Condition and Neuroticism Level Table 13. Differences in appraisal (stress), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being allocated to the mild or high stress condition in Study 5 Table 14. Differences in appraisal (stress), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 5 Table 15. Partial correlations (controlling for gender) between neuroticism, stress appraisal, appraised controllability, emotion-focused coping, and avoidance coping in the mild stress condition Table 16. Partial correlations (controlling for gender) between neuroticism, stress appraisal, appraised controllability, emotion-focused coping, and avoidance coping in the high stress condition Table 17. Results of the hierarchical regression testing the mediating effect of stressor appraisals on the relationship between neuroticism and emotion-focused coping in the mild stress condition Table 18. Results of the hierarchical regression testing the mediating effect of stressor appraisals on the relationship between neuroticism and emotion-focused Coping in the high stress condition Table 19. Results of the hierarchical regression testing the mediating effect of stressor appraisals on the relationship between neuroticism and avoidance coping in the high stress condition Table 20. Model fit indices obtained in the high stress and mild stress samples
Table 21. Item loadings on the neuroticism (N) and emotion-focused coping (EFC) factors (and standardised covariance between N and EFC factors) in the high stress and mild stress samples Table 22. Fit indices (and change in χ2 and CFI) obtained in each step of the measurement invariance analysis
XIV
Table 23. Internal consistencies of the subscales of the MACL, the RSES, the subscales of the Cyberball Appraisal Scale, and the subscales of the CITS-S obtained in Studies 3, 4, and 5
List of Figures
Figure 1: Screen shot of the Cyberball game
Figure 2: Variance in Appraisal (Excluded) scores in piloted probability conditions
Figure 3: Mean Appraised Control scores (and Standard Error) for high and low neuroticism participants in the excluded and ambiguous Cyberball conditions
15
CHAPTER 1: A Brief Review of the Coping Literature
Coping is a widely used but poorly defined construct (Stone & Neale, 1984)
that generally refers to the things that people do to avoid being harmed by life-
strains and stresses (Pearlin & Schooler, 1978). However, there is little coherence in
theory, research, and understanding of the construct of coping (Folkman &
Moskowitz, 2004; Monat & Lazarus, 1991). Moreover, the concept of coping has
not been confined solely to the realm of scientific study; rather an examination of
magazine articles and pseudo-psychological self-help book lists reveals that the
subject of coping has also received much lay attention (Monat & Lazarus, 1991).
Indeed, Monat and Lazarus (1991) suggest that coping is as much a colloquial term
as a scientific one. This thesis, however, will treat coping as a psychological
construct and draw solely on the body of scientific research literature. The aims of
this review chapter are threefold. Firstly, a definition of coping as a scientific
construct will be provided and various conceptualisations of the structure of coping
will be outlined. Secondly, a rationale for why the study of coping is important will
be provided. Finally, the effectiveness of various coping strategies will be
discussed, and the notion of matching the appropriate coping strategy to the specific
stressful situation will be introduced.
16
Defining Coping
A large proportion of contemporary coping research can be traced back to the
publication of Lazarus’s (1966) book, Psychological stress and the coping process.
Prior to the publication of Lazarus’ 1966 seminal work on stress and coping, most
research on coping had taken place within the framework of ego-psychology and the
concept of defense (Folkman & Moskowitz, 2004). Indeed, as early as 1933 Freud
proposed the concept of the defense mechanism, a primarily unconscious response to
internal threat or conflict (Freud, 1933). This concept was extended by researchers
such as Haan (1969), Menninger (1963), and Vaillant (1977). Research within this
framework was generally concerned with the influence of unconscious processes on
pathology (Folkman & Moskowitz, 2004).
However, Lazarus (1966) expanded the boundaries of coping beyond defense
and pathology, and examined the wide range of cognitive and behavioural responses
that ‘ordinary’ people use to manage distress associated with the problems of daily
life (Folkman & Moskowitz, 2004). Lazarus’ (1966) theory emphasised the
influence of cognitive appraisals on an individual’s emotional responses to a
troubled person-environment relationship, and focused on how people coped with
this stressful relationship (Folkman & Moskowitz, 2004; Lazarus, 1966). Influenced
by the cognitive revolution, and more specifically Lazarus’ (1966) cognitively
oriented theory of stress and coping, during the 1970s and 1980s coping research
moved towards a model that conceptualised coping as a primarily conscious
response to external stressful or negative life-events (Lazarus, 1993).
17
The most widely cited definition of coping is that of Lazarus and Folkman
(1984) who define coping as “constantly changing cognitive and behavioural efforts
to manage specific external and/or internal demands that are appraised as taxing or
exceeding the resources of the person” (p. 141). This definition has become widely
accepted (Tennen, Afleck, Armeli, & Carney, 2000), and is the definition of coping
that will be used by this thesis. From this perspective coping is viewed as a goal-
directed process in which the individual orients thoughts and behaviours towards the
goals of resolving the source of the stress and managing emotional reactions to the
stress (Lazarus, 1993). Key to this definition is the notion of cognitive appraisal;
essentially an event can only be considered stressful if it is perceived to be stressful
by the individual. Cognitive appraisal is a process through which the person
evaluates whether a particular encounter with the environment is relevant to his or
her well-being, and if so, in what way (Folkman, Lazarus, Gruen, & DeLongis,
1986).
Two categories of cognitive appraisal have been proposed (Folkman,
Lazarus, Gruen et al., 1986; Lazarus & Folkman, 1984). Primary appraisal refers to
an evaluation of what is at stake in a given stressful encounter. For example is
general well-being, health, self-esteem or anything else at risk (Folkman, Lazarus,
Gruen et al., 1986)? Lazarus and Folkman (1984) identify three sub-types of
primary stress appraisals; appraisals of harm or loss (referring to harm or loss which
has already occurred), threat appraisals (referring to anticipated harm or loss), and
challenge appraisals (focussing on the potential for gain or growth that may be
associated with a given stressful situation). In contrast, secondary appraisal refers
18
to evaluations of what can be done to overcome or deal with the stressful situation
(Folkman, Lazarus, Gruen et al., 1986). During secondary appraisal the various
coping options available to the individual are evaluated. Characteristics of the
stressor (e.g. controllability, severity, predictability and chronicity) may also be
evaluated and these stressor characteristics can have important implications for the
choice of coping strategies (Folkman & Moskowitz, 2004).
Conceptualising the Structure of Coping
While, for the moment at least, there appears to be a general consensus on
how coping is best defined, there is still little consensus about how best to
conceptualise or measure the central constructs in the field; namely the ways of
coping with stress (Skinner, Edge, Altman, & Sherwood, 2003). More specifically,
there has been little consensus regarding the dimensions or categories into which
specific coping strategies can or should be grouped (Compas, Connor-Smith,
Saltzman, Harding Thomsen, & Wadsworth, 2001). According to Skinner and her
colleagues (2003) the fundamental problem in identifying core coping categories is
that coping “is not a specific behaviour that can be unequivocally observed or a
particular belief that can reliably reported. Rather, it is an organizational construct
used to encompass the myriad actions individuals use to deal with stressful
experiences” (p. 217). Furthermore, numerous attempts to organise the structure of
coping have been made. Indeed in their recent review of the coping literature
Skinner and her colleagues (2003) identified more than 100 coping categorisation
19
schemes, along with multiple scoring systems for different coping measures. This
inconsistency and variability makes comparing findings across studies a daunting
task (Skinner et al., 2003). Some of the most widely used conceptualisations of the
structure of coping are briefly summarised below.
The Problem-Focused/Emotion-Focused Coping Distinction
Lazarus and Folkman (1984) propose that coping has two major functions.
Firstly, coping may reduce the stress associated with a negative event by directly
acting on the stressor. This is termed problem-focused coping. Secondly, coping
may change either a) the way the stressor is attended to or b) the relational meaning
of what is occurring. This is termed emotion-focused coping (Lazarus, 1993).
Historically, the distinction between problem-focused and emotion-focused coping
has been an extremely influential conceptualisation of coping. Indeed, Parker and
Endler (1992) suggest that “if there is any consensus in the coping literature, it is
primarily about the distinction between emotion-focused and problem-focused
coping strategies” (p. 323). However, this perspective does have a number of
criticisms. Coyne and Gottlieb (1996) argue that the distinction between problem
and emotion-focused coping is too broad and that many disparate types of coping are
placed into these two general categories. Moreover, a single coping strategy may be
directed towards both problem and emotion-focused goals (Compas, Worsham, Ey,
& Howell, 1996). For example, walking away from a conflict with a peer may serve
the emotion-focused goal of calming oneself down, and the problem-focused goal of
taking time to generate alternate solutions to the problem (Compas et al., 1996).
20
The Primary/Secondary Control Distinction
The primary/secondary control distinction is another common
conceptualisation of coping (Band & Weisz, 1988; Rothbaum, Weisz, & Snyder,
1982; Rudolph, Dennig, & Weisz, 1995). Within this framework coping efforts are
directed at maintaining, augmenting, or altering control over the environment and
the self. From this perspective primary control coping is aimed at influencing
objective conditions or events or directly regulating one’s emotions (Compas et al.,
2001). Secondary control coping involves efforts to fit with, or adapt to, the
environment as it is. Band and Weisz (1988, 1990) note that secondary control
strategies involve subtle psychological means of reducing stress (e.g. lowering
expectations to minimise future disappointment) which may be abstract in nature.
Relinquished control is the absence of any goal-directed activity or coping (Band &
Weisz, 1988, 1990). The distinction between primary and secondary control coping
has been criticised for not categorising various forms of disengagement coping, such
avoidance, denial and wishful thinking (Compas et al., 2001).
Approach/Avoidance Distinction
The approach/avoidance framework has also been used frequently in the
coping literature (Compas et al., 2001; Fields & Prinz, 1997). This model is also
sometimes referred to as the Engagement/Disengagement Model (Compas et al.,
2001). From this perspective approach coping refers to a disposition to seek out
information, exhibit concern, and make plans when faced with a stressful situation
(Fields & Prinz, 1997). In contrast, avoidance coping refers to a disposition to avoid
21
information, exhibit little concern, and distract oneself in the face of stressful
situations (Fields & Prinz, 1997). Similar to the distinction between problem-
focused and emotion-focused coping, the distinction between approach and
avoidance coping has been criticised for being overly broad and failing to
distinguish between more distinct subtypes of coping (Compas et al., 2001).
Convergence of Classification Systems
Despite the apparent diversity among theoretical approaches to
conceptualising coping, some convergence has emerged (Compas & Boyer, 2001;
Compas et al., 2001). All these frameworks note a basic distinction between two
overarching groups of coping strategies (Fields & Prinz, 1997). The first group of
coping strategies is aimed at affecting the stressor more directly, and involves efforts
to change or master some aspect of the individual, environment, or relation between
them that is perceived stressful. This group of strategies includes problem-focused,
primary control, and approach coping strategies (Fields & Prinz, 1997). The second
group of coping strategies involves efforts to manage the negative emotions
associated with a stressful event and includes emotion-focused, secondary control,
and avoidant coping strategies (Fields & Prinz, 1997).
While some convergence between classification systems has been noted,
there is still criticism that these two overarching categories are too broad (Fields &
Prinz, 1997). Therefore, some researchers have combined categories from various
frameworks when constructing coping measures. Most notably, Carver, Scheier, and
Weintraub (1989) designed a multidimensional scale (the COPE), which can provide
22
a measure of diverse strategies such as venting of emotions, behavioural
disengagement, mental disengagement, and alcohol/drug use. The COPE is a widely
used instrument and is the measure employed in the first study reported in this thesis
(see Chapter 3). Additionally, Endler and Parker (1990b) designed the Coping
Inventory for Stressful Situations (CISS), which included a task-oriented coping
scale, an emotion-oriented coping scale, and an avoidance scale (that included items
tapping distraction and social diversion). An updated version of the CISS, the
Coping Inventory for Task Stressors (CITS; Matthews & Campbell, 1998) has been
designed specifically for use in task situations, and is the measure used in
experimental Studies 2 to 6 in this thesis (see Chapters 5, 6, 7, 9, and 10). Most
recently, there has been an emphasis on hierarchical multi-dimensional classification
systems of coping strategies (see Morling & Evered, 2006, 2007; Skinner et al.,
2003 for a discussion of this issue).
Why Study Coping?
The lack of clarity and consensus regarding the conceptualisation of coping
has resulted in numerous problems for researchers in the field of coping. These
include confusion over how best to measure coping, difficulties in comparing
findings across studies, and difficulties in documenting fundamental differences in
coping as a function of age, gender, personality, and other individual differences
(Compas et al., 2001). Given these difficulties and limitations, why do researchers
persist in studying this construct? Compas and his colleagues suggest that research
23
on the nature of coping processes is of both basic and applied importance (Compas
et al., 2001). From the perspective of basic research, coping represents an important
aspect of the more general processes of self-regulation of emotion, cognition,
behaviour, and physiology; and regulation of the environment (Compas et al., 2001).
Findings from research examining how individuals cope with stressful situations are
likely to provide valuable information on the nature of these self-regulatory
processes (Compas et al., 2001).
Coping research is also likely to have applied importance. Generally,
researchers agree that the study of coping is fundamental if we are to understand
how stress impacts on people for better or worse (Skinner et al., 2003). Coping
researchers argue that how people deal with stress can reduce or amplify the effects
of stressful situations on both short-term functioning and, in the long-term, the
development and maintenance of physical and mental health or disorder; although,
as yet this has not been unequivocally documented (Skinner et al., 2003).
Importantly, the ways in which individuals cope with stress are argued to be
potentially important mediators and moderators of the impact of stress on adjustment
and psychopathology (Compas et al., 2001). As the removal of all stress from daily
life is impossible, understanding how to effectively cope with life’s stressors and to
reduce the impact of these stressors on physical and mental health is of great
practical and applied importance (Compas et al., 2001). Given this, it is vitally
important that researchers identify what coping strategies are successful, and in what
situations they are or are not effective.
24
Effectiveness of Coping Strategies
Over the past 25 years a substantial body of research has examined the
impact of stress, and the ways of coping with it, on health and well-being. Despite
this sizeable body of literature, there is a lack of consistency across studies with
regard to which coping strategies are adaptive or maladaptive for dealing with
stressful events (DeLongis & Holtzman, 2005). Questions remain as to which
coping responses are most effective in the short and long term, in which contexts,
and for whom (Carver & Scheier, 1994). However, some general trends have
emerged1 (Matthews, Schwean, Campbell, Saklofske, & Mohamed, 2000; Zeidner &
Saklofske, 1996). Much of the research examining the effectiveness of coping
strategies has focused on the problem-focused versus emotion-focused coping
distinction.
Many research psychologists have emphasised the positive effects of
problem-focused coping and the negative effects of emotion-focused coping on
diverse psychological outcomes (Aldridge & Roesch, 2008; Matthews et al., 2000;
Moskowitz, Hult, Bussolari, & Acree, 2009; Roesch & Weiner, 2001). In
particular, this is argued to be the case when the stress experienced can be alleviated
by subjects’ responses (Lazarus & Folkman, 1984; Zeidner & Saklofske, 1996). A
number of empirical studies appear to offer support for this contention. Blalock and
Joiner (2000), Ebata and Moos (1991), as well as Sandler, Tein, and West (1994) all
report that problem-focused coping strategies, such as problem-solving, are
1 See Zeidner & Saklofske (1996) for an excellent review chapter on adaptive and maladaptive coping.
25
associated with increased feelings of efficacy and reduced levels of stress and
anxiety. In contrast, emotion-focused strategies (as well as avoidant strategies) are
reported to be associated with higher levels of anxiety and stress (Blalock & Joiner,
2000; Endler & Parker, 1990c; R. Gomez, 1998a; Matthews et al., 2000; Moskowitz
et al., 2009). Research examining mental health problems has also reported that
depressed individuals are more likely to employ emotion-focused strategies than
task-focused or problem-solving strategies (Zeidner, 1994). Additionally, loneliness
has been reported to be negatively associated with problem-focused coping and
positively associated with emotion-focused coping (Saklofske & Yackulic, 1989).
As well as being associated with depression, general anxiety, and loneliness,
coping responses are also reported to be predictors of post-disaster stress (Zeidner &
Saklofske, 1996). Studies examining US and Israeli soldiers (involved in combat)
have reported that severity of Post Traumatic Stress Disorder (PTSD) was positively
associated with emotion-focused coping and negatively associated with problem-
focused coping (Nezu & Carnevale, 1987; Solomon, Avitzur, & Mikulincer, 1989;
Solomon, Mikulincer, & Avitzur, 1988; Solomon, Mikulincer, & Benbenishty,
1989). Additionally, PTSD in soldiers can be predicted by the use of emotion-
focused strategies to cope with combat stress (Solomon, Mikulincer et al., 1989;
Zeidner & Saklofske, 1996). Research with children has revealed consistently that
problem-focused and approach/engagement coping are associated with fewer
internalising and externalising problems, and that disengagement (avoidant) coping
is associated with more internalising and externalising problems (Aldridge &
26
Roesch, 2008; Compas et al., 2001; Connor-Smith, Compas, Wadsworth, Harding
Thomsen, & Saltzman, 2000).
The research on the adaptiveness of avoidance coping is mixed (Zeidner &
Saklofske, 1996). There is a substantial body of literature that indicates that in
general avoidance coping is related to distress (Aldridge & Roesch, 2008; Billings &
Moos, 1981; Connor-Smith & Compas, 2002, 2004; Endler & Parker, 1990c;
Zeidner & Saklofske, 1996). Indeed, a literature review by Carver and Scheier
(1994) reports that avoidance coping (including strategies such as wishful thinking,
escapism, denial, self-distraction and mental disengagement) tend to work against
people rather than to their advantage. In particular, as mentioned previously, there
appears to be a relationship between increased reliance on avoidance coping and
depressive symptomatology (Aldridge & Roesch, 2008; Connor-Smith & Compas,
2002, 2004; Endler & Parker, 1990c; Zeidner, 1994; Zeidner & Saklofske, 1996).
However, there is also evidence suggesting that in some situations avoidance coping
can be adaptive. Specifically, cognitive avoidance may be an effective way to cope
with short-term stressors, such as noise, pain, and uncomfortable medical procedures
(Zeidner & Saklofske, 1996). Indeed, Carver, Scheier, and Pozo (1992) suggest that
avoidance coping may be useful at times because it gives the person a psychological
breather and an opportunity to escape from the constant pressures of the stressful
situation.
Overall, these results seem to suggest that problem-focused coping is more
adaptive than emotion-focused and avoidant coping, and this is indeed a widely held
view (Compas et al., 2001). Zeidner and Saklofske (1996) suggest that this may be
27
because problem-focused or active coping provides a sense of mastery over the
stressor. However, contradictory findings have been reported. For example, studies
have reported that denial (an emotion-focused/avoidant strategy often viewed as
maladaptive) can have favourable consequences for patients in post-coronary
hospital care (Levenson, Kay, Monteferrante, & Herman, 1984; Levine et al., 1987).
Pearlin and Schooler (1978) suggest that emotion-focused coping strategies are
increasingly used when the source of stress is unclear, when there is a lack of
knowledge about stress modification, or when the person can do little to eliminate
the stress. The issue of determining coping effectiveness remains one of the most
perplexing in coping research (Somerfield & McCrae, 2000). Recently, Folkman
and Moskowitz (2004) have argued that the adaptive qualities of coping processes
need to be evaluated in the specific stressful context in which they occur. They
suggest that a given coping response may be effective in one situation but not in
another, depending on situational characteristics such as controllability of the
stressor (Folkman & Moskowitz, 2004). This notion has been extended in the form
of the goodness of fit hypothesis.
The Goodness of Fit Hypothesis
The goodness of fit hypothesis highlights the importance of the match
between an individual’s coping effort and the specific situation. Essentially, from
this perspective it is argued that problem-focused coping should be more beneficial
in controllable situations; where there are more opportunities to actually change the
28
circumstances or have an impact on the stressful event. In contrast, emotion-focused
strategies should be more useful in less controllable situations, which by definition
allow less change of the circumstances of the stressful situation (e.g. Conway &
Terry, 1994; Park, Armeli, & Tennen, 2004). This fit between appraised
controllability and the use of problem/emotion-focused coping strategies is posited
to contribute to how well individuals adapt to stressors above and beyond the effects
of coping strategy use (Park et al., 2004; Park, Folkman, & Bostrom, 2001).
Essentially, the goodness of fit hypothesis proposes the use of situationally
appropriate coping strategies (Zeidner & Saklofske, 1996).
There is research evidence supporting the goodness of fit hypothesis
(Folkman & Moskowitz, 2004; Roesch & Weiner, 2001). For example, Christensen,
Benotch, Wiebe, and Lawton (1995) studied coping and adherence in a sample of
haemodialysis patients and found that problem-solving was associated with more
favourable adherence when the stressor involved a controllable aspect of the
haemodialysis context. For less controllable aspects of the treatment emotional self-
control was associated with more favourable adherence (Christensen et al., 1995).
Similarly, Park and colleagues (2001) found a match between problem-focused
coping and controllability in a sample of HIV+ men; however, the evidence for a
match between emotion-focused coping and uncontrollable stressors was less clear.
In contrast, Terry and Hynes (1998) examined women coping with an
uncontrollable stressor, in vitro fertilisation, and found that emotion-focused coping
was associated with better adjustment. Terry and Hynes (1998) also found,
consistent with the goodness of fit hypothesis, that direct efforts to manage the
29
problem were associated with poorer adjustment. Taken together, these findings
suggest that the controllability of the stressful situation can impact on the choice of
coping strategy (Roesch & Weiner, 2001); although more research is needed to
document this conclusively.
Research examining the goodness of fit hypothesis suggests that simply
trying to classify coping strategies as either effective or ineffective is problematic.
The effectiveness of coping strategies is likely to interact with stressor variables;
such as controllability, severity and chronicity (Park et al., 2004). Moreover, the
somewhat inconsistent findings of studies assessing the goodness of fit hypothesis
have prompted Park and her colleagues (2004) to suggest that the relationships
between stressor appraisals, coping strategy use, and outcomes may be even more
complex than previous models of appraisal-coping have proposed (e.g. Folkman,
1984, 1992). It is possible that person-related variables (e.g. personality traits,
previous experiences, and available resources) may also influence coping
strategy/stressor fit and adaptation or outcomes (Park et al., 2004). For example,
Connor-Smith and Flachsbart (2007) argue that personality traits may influence both
the choice of coping strategies and the effectiveness of coping strategies; with
strategies that are beneficial for some individuals being less effective, or even
harmful, for those with different personality traits. Moreover, personality variables
may also be associated with both the likelihood of experiencing stressful life-events
(e.g. Headey & Wearing, 1989; Magnus, Diener, Fujita, & Pavot, 1993) as well as
appraisals of these stressful events/situations (e.g. Engelhard, 2007; Mak, Blewitt, &
Heaven, 2004). This thesis aims to examine personality-related differences in
30
stressor appraisals and coping strategy use in situation specific stressors. In
particular, the role of neuroticism in the stress-coping process will be a specific
focus. Whilst acknowledging that research examining the effectiveness of coping
strategies (and any individual differences associated with coping effectiveness) is of
importance, this thesis will be specifically exploring the relationships between
personality and coping strategy use in the context of stressful situations. Chapter 2
reviews the body of literature exploring the relationships between personality-related
variables and both stressor appraisals and coping in more detail.
Summary
The scientific study of coping is argued to be important for both basic and
applied research and some consensus about how best to define coping appears to
have been reached; with Lazarus and Folkman’s (1984) definition becoming widely
accepted. However, conceptual confusion over how best to classify and measure
coping strategies has hampered progression in this research area and made making
comparisons across studies incredibly difficult. Bearing these limitations in mind,
two substantive conclusions can be reached. Firstly, although some inconsistent
findings have been reported regarding which coping strategies are (in)effective,
research has established a link between the construct of coping and psychological
health and well-being. In particular, research examining depression, anxiety, and
PTSD has revealed that coping is an important variable contributing to mental
health. Secondly, trying to classify coping strategies as either effective or
31
ineffective is overly simplistic. Choice of coping strategies (and the effectiveness of
coping strategies) is likely to interact with both stressor variables (such as
controllability, severity, and chronicity) and person-related variables (such as
personality, previous experiences, and available resources).
32
CHAPTER 2: Reviewing the Relationship between Personality and Coping
As mentioned in the previous chapter, the most widely accepted definition of
coping in the research literature is that of Lazarus and Folkman (1984), who define
coping as cognitive and behavioural efforts to manage specific demands that are
appraised as taxing or exceeding the resources of the person. One of the central
features of this perspective is that the stress-coping process unfolds in a dynamic
interplay between the person and the stressful situation (O'Brien & DeLongis, 1996).
Naturally, this has led to substantial interest in the role that personality-related
variables may have in the stress-coping process. Historically, the study of coping
has been approached from two distinct viewpoints; the inter-individual (style) and
intra-individual (process) perspectives (Folkman, Lazarus, Dunkel-Schetter,
DeLongis, & Gruen, 1986).
When studied from the inter-individual approach, it is argued that individuals
may be predisposed to consistently use certain types of coping strategies (Lazarus,
1993). From this perspective the coping scores of an individual are measured and
aggregated over a number of stressful situations in order to represent an index of the
individual’s coping style (Parker & Endler, 1992). The individual’s characteristic
coping style can then be compared with the coping style of others. This approach
allows for the assessment of individual differences in the use of coping strategies
(Parker & Endler, 1992). Traditionally, the study of relationships between
personality traits and coping has stemmed from this inter-individual approach to the
study of coping.
33
In contrast, the intra-individual approach to the study of coping attempts to
understand the process of coping; the impact that specific stressful situations have
on the use of coping strategies (Lazarus, 1993; Parker & Endler, 1992). This
approach measures the behaviours and cognitions of the same person across
different types of stressful situations in order to determine how the use of coping
strategies changes in response to particular types of stressors (Lazarus, 1993; Parker
& Endler, 1992). This perspective de-emphasises the role of stable traits in the
stress-coping process, and dominated coping research throughout the 1980s and
early 1990s2. More recently a viewpoint that proposes an interaction between
person-related variables and stressor-related variables has become widely accepted3
(Suls, David, & Harvey, 1996). This view, which proposes that both stressor and
person-related variables play an important role in the coping process, is the
perspective taken by this thesis.
Personality and Coping
Broadly speaking personality traits can be defined as “characteristic patterns
of thoughts, feelings, and behaviours over time and across situations” (Connor-
Smith & Flachsbart, 2007, p. 1080). It is therefore no surprise that personality traits
have been identified as likely candidates for researchers seeking predictors of inter-
2 See the work of Ptacek and colleagues (Ptacek, Pierce, & Elliott, 2003; Ptacek, Pierce, & Thompson, 2006; Ptacek, Smith, Espe, & Rafferty, 1994) for a discussion of the relationship between dispositional coping reports (style perspective) and situational coping reports (process perspective). 3 See Suls, David, and Harvey (1996) for an excellent review of three generations of research examining the relationship between personality and coping
34
individual differences in coping. Consistent with the inter-individual approach, this
research has revealed some evidence of consistency and stability in coping over time
and across stressful situations (Parkes, 1986; Terry, 1991, 1994). For example,
Parkes (1986) reported that stable traits (neuroticism and extraversion) were
significant predictors of coping scores over a number of different stressful episodes.
These findings were obtained in a sample of student nurses who were experiencing a
ward environment for the first time. Additionally, two longitudinal studies (Terry,
1991, 1994) reported that stable factors (including self-esteem and control beliefs)
influenced coping behaviour at all testing times. Furthermore, there is also evidence
that personality traits are significant predictors of coping strategy use in both the
short and long-term (Vollrath, Torgersen, & Alnaes, 1995). Indeed, some
researchers go so far as to say that coping can be described as “personality in action
under stress” (Bolger, 1990, p. 525) and that “coping ought to be redefined as a
personality process” (Vollrath, 2001, p. 341). These arguments have been supported
by correlations between various personality traits and coping subscales4 (Connor-
Smith & Flachsbart, 2007; Fickova, 2001; McWilliams, Cox, & Enns, 2003).
A large body of research has shown that personality-related variables are
associated with coping responses. For example, Strutton and Lumpkin (1992)
reported that in a sample of sales-people, participants who scored highly in
dispositional optimism were observed to rely on problem-focused coping strategies,
while participants who scored highly on pessimism were observed to rely on
emotion-focused coping. Research has also linked traits such as self-confidence,
4 See Connor-Smith and Flachsbart (2007) for a meta-analysis of relationships between personality and coping.
35
Type A Behaviour pattern, hostility, self-efficacy, and locus of control with coping
responses in meaningful ways (e.g. Fournier, de Ridder, & Bensing, 2002; A.
Gomez, 1997; R. Gomez, 1998a, 1998b; Hart & Hittner, 1995; Lee, Ashford, &
Jamieson, 1993; McCrae & Costa, 1986; Scheier, Weintraub, & Carver, 1986;
Strutton & Lumpkin, 1992). However, the magnitude and direction of correlations
have varied across studies and some studies have failed to demonstrate expected
relationships despite adequate statistical power and the use of reliable and valid
measures (Connor-Smith & Flachsbart, 2007).
This inconsistency suggests that the relationship between personality traits
and coping may be influenced by situation-specific variables such as stressor
severity, predictability and controllability (Connor-Smith & Flachsbart, 2007).
Importantly, this is consistent with a perspective which posits an interaction between
person and stressor-related variables in determining coping efforts and responses
(Suls et al., 1996). Additionally, the fact that investigators have tended to focus on
narrow, unidimensional personality traits (such as locus of control and optimism)
has resulted in difficulties summarising and generalising findings across studies
(Penley & Tomaka, 2002; Watson & Hubbard, 1996). More recently, research
emphasis has shifted from narrow, unidimensional traits to broader, more
comprehensive conceptualisations (Penley & Tomaka, 2002). Since the early 1990s
the five-factor model of personality (big five) has emerged as the predominant
model for specifying personality structure (McCrae & John, 1992). The five-factor
model provides a useful context and framework for systematically assessing
individual differences in coping (Watson & Hubbard, 1996).
36
The Five-Factor Model of Personality
The five factor model is a taxonomy of broad personality dimensions which
are posited to represent the minimum number of traits necessary for adequately
describing personality (David & Suls, 1999). Researchers have consistently reported
that five robust factors (neuroticism, extraversion, conscientiousness, agreeableness,
and openness to experience) are sufficient to represent the underlying structure of
personality (Borgatta, 1964; Costa & McCrae, 1985; John, 1990; Norman, 1963;
Tupes & Christal, 1992). Whilst a detailed discussion of the five factor model of
personality is beyond the scope of this review chapter, a brief description of each of
the five factors is warranted. Generally people who score high in neuroticism tend
to be more likely to experience negative emotions such as anxiety, depression, and
anger. Individuals who score high in extraversion tend to be energetic, cheerful, and
gregarious. People who score highly on openness to experience tend to be
untraditional, imaginative, and appreciative of aesthetic experiences. Agreeable
individuals are characterised as being helpful, trusting, and straightforward.
Individuals high in conscientiousness are characterised as reliable, hard-working,
and self-disciplined (Costa & McCrae, 1992; Goldberg, 1990; John, 1990).
These five factors have been obtained using both peer ratings (Norman,
1963; Tupes & Christal, 1992) and self-report (Costa & McCrae, 1988; McCrae &
Costa, 1985, 1987) methodologies. Moreover, closely parallel structures have been
identified in languages other than English, establishing some degree of cross-cultural
37
validity for the model (Barbaranelli, Caprara, Rabasca, & Pastorelli, 2003; John,
Goldberg, & Angleitner, 1984; Muris, Meesters, & Diederen, 2005).
Why Might General Personality Traits be Associated with Coping?
Why might general personality traits, such as the big five factors, be
expected to correlate with coping responses? Watson and Hubbard (1996) offer two
considerations. Firstly, there is no reason to believe that coping responses differ
fundamentally from other cognitive, affective, and behavioural responses. Coping
responses are distinguished from other adaptational responses only in that they are
inextricably linked to stress. Because individuals show some degree of consistency
in responses across different events and experiences (Diener & Larsen, 1984) they
should respond to stress in ways that relate systematically to the ways in which they
respond to other experiences (Watson & Hubbard, 1996). Secondly, dispositional
differences are likely to have important implications for the resources (and therefore
options for coping) that are available to the stressed individual (Watson & Hubbard,
1996). Furthermore, Connor-Smith and Flachsbart (2007) argue that personality
may affect coping strategy selection directly by constraining or facilitating the use of
specific coping strategies, or indirectly, by influencing the nature and severity of
stressors experienced and how these stressors are appraised.
In their recent review and meta-analysis of the personality and coping
literature, Connor-Smith and Flachsbart (2007) argue that direct effects of
personality on coping may begin early in childhood, with biologically based
appetitive, defensive, and attentional systems providing a framework in which
38
coping develops (Connor-Smith & Flachsbart, 2007). Additionally, Connor-Smith
and Flachsbart (2007) suggest that by influencing approach to rewards, withdrawal
from threats, and engagement or disengagement of attention, biological systems may
affect coping responses across the life-span (Connor-Smith & Flachsbart, 2007).
Personality may also influence coping responses indirectly. Connor-Smith and
Flachsbart (2007) argue that because coping responses are determined by stress
exposure, stress-reactivity, and the situational demands of the stressor, the influence
of personality on the frequency, intensity, nature and appraisal of the stressful
situations experienced may partially account for relationships between personality
and coping (Connor-Smith & Flachsbart, 2007).
Relationships between the Big Five Factors, Stress Appraisals, and Coping
Research has linked the big five factors with both appraisals of stressful
situations and coping with these situations. The role of neuroticism in the stress-
coping process has been a particular focus. Neuroticism is positively associated
with subjective measures of general stress level as well as the occurrence of stressful
life-events, even when these events are objectively defined (Headey & Wearing,
1989; Magnus et al., 1993). Moreover, neuroticism is also associated with negative
appraisals of the environment (e.g. Costa & McCrae, 1987; Engelhard, 2007;
Gallagher, 1990; Mak et al., 2004; Schneider, 2004). Indeed, individuals high in
neuroticism (and related constructs such as negative affectivity and trait anxiety) are
thought to interpret ambiguous stimuli in a negative or threatening manner, and
therefore are more likely to see threats or problems where others do not (Watson &
39
Clark, 1984). Specifically in relation to appraisal, findings suggest that individuals
high in neuroticism are more likely to appraise stressful situations as threats rather
than challenges (Costa & McCrae, 1985; Gallagher, 1990).
Research examining the use of coping strategies has revealed that individuals
who score highly in neuroticism tend to rely on passive, less effective forms of
emotion-focused and avoidance coping (Watson & Hubbard, 1996). Neuroticism is
typically positively associated with wishful thinking (Rim, 1987; Watson &
Hubbard, 1996), behavioural and mental disengagement (Carver et al., 1989), escape
avoidance, hostile reactions, emotional venting (McCrae & Costa, 1986; O'Brien &
DeLongis, 1996; Watson & Hubbard, 1996) and general emotion-focused coping
(Endler & Parker, 1990b; Matthews & Campbell, 1998; Matthews et al., 2006).
Neuroticism is also negatively associated with problem-focused coping (Endler &
Parker, 1990b; Matthews & Campbell, 1998; Matthews et al., 2006) and direct
coping (Parkes, 1986). These findings are broadly consistent with a personality type
categorized as prone to experience negative emotions (Watson & Hubbard, 1996).
Additionally, according to the goodness of fit hypothesis (see Chapter 1) when
stressful situations are appraised as uncontrollable individuals should engage in
more emotion-focused and avoidance coping (Conway & Terry, 1994; Park et al.,
2004). Given that neuroticism is associated with negative appraisals of stressful
situations as well as the use of passive, avoidant, and emotion-focused coping
strategies (Watson & Hubbard, 1996), it is possible that stressor appraisals may
mediate the relationship between neuroticism and coping. A secondary aim of this
thesis is to explore this possibility.
40
Less research has examined the role of extraversion in the stress-coping
process; however, some replicable findings have emerged. Individuals high in
extraversion are reported to appraise stressful situations as challenges rather than
threats (Costa & McCrae, 1985; Gallagher, 1990). In relation to coping, individuals
scoring high in extraversion appear to be more likely to engage in problem-focused
coping and seek social support. High extraversion individuals are more likely to
take action (McCrae & Costa, 1986; Parkes, 1986), engage in positive thinking
(McCrae & Costa, 1986; Rim, 1987), seek social support (Amirkhan, Risinger, &
Swickert, 1995; Watson & Hubbard, 1996), and generally employ more problem-
focused coping strategies (Matthews & Campbell, 1998; Matthews et al., 2006;
McCrae & Costa, 1986; O'Brien & DeLongis, 1996; Watson & Hubbard, 1996).
Extraversion is negatively associated with emotion-focused coping and avoidance
strategies (Amirkhan et al., 1995; Matthews & Campbell, 1998; Matthews et al.,
2006; O'Brien & DeLongis, 1996). These findings are broadly consistent with a
personality type categorized as cheerful and gregarious (Watson & Hubbard, 1996).
In comparison to research focusing on neuroticism and extraversion, little
research has examined the relationships between the other dimensions of the five-
factor model and coping strategy use; however, some patterns are emerging.
Openness to experience has been negatively related to religious coping (McCrae &
Costa, 1986; Watson & Hubbard, 1996). Agreeableness has been found to be
positively related to support seeking (O'Brien & DeLongis, 1996), positive
reappraisal, active coping, and planning (Watson & Hubbard, 1996). People scoring
high in conscientiousness are reported to use more problem-focused strategies;
41
including active coping, planning, suppression of competing activities, and restraint
coping (Watson & Hubbard, 1996). Conscientiousness is also negatively related to
emotion-focused coping; particularly alcohol and drug use, mental and behavioural
disengagement, denial (Watson & Hubbard, 1996), escape-avoidance, and accepting
responsibility (O'Brien & DeLongis, 1996). These findings appear broadly
consistent with personality traits characterized respectively as imaginative, helpful
and trusting, and self-disciplined; however, more replication is needed before any
firm conclusions can be drawn.
Limitations of Previous Research
Although consistent relationships between personality and coping have
emerged in previous research (Connor-Smith & Flachsbart, 2007), these findings
have several important limitations. Typically researchers have utilized one of two
methods to measure coping: 1) participants are asked how they usually or generally
cope with stressful situations; or 2) participants are asked to recall coping efforts
used in the past week, month, year, or more; (David & Suls, 1999). Indeed, in a
recent meta-analysis, Connor-Smith and Flachsbart (2007) reported that of 124
research papers published between 1980 and 2004 examining relationships between
personality and coping 103 of these papers (83.06%) utilised either a dispositional or
retrospective recall methodology.
Whilst an understanding of the relationship between personality and general
coping style may be of interest, it has been suggested that a using dispositional
and/or retrospective coping measures may overestimate the relationship between
42
personality and coping strategy use in the context of any given stressful situation.
Firstly, David and Suls (1999) suggest that measures which tap how one ‘usually’
copes are likely to reflect dispositions to a greater extent than situation specific
measures. Indeed, by asking how one usually copes with stress, researchers are
likely to be priming a trait-like response which may be systematically related to
personality. Secondly, previous research has revealed that the more time that
elapses between an event and its assessment, the more likely individuals will be
biased towards giving dispositional reports of their behaviour (Moore, Sherrod, Liv,
& Underwood, 1979). Indeed, Ross (1989) argues that individuals reconstruct
previous experiences by organizing available memory fragments around an implicit
theory of what they believe most likely occurred. Again, it is likely that when recall
periods are long, any retrospective contamination that occurs may be systematically
related to measures of personality (David & Suls, 1999). Another problem with
using retrospective reports is that they may be coloured by the outcome of the event.
Specifically, whether the individual experienced an (un)successful outcome is likely
to influence their memory of the event (Coyne & Gottlieb, 1996). Finally, when
using a retrospective methodology different participants may recall different
stressful events; this makes it extremely difficult (if not impossible) to assess
individual differences in coping, as different events may evoke different coping
strategies. Importantly, personal characteristics, history, and circumstances may
determine which incidents people report as being the most stressful, as well as how
these incidents unfolded and what options were available for coping with them
(Coyne & Gottlieb, 1996).
43
Due to these limitations it has been suggested that examining the relationship
between personality and coping with a specific stressful events, where coping can be
measured either before, during, or immediately after the stressful situation, would
allow for much firmer conclusions regarding the relationships between personality
and coping to be made (Bolger, 1990; Stone, Kessler, & Haythornthwaite, 1991). In
a series of studies examining task-induced stress (using a number of experimental
tasks including rapid information processing, mental arithmetic, working memory
tasks, and driving simulation) and individual differences in coping, Matthews and
his colleagues (Matthews & Campbell, 1998; Matthews et al., 2006) found that
neuroticism was significantly correlated with emotion-focused and avoidant coping
strategies and extraversion was significantly correlated with task-focused coping
strategies. These data are consistent with previous findings using dispositional and
retrospective coping measures and appear to be a promising start in elucidating the
relationship between personality and coping with situation specific stressors.
However, despite the compelling arguments that dispositional/retrospective coping
measurement may overestimate the relationship between personality variables and
coping strategy use in the context of situational stressors, the literature still lacks
empirical tests of this notion. Study 1 addressed this issue.
Summary
Investigators taking an interactionist perspective towards coping research
argue that both stressor variables and person-related variables impact on coping
44
strategy choice. Early research examining the link between person-related variables
focused on a variety of individual differences (including optimism/pessimism, locus
of control, and type A behaviour pattern). More recently the five-factor model of
personality has provided a cohesive framework for examining the relationship
between personality and coping; which should allow for greater ease in documenting
individual differences in coping across studies. A substantial body of research has
revealed that the big five personality factors are systematically related to coping
strategy use (although the majority of this research has focused on neuroticism and
to a lesser extent extraversion). However, previous research is limited in its reliance
on dispositional and retrospective coping measures. Study 1 aimed to 1) empirically
test the notion that dispositional measures of coping may overestimate the
relationship between personality-related variables and coping in the context of a
specific stressor, and 2) examine the relationship between neuroticism, extraversion,
and coping in the context of a situation-specific stressor; approaching university
examinations.
45
CHAPTER 3: Study 1 – Does Dispositional Coping Measurement Overestimate
the Relationship between Personality and Coping?
Aims and Hypotheses of Study 1
Study 1 utilized both a dispositional and situational coping measure to
empirically test the argument that the relationships obtained between personality
dimensions and coping strategy use may be overestimated when coping is measured
using a dispositional measure (David & Suls, 1999). Additionally, Study 1 aimed to
explore the relationship between personality-traits and coping in the context of a
situation-specific stressor. As mentioned in Chapter 2, neuroticism and extraversion
are the personality factors which have received the most attention in the research
literature. Both neuroticism and extraversion have been consistently associated with
coping responses (McCrae & Costa, 1986; Watson & Hubbard, 1996). Very little
research has examined the relationships between the other dimensions of the five-
factor model and coping strategy use (Watson & Hubbard, 1996). Given that the
aim of Study 1 was to explore the possibility that dispositional coping measurement
may inflate the relationship between personality and coping, Study 1 focused
specifically on neuroticism and extraversion, as these are the factors that are
consistently associated with self-reported coping.
Individuals were assigned to either a dispositional condition (completed a
dispositional version of the coping measure, which asked how do you usually cope
with stressful situations), or a situational condition (completed the coping measure
46
specifically in relation to upcoming university examinations). Examinations were
chosen as a situation-specific stressor because, due to their anticipatory nature,
coping can be measured concurrently with the stressor. It was anticipated that this
would minimise the problems associated with retrospective reporting.
Based on previous research the following hypotheses were proposed. Firstly,
it was hypothesized that neuroticism would be significantly related to general stress
level in the dispositional condition, and with negative stress and controllability
appraisals in the situational condition. Extraversion was predicted to be negatively
related to stress level in the dispositional condition and associated with more
positive appraisals of the examinations in the situational condition. Secondly, it was
predicted that when coping was measured using the dispositional measure the
correlations obtained between neuroticism, extraversion and the various coping
subscales would be consistent with previous research (see Chapter 2 for a summary
of previous findings). Thirdly, it was predicted that if dispositional measures do
overestimate the relationships between personality traits and coping then the
correlations obtained between neuroticism, extraversion, and coping subscales in the
context of the approaching examinations should be smaller than the correlations
obtained using the dispositional coping measure.
47
Method
Participants
One hundred and fifty-six psychology students (42 male; 114 female) aged
between 17 and 53 years participated in Study 1 (mean age = 22.58 years). First
year students gained academic credit in exchange for participating in the study.
Participants were allocated to either the dispositional condition (n = 79) or the
situational condition (n = 77).
Materials
Personality
The NEO-FFI (Costa & McCrae, 1992) is a short-form version of the revised
NEO Personality Inventory (NEO-PI-R). It contains 60 items with 12 items
assessing each of the Big Five personality factors. Items are rated on a five point
scale (0: Strongly disagree; 4: Strongly agree). The NEO-FFI scales are highly
correlated with the corresponding scales on the full NEO-PI-R; convergent
correlations range from .77 to .94 across samples (Watson & Hubbard, 1996). Costa
and McCrae (1992) have reported internal consistencies of the five subscales to
range between .68 (agreeableness) and .86 (neuroticism).
Stress and Stressor appraisal
The stress subscale of the Depression Anxiety Stress Scales 21 (DASS 21)
(Lovibond & Lovibond, 1995) was used to assess the current stress level of
48
participants in the dispositional condition. The DASS 21 is a 21 item scale made up
of three seven item subscales measuring severity of depression, anxiety, and stress
symptoms experienced over the past week. Responses are made on a four-point
scale (0: Not at all; 3: A lot or most of the time). The stress items of the DASS 21
are categorized by tension or stress and characterized by non-specific arousal
(Lovibond & Lovibond, 1995). Studies using normal samples report the stress
subscale to have an internal consistency ranging between .90 and .93 (Crawford &
Henry, 2003; Lovibond & Lovibond, 1995).
In order to assess the stressor appraisals of the situational condition an eight
item Exam Appraisal Scale was constructed (see Appendix A). Responses were
made on a four-point scale (1: Not at all; 4: Extremely). The Exam Appraisal Scale
consists of two sub-scales; five items measuring how stressful and important the
examinations were perceived to be (e.g. ‘to what extent do you find the upcoming
event stressful?’ α = .83), and three items measuring controllability and
predictability of the outcome (e.g. ‘to what extent do you feel you have control over
the outcome?’ α = .72). A principal components analysis (using oblique rotation)
revealed that the two predicted components emerged clearly, and accounted for
61.13% of the overall variance.
Coping
The COPE (Carver et al., 1989) is a 60 item scale used to assess the use of a
variety of coping strategies. It contains 15 sub-scales (active coping, planning,
seeking instrumental social support, seeking emotional social support, suppression
49
of competing activities, turning to religion, positive reinterpretation and growth,
restraint coping, acceptance, focusing on and venting emotions, denial, behavioural
disengagement, mental disengagement, alcohol and drug use, and humour). All the
subscales of the COPE consist of four items.
Responses are made on a four-point scale (1: I usually don’t do this at all; 4:
I usually do this a lot). Subscale scores are calculated by summing responses to the
four items that make up each subscale. Internal consistencies of the individual
subscales are acceptably high (considering the scales consist of only four items)
ranging between .45 and .92, with a median value of 0.71 (Carver et al., 1989). Both
a dispositional and a situational version of the COPE were used. Individuals in the
dispositional condition completed the dispositional version of the COPE which asks
people to describe what they typically do under stress (see Appendix B). Individuals
in the situational condition completed a situational version of the COPE that had
been altered to ask participants how they were coping with a specific stressful event;
the upcoming examinations (see Appendix C).
Procedure
Participants in the dispositional condition were recruited during the middle of
semester in order to prevent the possibility of examination stress confounding their
responses to the coping measure. Participants in the situational condition were
recruited two weeks prior to the commencement of end of year examinations. After
reading an information sheet and signing a consent form, participants assigned to the
dispositional condition completed the NEO-FFI, the DASS 21, and the dispositional
50
version of the COPE. Participants assigned to the situational condition completed
the NEO-FFI, the Exam Appraisal Scale, and the situational version of the COPE.
Results
In this thesis all numerical values will be rounded to two decimal places,
with the exception of p values, which will be rounded to three decimal places. The
dispositional and situational conditions did not differ significantly in neuroticism;
F(1, 153) = .24, p = .627; or extraversion; F(1, 154) = .18, p = .671. Critical α level
was set at .05 for all analyses and sample sizes in each condition varied between 75
and 79 due to missing data.
Personality, Stress, and Stressor Appraisal
Firstly, correlations were calculated between neuroticism, extraversion and
the DASS stress subscale for the dispositional condition; and neuroticism,
extraversion and the two exam appraisal subscales for the situational condition.
These results are summarized in Table 1. In the dispositional condition neuroticism
was positively correlated with current stress level. Extraversion was negatively
correlated with current stress level. In the situational condition neuroticism was
positively correlated with appraisals of how stressful and important the exams were
perceived to be and negatively with appraisals of controllability and predictability of
the outcome. Extraversion was not significantly correlated with appraisals of the
approaching exams.
51
Table 1. Correlations between neuroticism (N), extraversion (E), current stress level (in the dispositional condition) and stress appraisal (in the situational condition) obtained in Study 1
N E
DASS 21 stress
(Dispositional Condition)
.58*
-.24*
Stress Appraisal
(Situational Condition)
.30*
-.11
Outcome Predictability
(Situational Condition)
-.31* -.08
* p < .05
Correlations between Personality and Dispositional Coping
Secondly, correlations between neuroticism, extraversion and coping scores
obtained by individuals who completed the dispositional version of the COPE were
calculated. These correlations are summarized in Table 2. Significant negative
correlations were obtained between neuroticism and active coping, turning to
religion, positive reinterpretation, and mental disengagement. Neuroticism was
positively correlated with focus on/venting emotions and behavioural
disengagement. Extraversion was positively correlated with active coping, seeking
instrumental social support, seeking emotional social support, positive
reinterpretation, and humour. Extraversion was negatively correlated with
behavioural disengagement.
52
Correlations between Personality and Situational Coping
Thirdly, correlations obtained between neuroticism, extraversion, and the
coping scores obtained by individuals who completed the situational version of the
COPE (in the context of the examination stressor) were calculated. These
correlations are also summarized in Table 2. Neuroticism was positively correlated
with seeking instrumental social support, focussing on/venting emotions, denial, and
behavioural disengagement. Extraversion was negatively correlated with planning.
Comparing the size of correlations between Personality and Coping in the
Dispositional and Situational Groups
The correlations obtained between neuroticism, extraversion, and coping in
the two conditions were compared using the methodology suggested by Cohen and
Cohen (1983). The p values associated with differences (Diff) between the size of
the correlations obtained in the dispositional condition and the situational condition
are summarised in Table 2. With regard to extraversion, all of the six significant
correlations obtained in the dispositional group were reduced to non-significance in
the situational group. For three of the six significant correlations, the difference
between the values obtained in the dispositional and situational groups were
statistically significant (see Table 2). With regard to neuroticism, four of the six
significant correlations obtained in the dispositional group were reduced to non-
significance in the situational group. For three of the six significant correlations, the
difference between the size of the correlations obtained in the dispositional and
situational groups were statistically significant (see Table 2).
53
Table 2. Correlations (and comparisons of these correlations) between neuroticism (N), extraversion (E), and coping strategy use in the dispositional (DC) and situational (SC) conditions obtained in Study 1 N p E p
DC SC Diff DC SC Diff
Active Coping
-.28*
-.05
.073
.30*
-.07
.010*
Planning -.14 .16 .032* .19 -.26* .002*
Instrumental Social Support -.14 .24* .009* .26* .06 .103
Emotional Social Support -.12 .08 .109 .28* .20 .301
Suppression .01 .09 .312 .04 -.17 .097
Religion -.28* .21 .001* .13 .08 .378
Positive Reinterpretation -.41* .01 .003* .35* -.10 .002*
Restraint -.08 .15 .078 .01 -.07 .312
Acceptance .09 -.11 .110 .14 .06 .310
Focus On/Vent Emotions .33* .49* .118 .02 -.20 .086
Denial -.13 .36* < .001* -.09 -.02 .333
Mental Disengagement -.27* .16 .004* -.17 -.01 .161
Behavioural Disengagement .34* .39* .362 -.28* .06 .017*
Alcohol/Drug Use .17 .05 .228 .06 .18 .228
Humour -.15 .07 .088 .26* .21 .373
* Significant at p < .05
54
Discussion
Study 1 aimed to empirically test the argument that the relationships obtained
between personality characteristics and coping strategy use may be overestimated
when coping is assessed using dispositional measures (David & Suls, 1999). Based
on previous research the following hypotheses were proposed. Firstly, it was
hypothesized that neuroticism would be significantly related to general stress level
in the dispositional condition, and to negative appraisals of the examinations in the
situational condition. Extraversion was predicted to be negatively related to current
stress level in the dispositional condition, and to be associated with more positive
appraisals of the approaching exams in the situational condition. Secondly, it was
predicted that when coping was measured dispositionally the correlations obtained
between the personality variables and coping would be consistent with the
relationships obtained between neuroticism, extraversion, and coping strategy use
previously reported in the literature. Thirdly, if dispositional coping measures do
overestimate the relationships between personality and coping strategy use then,
when coping was measured concurrently (using the examinations as a specific
stressor), it was predicted that these correlations would be reduced.
Personality, Stress, and Stressor Appraisal
As predicted, in the dispositional condition neuroticism was positively
correlated with current stress level and extraversion was negatively correlated with
current stress level. These findings are consistent with previous research (Costa &
55
McCrae, 1987). However, neuroticism was the only personality characteristic
significantly correlated with appraisals of the approaching exams. Neuroticism was
correlated positively with appraisals of how stressful and important the exams were
perceived to be, and negatively with appraisals of controllability and predictability
of the outcome. These findings are consistent with previous studies which have
shown substantial correlations between neuroticism and subjective measures of
perceived stress and distress in the context of stressful situations5 (e.g. Costa &
McCrae, 1990; Engelhard, 2007; Gallagher, 1990; Mak et al., 2004; Watson &
Clark, 1984). This consistent link between neuroticism and both general stress level,
as well as negative reactions to specific events or situations, has prompted
researchers to conclude that neuroticism is likely to play an especially important role
in the stress-coping process (Watson & Hubbard, 1996). It has been argued that
individuals who score highly in neuroticism tend to make negative appraisals of the
environment and interpret ambiguous stimuli in a negative or threatening manner.
Watson and Hubbard (1996) therefore suggest that individuals high in neuroticism
are likely to see threats and problems where others do not. The findings obtained in
the situational condition, appear to be consistent with this theory.
Personality and Coping
In the dispositional condition significant negative correlations were obtained
between neuroticism, active coping and positive reinterpretation. These findings are
5 It should be noted that a limitation of this study was that the stress measures used in the situational and dispositional groups were not truly parallel. The situational appraisal scale assesses the stressfulness and controllability of an external event (examinations). In contrast, the DASS items (e.g. I can’t wind down; I can’t relax) may be a reflection of emotional distress or neuroticism.
56
broadly consistent with previous research which has reported a negative relationship
between neuroticism and problem-focused coping (Endler & Parker, 1990b), direct
coping (Parkes, 1986), and the use of more adaptive emotion-focused strategies (e.g.
positive reinterpretation). Neuroticism was positively correlated with focus
on/venting emotions and behavioural disengagement. These findings are broadly
consistent with previous research reporting that neuroticism is associated with more
passive and maladaptive emotion-focused and avoidance coping strategies (e.g.
Carver et al., 1989; Endler & Parker, 1990b; McCrae & Costa, 1986; O'Brien &
DeLongis, 1996; Watson & Hubbard, 1996). However, unlike much previous
research, in the dispositional condition neuroticism was not significantly correlated
with behavioural disengagement or denial (Carver et al., 1989). In the dispositional
condition extraversion was significantly correlated with active coping, seeking
instrumental and emotional social support, positive reinterpretation, and humour.
These findings are broadly consistent with a personality type conceptualized as
cheerful and gregarious, and replicate previous findings suggesting extraversion is
positively associated with problem-focused and active coping (McCrae & Costa,
1986; Parkes, 1986), seeking social support (Amirkhan et al., 1995; Watson &
Hubbard, 1996), and engaging in positive thinking (McCrae & Costa, 1986; Rim,
1987). In the dispositional condition, extraversion was also negatively correlated
with behavioural disengagement. This is consistent with previous research
suggesting that extraversion is negatively associated with avoidance coping
(Matthews & Campbell, 1998; Matthews et al., 2006; O'Brien & DeLongis, 1996).
57
Overall the relationships obtained between neuroticism, extraversion, and
dispositional coping in Study 1 appear consistent with previous findings obtained
using dispositional and retrospective coping measures. The results obtained in Study
1 therefore offer support for the notion that, at least when measured dispositionally,
coping is systematically related to neuroticism and extraversion. However, before
any firm conclusions can be made regarding the relationships between personality
variables and coping strategy use, it is imperative to ensure that dispositional coping
measurement is not overestimating the relationship between personality
characteristics and coping strategy use (David & Suls, 1999). In Study 1, when
coping was measured in the context of the examinations, only neuroticism was
found to be related to the self-reported use of multiple coping strategies. However;
it is notable that in the situational condition neuroticism was positively correlated
with passive and maladaptive strategies such as focus on/venting emotions, denial
and behavioural disengagement. None of the negative correlations between
neuroticism and problem-focused/active coping strategies, reported in previous
research (and in the dispositional group), were obtained when the approaching
exams were used as a situation-specific stressor. Situational reports may therefore
be particularly important in capturing maladaptive coping behaviours.
In general, the significant correlations obtained between extraversion and the
various coping subscales in the dispositional condition were reduced to non-
significance when the examinations were used as a situation-specific stressor.
Importantly, for three of these subscales (active coping, positive reinterpretation, and
behavioural disengagement) the differences between the correlations obtained in the
58
dispositional and situation groups were statistically significant. Taken together,
these results offer some support for David and Suls’ (1999) contention that the
associations between personality traits and coping strategy use may be overestimated
when coping is assessed using dispositional measures; however, these findings
should be replicated in the context of other situation-specific stressors and the use of
a within-subjects design would allow firmer conclusions to be drawn6. However, the
current results suggest that any relationships between personality variables and
coping obtained with dispositional coping measures need to be interpreted with
caution (specifically if researchers are interested in predicting coping behaviour in
the context of a given stressful situation).
Neuroticism and Situational Coping
Why might neuroticism be associated with situational coping when
extraversion appears not to be? There are numerous reasons to believe that
neuroticism plays a particularly important role in the stress-coping process. Indeed,
Watson and Hubbard (1996) argue that, although previous research has
demonstrated significant links between neuroticism and virtually every stage in the
stress-coping-disorder process, the available data on the other big five factors are
much more sparse. Prospective studies have revealed that neuroticism predicts the
subsequent occurrence of real stressful life-events, even when these events are
objectively defined (Headey & Wearing, 1989; Magnus et al., 1993). This has led
6 It should however be noted that a narrowly focused situational stressor may equally lead to an underestimation of the relationship between personality and coping; either because the personality trait is not relevant to the specific stressor or because the situation offers only a narrow range of coping options and limited scope for the influence of personality.
59
Watson and Hubbard (1996) to suggest that, to some extent, individuals who score
high in neuroticism create problems for themselves. Given the fact that neuroticism
appears to be associated with negative life-events, such as divorce or being fired
(Ormel & Wohlfarth, 1991), it is likely that neuroticism might be associated with
coping during the evolution of stressful life-events.
Additionally, the current study, as well as previous research, has reported
that neuroticism is linked to subjective stress level and, more importantly, negative
appraisals of stressful situations (Watson & Hubbard, 1996). In the current study,
neuroticism was correlated positively with appraisals of how stressful and important
the exams were perceived to be and negatively with appraisals of controllability and
predictability of the outcome. Importantly, in the transactional model of stress and
coping, cognitive appraisals are a vital link between the stressful situation and
coping. Indeed, a situation can only be considered stressful if it is appraised by the
individual as being stressful (Lazarus & Folkman, 1984). It may be this relationship
between neuroticism and situational appraisal that underlies the neuroticism-
situational coping relationship. Of particular importance is the negative relationship
between neuroticism and appraisals of controllability. According to the goodness of
fit hypothesis (Park et al., 2004; Park et al., 2001) individuals attempt to match their
choice of coping strategy to the given stressful situation. In essence, if a stressor is
appraised as controllable then individuals should attempt to directly act upon, or
approach the stressor, by using problem-focused coping strategies. If a stressor is
appraised as uncontrollable then individuals should attempt to change their
emotional response to the situation, by using emotion-focused or avoidant coping
60
strategies, such as positive reinterpretation, mental and behavioural disengagement
(Park et al., 2004; Park et al., 2001). This raises the possibility that stressor
appraisals may mediate the relationship between neuroticism and coping strategy
use. Studies 2-5 (Chapters 5, 6, and 7) examine this idea in more detail.
A final methodological consideration is that the examination situation was
not a time-constrained or task-specific stressor. Participants were asked during the
two weeks before the examinations commenced how they were coping with stress
associated with the approaching examinations. By recalling coping efforts over the
preceding two weeks it is possible that participants were providing responses to the
coping questions that were still to some degree dispositional. It is also important to
note that, because the examinations were not a novel situation, participants may also
have drawn on previous experiences of how they typically or previously coped with
examinations. This may have resulted in retrospective recall biases contaminating
responses to the situational coping measure7.
Additionally, the examination period is not equivalent for all students. The
number of exams that students write, and individual exam timetables can vary
substantially. It is likely that these unmeasured variables may have an important
impact on both appraisals of the exams, as well as coping in the context of university
exams. In order to clarify the relationship between neuroticism and situational
coping future research should attempt to replicate these findings using novel task-
7 It should also be noted that if individuals aggregate memories of how they coped with different stressful situations then two possible factors influencing dispositional vs. situational coping reports may include 1) how representative coping with the specific situation (examination stress) is of coping in general, and 2) how salient memories of coping with exam stress are in the individual’s judgement of dispositional coping. Future research should explore this issue in more detail.
61
specific stressors in which coping could be assessed immediately after participation.
Laboratory-based stressors are likely to be extremely useful in this context. This is
for two major reasons. Firstly, laboratory-stressors overcome the reliance on
dispositional and retrospective coping measures that have limited much of the
previous research (David & Suls, 1999). Secondly, presenting the same objective
stressor to all participants minimizes contextual confounds and allows firmer
conclusions regarding individual differences in the coping process to be made
(Connor-Smith & Flachsbart, 2007). However, generic coping measures (such as
the COPE) are inappropriate for use in highly constrained laboratory tasks, as many
of the coping strategies measured are simply not possible under the constraints of
experimental tasks (for example engaging in alcohol or drug use). The Coping
Inventory for Task Stressors (Situational Version, CITS-S; Matthews & Campbell,
1998) has been designed specifically for use in task situations and is the measure
that will be used for the remainder of this thesis.
Conclusion
Study 1 replicated previous research which reports that neuroticism and
extraversion are systematically related to self-reported coping strategy use when
coping is measured dispositionally. However, only neuroticism was found to be
systematically related to coping in the context of a situation-specific stressor; the
approaching university examinations. Based on these results, as well as the fact that
neuroticism is the most widely studied personality-trait within the coping literature
62
(Watson & Hubbard, 1996), neuroticism will become the focus of the remainder of
this thesis. The findings of Study 1 offer some empirical support for the notion that
(with the possible exception of neuroticism) dispositional coping measures may
inflate the relationship between personality and coping with situation-specific
stressors. Neuroticism was associated with maladaptive emotion-focused and
avoidant coping strategies in situational condition, and this is consistent with
previous findings obtained using dispositional and retrospective coping measures.
However, future research should examine the relationship between neuroticism and
coping using a novel, time-constrained stressor, in order to ensure that situational
coping reports are not being contaminated by retrospective or dispositional reporting
biases or situational differences/inconsistencies (e.g. number of exams and
differences in exam timetables). Chapter 4 will introduce a ball-throwing paradigm
(Williams & Jarvis, 2006; Williams & Sommer, 1997), which has been used
successfully in ostracism research, as a potential laboratory-based social stressor
with which to examine the relationship between neuroticism and situational coping.
63
CHAPTER 4: Social Ostracism and Williams’ Ball-Throwing Paradigm
Social ostracism can be defined as the act of being excluded or ignored by
another individual or group of individuals (Williams & Sommer, 1997). Research
has consistently demonstrated that the effects of social ostracism are both ubiquitous
and powerful (Williams, Cheung, & Choi, 2000; Williams et al., 2002; Zadro,
Williams, & Richardson, 2004). In animals, it has been reported that primates
exclude non-contributing members in order to regulate social behaviour (Goodall,
1986; Lancaster, 1986). In humans social exclusion has been observed in both
primitive (Boehm, 1986; Mahdi, 1986) and modern cultures (Woods, 1978), military
academies (Davis, 1991), schools and academic institutions (Heron, 1987),
workplaces (Faulkner & Williams, 1999; McInnis & Williams, 1999; Williams &
Sommer, 1997), religious groups (Gruter, 1986), and within interpersonal
relationships (Sommer, Williams, Ciarocco, & Baumeister, 2001; Williams &
Sommer, 1997). The prevalence of social ostracism across time, cultures, and
species has led to substantial interest in this construct.
This review chapter has four aims. Firstly, the negative outcomes that are
associated with social ostracism will be outlined. Secondly, a model purporting that
ostracism threatens four fundamental human needs will be briefly discussed.
Thirdly, a ball-throwing paradigm that has been used successfully in laboratory
studies to induce feelings of ostracism will be introduced and research using this
paradigm will be briefly summarised. Finally, it will be argued that the ball-
throwing paradigm may be a promising laboratory-stressor with which to examine
neuroticism-related differences in coping with social stress.
64
The Negative Outcomes of Social Ostracism
Early research in the field of ostracism examined the effects of being
physically isolated from other people (e.g. Schacter, 1951, 1959). Subsequent
research has focused social ostracism; being ignored or excluded in the presence of
others (Williams, Cheung et al., 2000). This review chapter will focus specifically
on research examining social ostracism. Social ostracism is argued to differ from
other forms of rejection in several ways. First, unlike verbal or physical abuse,
ostracism is generally ambiguous (the person being ostracized can be unsure as to
both whether and why ostracism is occurring). Second, ostracism tends to leave
individuals feeling shunned or ignored rather than spotlighted and ridiculed
(Williams et al., 2002). Research examining social ostracism reveals that it is
associated with a variety of negative outcomes.
Individuals who experience social ostracism report depressed mood,
loneliness, anxiety, frustration, feelings of invisibility and helplessness, and anti-
social behaviour (Baumeister & Tice, 1990; Geller, Goodstein, Silver, & Sternberg,
1974; Leary, 1990; Pepitone & Wilpizeski, 1960; Sommer et al., 2001; Twenge,
Baumeister, Tice, & Stucke, 2001; Williams & Sommer, 1997). Additionally,
research suggests that the negative effects of social ostracism increase when targets
attribute being ostracized as due to their own personal shortcoming or limitations
(Nezlek, Kowalski, Leary, Blevins, & Holgate, 1997; Pepitone & Wilpizeski, 1960)
and when the social exclusion is obvious rather than ambiguous (Snoek, 1962).
Furthermore, it appears that even merely imagining (Craighead, Kimball, & Rehak,
65
1979) or role-playing (Williams, Bernieri, Faulkner, Gada-Jain, & Grahe, 2000)
being socially excluded is associated with negative self-evaluations. Similarly, the
negative outcomes of ostracism have prevailed in the face of factors that were
proposed to diminish its negative impact; such as manipulating the in/out group
status of the source of ostracism (Eisenberger, Lieberman, & Williams, 2003;
Gonsalkorale & Williams, 2006). Finally, being socially excluded or isolated has
also been linked to detriments in physical health (Cacioppo & Hawkley, 2003).
Specifically social exclusion has been linked with an increased risk of heart attack
(Case, Moss, Case, McDermott, & Elberly, 1992) and poorer blood pressure
regulation (Uchino, Cacioppo, & Kiecolt-Glaser, 1996).
So why are the effects of being socially excluded so powerful? From an
evolutionary perspective it is argued that group affiliation and acceptance are not
only emotionally desirable, but are evolutionarily adaptive as well (Buss, 1990). For
social animals group inclusion meant fulfillment of needs such as nutrition, security,
and sexual partners (Buss & Kenrick, 1998). In contrast, social exclusion threatened
both survival and gene transmission (Baumeister & Leary, 1995). Williams and
Zadro (2001) suggest that because of the adaptive nature of social groups, and the
dire consequences of isolation in our evolutionary past, individuals are motivated to
maintain their group acceptance and are therefore sensitively attuned to information
relative to this motive. Indeed, it is argued that early detection of anything signaling
social exclusion (or possible future social exclusion) would have evolved as an
important mechanism (Williams & Zadro, 2001).
66
Consistent with this argument, neuropsychologists have identified a
biological system that may function to protect individuals from social exclusion.
Social Pain/Physical Pain Overlap Theory hypothesizes that social pain (the pain
that is experienced when social relationships are damaged or lost) and physical pain
(the pain that is experienced upon physical injury) share parts of the same
underlying processing system (Eisenberger & Lieberman, 2004, 2005; MacDonald,
Kingsbury, & Shaw, 2005). Eisenberger, and her colleagues (2003) examined brain
activity using functional magnetic resonance imaging and found that the anterior
cingulate cortex, the same area of the brain which mediates sensations of physical
pain (e.g. Price, 2000), shows increased activity in socially excluded individuals.
From this perspective, the negative outcomes associated with social ostracism may
act as an alarm system, which allows the excluded individual to alter their behaviour
in order to be socially accepted (Eisenberger & Lieberman, 2004). This argument is
also supported by research which shows that when the opportunity is presented
socially ostracized individuals attempt to get back into the good graces of the
individuals who have excluded them (Snoek, 1962; Williams & Sommer, 1997).
Similarly, Williams and colleagues (2000) report that after experiencing social
ostracism individuals are more likely to conform with incorrect responses on a
subsequent embedded figures task.
Williams’ Need-Threat Model of Ostracism
Williams (1997) proposes that four fundamental human social needs are
uniquely and simultaneously effected by ostracism; the need for belonging, control,
67
self-esteem, and meaningful existence. Indeed, there is evidence supporting the
notion that each of these four needs is fundamental to the well-being of an
individual. Baumeister and Leary (1995) argue that the need to belong (and form
interpersonal attachments) is a fundamental motive for social interaction and has
important consequences for social functioning. An abundance of research has
revealed that without control (or at least perceptions of control) individuals may
exhibit learned helplessness and become depressed (Bandura, 1995; Seligman, 1975;
Skinner, 1996). With regard to self-esteem, theory and research both suggest that it
could be the primary determinant of self-efficacy and mental health (Barnett &
Gottlib, 1988; Leary, Tambor, Terdal, & Downs, 1995). Finally, individuals have a
fundamental need to believe that their existence is meaningful (James, 1890; Mead,
1934), and to avoid contemplating their mortality (Greenberg, Pyszczynski, &
Solomon, 1986).
Williams (1997) acknowledges that these four fundamental needs are not
exclusively threatened by social ostracism but he does argue that ostracism acutely
and immediately threatens all four of these needs. By definition, when an individual
is ostracized any connection between a source and a target of ostracism is severed.
Therefore, any sense of belonging is shattered. Williams (1997) also argues that
because ostracized individuals are given a very real insight into their social death,
they also suffer deficits in the fulfillment of their need for meaningful existence.
Furthermore, Williams (2001) suggests that the causes of social ostracism tend to be
ambiguous. Therefore, the ostracized individual is left to generate a list of personal
flaws and deficits that may account for their social exclusion (Williams, 2001).
68
Sommer and colleagues (2001) have reported that focusing on personal flaws
diminishes self-esteem. Finally, Zadro and colleagues (2004) report that social
ostracism is a greater threat to control than other aversive inter-personal behaviours,
because ostracized individuals are given little or no opportunity to defend
themselves.
According to the Need-Threat model of ostracism individuals pass through
up to three stages in their response to social ostracism. The immediate reaction
when the four fundamental needs are threatened is physiological arousal, worsened
mood, lowered self-esteem and apprehension (Williams & Sommer, 1997; Williams
& Zadro, 2001). Over the short-term ostracized individuals attempt to restore their
threatened needs through behavioural, cognitive, or emotional means. For example
laboratory studies have demonstrated that after being ostracized individuals attempt
to work harder for the group (Williams & Sommer, 1997) and are more likely to
conform to erroneous visual perceptions (Williams, Cheung et al., 2000). Finally,
chronic and long-term social ostracism is argued to have negative outcomes such as
depression and even suicide (Williams & Zadro, 2001).
Laboratory studies of ostracism have consistently demonstrated the utility of
Williams’ Need-Threat model (Williams, 1997; Williams & Zadro, 2001) in
understanding the aversive impact of social ostracism. Experiments have
demonstrated that, regardless of whether the source of ostracism is role-played
(Williams, Bernieri et al., 2000), by diary (Williams, Wheeler, & Harvey, 2001),
interview (Zadro & Williams, 1999), or experimental manipulation (Williams,
69
Cheung et al., 2000; Zadro et al., 2004), all four fundamental needs are consistently
threatened
The Ball-Throwing Paradigm and Cyberball
Williams and Sommer (1997) developed a ball-throwing paradigm for the
study of social ostracism. They led participants to believe they were playing a
spontaneous ball-throwing game with two other individuals. Unbeknown to the
participants, the other players were confederates of the experimenter. Participants
were randomly assigned to either an included or ostracized condition. If participants
were assigned to the ostracized condition then, after a few initial throws, the
confederates stopped throwing the ball to the participants. In effect, the participant
was excluded and left to watch the confederates throw the ball to each other.
Williams and Sommer (1997) reported that after only approximately one minute of
being excluded by other players (who were strangers the participant was likely never
to see again) participants started to slump and look dejected. In post-task self-report
measures all four fundamental needs as well as participants’ mood were reduced.
More recently, Williams and his colleagues (Williams, Cheung et al., 2000;
Zadro et al., 2004) have developed Cyberball; a virtual analogue of the ball-throwing
game (Williams & Jarvis, 2006). The procedure for utilizing Cyberball involves
informing participants that they will be engaging in a mental visualization
experiment; and that to assist them in practicing their mental visualization skills they
will be playing a ball-throwing game over the internet with two other participants.
70
In reality participants play a predetermined computer program designed to either
include or exclude them. Participants are told that their performance in the ball-
throwing game is unimportant; instead the game is merely a means for them to
engage their mental visualization skills. Participants are asked to visualize the
situation, themselves, and the other players and maintain this mental representation
throughout the game. The game depicts three ball-throwers with the middle one (the
hand) representing the participant (see Figure 1). If desired, photographs identifying
the two computerized ‘players’ and the participant can be displayed on the computer
screen.
Figure 1: Screen shot of the Cyberball game
The game is animated and shows the cartoon character throwing a ball to one
of the other two characters. When the ball is thrown to participants they are told to
click on one of the other two characters, to indicate their chosen recipient. The ball
is then thrown to that character. The game can be set for any number of throws,
although 40 throws is generally used in Cyberball research (e.g. Zadro et al., 2004).
71
Participants are randomly allocated to either an included condition (receive the ball
approximately one third of the time) or an excluded condition (receive the ball twice
at the very start of the game and then do not receive the ball again). After
completing the Cyberball game participants complete a post-experiment
questionnaire tapping the four fundamental needs as well as mood. These short
scales have been developed by Williams and his colleagues. Once participation in
the experiment is complete individuals are thoroughly debriefed about the actual
purpose of the experiment(s).
Research using the Cyberball paradigm has revealed that even when social
ostracism occurs over the computer (where the ostracizing others are unknown,
unseen, and will never be met) it elicits aversive reactions (Williams, Cheung et al.,
2000). Moreover, subsequent experiments using Cyberball have found similar
results despite increased psychological and physical distance between the ostracized
individuals and the sources of ostracism. For example, Zadro and colleagues (2004)
reported that even after participants were informed they were playing Cyberball
against a computer (and not real people) excluded participants reported lower levels
of the four fundamental needs in comparison to included participants. Similarly,
Eisenberger and colleagues (2003) informed participants that a technical
malfunction meant that they could not participate in the game, but could only watch
it. They reported that those watching the excluded condition reported lower need
levels than those watching the included condition (Eisenberger et al., 2003). Finally,
studies using the Cyberball paradigm have found no individual differences variables
which moderate the aversive effects of ostracism. Additionally, studies
72
manipulating the identity of the sources of ostracism have found no moderation
effects. For example, Gonsalkorale and Williams (2006) used the Cyberball
paradigm to demonstrate that ostracism by a despised political group was no less
aversive than exclusion by ‘in-group’ members from admired political parties.
Studies using the ball-throwing paradigm reveal that social ostracism results
in a quick and painful reaction; regardless of the source (Gonsalkorale & Williams,
2006), context (Eisenberger et al., 2003), or setting (Williams, Cheung et al., 2000)
of the ostracism. The fact that, at present, no variables which moderate the
deleterious effects of ostracism have been identified has led Williams (2001) to
conclude that individuals react negatively (even to small amounts of ostracism in
seemingly unimportant conditions) as the result of an evolutionary system designed
to protect early humans. According to Williams, humans appear to be innately
prepared to detect and respond to threats involving their acceptance by other people.
Such reasoning is consistent with Social Pain/Physical Pain Overlap Theory, as
reductions in the feelings of belonging, self-esteem, control, and meaningful
existence may be conceptualized as the emotional output of an early warning system
that is designed to detect potential social exclusion well in advance of actual
rejection, so that the individual may act to protect relationships that may be in
jeopardy.
The Ball-Throwing Paradigm and Coping Research
As mentioned in Chapters 2 and 3, previous research examining the
relationship between personality variables and coping has been limited in its reliance
73
on dispositional and retrospective coping measures (David & Suls, 1999). David
and Suls (1999) argue that obtaining coping reports immediately after a stressful
event has occurred would allow for firmer conclusions regarding the relationship
between personality traits and coping to be made. Study 1 attempted to examine the
relationship between personality and coping utilizing approaching university
examinations as a situation-specific stressor. However, using approaching
examinations as a stressor had a number of limitations. Firstly, individuals are likely
to have experienced an event of this nature before, and this may impact on their self-
reporting. Secondly, methodological limitations of Study 1 meant individuals could
respond to the measures at any time within a two week period. It was concluded that
a novel, time-constrained stressor would allow for a better examination of possible
relationships between personality and coping. Furthermore, in a recent meta-
analysis of experiments examining personality and coping, Connor-Smith and
Flachsbart (2007) conclude that that the impact of personality on coping could best
be identified through the use of standardised laboratory-stressors; as presenting the
same objective stressor to all participants minimizes confounds and allows for
immediate self-reports of appraisal and coping (Connor-Smith & Flachsbart, 2007).
The Cyberball paradigm appears promising as laboratory-based stressor for a
number of reasons. Firstly, being excluded during the ball-throwing game is
associated with a variety of negative outcomes. In particular, the negative impact
that being excluded during the Cyberball game has on mood and self-esteem,
suggests that social exclusion evokes a negative affective response. However, these
findings have been obtained using short scales developed by Williams and his
74
colleagues, and should be replicated using well-known and well-validated mood and
self-esteem measures. Secondly, Williams argues that over the short-term ostracized
individuals attempt to restore their threatened needs through behavioural, cognitive,
or emotional means. Therefore, it is likely that being ostracized during the ball-
throwing game will elicit a coping response. Thirdly, the ball-throwing game is
novel. Individuals are unlikely to have experienced this situation before. This
should minimize the influence of prior experiences on coping self-reports. Fourthly,
participants can complete all measures immediately after playing the Cyberball
game, thereby overcoming the limitations of dispositional and retrospective. Fifthly,
all participants are faced with exactly the same scenario. Therefore, any differences
in how participants respond to the ball-throwing game should reflect person-related
differences rather than stressor-related differences. Finally, Cyberball represents a
controlled analogue of a social situation. Previous research has identified that
cognitive tasks, such as rapid information processing, driving simulation, and
working memory tasks (Matthews & Campbell, 1998; Matthews et al., 2006), and
physical stimuli (e.g. cold and noise) can be used in laboratory settings to elicit a
coping response. Given the immense difficulties in controlling variables in real-
world social settings, Cyberball appears promising as a controlled laboratory-
stressor with which to examine how individuals cope with social stress.
75
Summary
Studies using the ball-throwing paradigm reveal that social ostracism results
in a quick and painful reaction; regardless of the source (Gonsalkorale & Williams,
2006), context (Eisenberger et al., 2003), or setting (Williams, Cheung et al., 2000)
of the ostracism. Moreover, Williams and his colleagues argue that ostracized
individuals attempt to restore their threatened needs through behavioural, cognitive,
or emotional means. This raises the possibility of using the Cyberball game as a
controlled stressor with which to examine coping with social stress. Importantly,
Cyberball represents a controlled social situation, in which excluded participants all
face exactly the same scenario. This should provide clarity in interpreting individual
differences in coping. Moreover, by responding to all measures immediately after
completing the game, limitations of dispositional and retrospective self-reporting are
minimized. Studies 2, 3, and 4 (Chapters 5, 6, and 7) examined, in more detail, the
potential of the Cyberball game as a laboratory-based stressor with which to
examine neuroticism-related differences in coping with social exclusion. As
mentioned in Chapter 3, the use of generic coping measures is not appropriate in the
context of constrained laboratory tasks. Therefore, the CITS-S (Matthews &
Campbell, 1998), which has been designed specifically for use in task situations (see
page 81), will be the coping measure utilized for the remainder of this thesis.
76
CHAPTER 5: Study 2 – Cyberball as an Experimental Stressor
Aims and Hypotheses of Study 2
Study 2 had four major aims. Firstly, previous research has revealed that
being ostracized during the Cyberball game has a negative effect on mood and self-
esteem (at least when short scales designed by the researchers are used, Williams &
Sommer, 1997; Zadro et al., 2004). Study 2 aimed to replicate this negative effect of
ostracism on mood and self-esteem using well-known and well-validated outcome
measures. Based on previous findings using the ball-throwing paradigm (Williams &
Sommer, 1997; Zadro et al., 2004), it was hypothesized that individuals who were
ostracized during the game would report worse mood and lowered self-esteem when
compared to individuals who were included during the game; even when widely-
used outcome measures were employed.
Secondly, Study 2 aimed to assess the utility of Cyberball as an experimental
social stressor by determining whether it can evoke a coping response. If being
excluded during the Cyberball game can evoke both a negative affective response
and a coping response, it may be a useful laboratory-based stressor with which to
examine individual differences in coping with social stress. It was predicted that
individuals who were excluded during Cyberball would engage in significantly more
emotion-focused and avoidance coping than individuals included during the game.
Due to the ambiguity in the definition of the task (participants believe their primary
77
task is mental visualization); no predictions were made regarding task-focused
coping.
Thirdly, Study 2 aimed to examine neuroticism-linked differences in both
appraisal and coping during the Cyberball game. In accordance with previous
research (e.g. Engelhard, 2007; Gallagher, 1990; Mak et al., 2004; Schneider, 2004)
as well as the results obtained in Study 1 (using approaching exams as a situational
stressor), high neuroticism participants were predicted to appraise the Cyberball
game more negatively than low-neuroticism participants. With regard to coping,
based on previous research using retrospective and dispositional coping measures
(e.g. Bouchard, 2003; Carver et al., 1989; Endler & Parker, 1990b; R. Gomez,
Holmberg, Bounds, Fullarton, & Gomez, 1999), the results obtained in Study 1, as
well as the laboratory-based findings of Matthews and colleagues using cognitive
tasks (Matthews & Campbell, 1998; Matthews et al., 2006), it was predicted that
neuroticism-related differences in emotion-focused and avoidance coping would be
obtained in the excluded Cyberball condition.
Finally, according to the goodness of fit hypothesis (see Chapter 1) stressor
appraisals (in particular controllability appraisals) should influence the relationship
between stressful situations and coping strategy choice. Given that neuroticism is
associated with negative appraisals of the environment (e.g. Engelhard, 2007;
Gallagher, 1990; Mak et al., 2004; Schneider, 2004), it may be the case that the
relationship between neuroticism and emotion-focused/avoidance coping is
mediated by negative stressor appraisals. If neuroticism-related differences in both
appraisals of the Cyberball game and coping strategy use were obtained, then the
78
final aim of Study 2 was to determine whether stressor appraisals mediate the
relationship between neuroticism and self-reported coping.
Method
Participants
Three hundred and thirty-four undergraduate psychology students were
screened on neuroticism, and the top and bottom quartiles were invited to participate
in the study. Eighty-nine individuals aged between 17 and 55 years (mean age =
20.44 years) actually took part in the study. The sample was stratified so that
approximately equal numbers of low neuroticism (n = 43) and high neuroticism (n =
46) participants were recruited. Approximately equal numbers of male (n = 42) and
female (n = 47) participants were recruited. Participants were randomly assigned to
either the included condition (n = 44) or the excluded condition (n = 45).
Materials
Cyberball8
Cyberball (Williams & Jarvis, 2006) is an online animated ball-throwing
computer game. Participants were told they would be taking part in a mental
visualization experiment and that to assist them in practicing their skills at mental
visualization they would be participating in an online ball-throwing game with two 8 A downloadable version of this game is available at: http://www2.psych.purdue.edu/~kip/Announce/cyberball.htm
79
additional participants. Participants were told that their performance in the game
was unimportant and that the game was merely a means for them to engage their
mental visualization skills. They were asked to visualize the situation, themselves,
and the other ‘players’. The game depicts three ball-throwers, the middle one
representing the participant (see Figure 1 in Chapter 4 for a screenshot of Cyberball).
The game is animated and shows the icon throwing a ball to one of the other two
players. Participants could choose who they wished to throw the ball to by clicking
on the photograph of one of the other two ‘players’. The game was set for a total of
40 throws. Participants were randomly allocated to either an included condition
(received the ball approximately one third of the time) or an excluded condition
(received the ball twice at the start of the game but were then excluded for the
remainder of the game).
Measures
Neuroticism Screening
A 10-item scale (see Appendix D) compiled from the International
Personality Item Pool (Goldberg et al., 2006) was used to screen participants on
neuroticism. Items are responded to on a five-point scale (0: Very inaccurate; 4:
Very accurate). The neuroticism scale has an internal consistency of .86 (Goldberg
et al., 2006) and correlates highly with other neuroticism measures (e.g. a correlation
of .84 with the NEO-FFI neuroticism subscale; Gow, Whiteman, Pattie, & Deary,
2005).
80
NEO-Five Factor Inventory (NEO-FFI)
Participants completed the NEO-FFI (Costa & McCrae, 1992) at time of
testing. The NEO-FFI is a short-form version of the revised NEO Personality
Inventory (NEO-PI-R). It contains 60 items with 12 items assessing each of the Big
Five personality factors. Items are rated on a five point scale (0: Strongly disagree;
4: Strongly agree). The NEO-FFI scales are highly correlated with the
corresponding scales on the full NEO-PI-R; convergent correlations range from .77
to .94 across samples, (Watson & Hubbard, 1996). Costa and MaCrae (1992) have
reported internal consistencies of the five subscales to range between .68
(agreeableness) and .86 (neuroticism). The neuroticism (α = .86) subscale was of
particular interest in Study 2.
Mood
The UWIST Mood Adjective Checklist (MACL; Matthews, Jones, &
Chamberlain, 1990) was used to measure three bipolar mood dimensions; energetic
arousal (vigorous vs. tired), tense arousal (nervous vs. relaxed), and hedonic tone
(pleasant vs. unpleasant mood). Subscales consist of eight items (with the exception
of the hedonic tone subscale which contains 13 items) and are responded to on a
four-point scale (1: Definitely; 4: Definitely not). In Study 2 items were scored so
that a higher score indicated a higher level of the named dimension; a high score on
the energetic arousal scale indicated a high level of energy, a high score on the tense
arousal scale indicated high levels of tension, and a high score on the hedonic tone
scale indicated a high level of pleasant mood. The internal consistencies for the
81
energetic arousal, tense arousal, and hedonic tone sub-scales are .83, .88, and .88
respectively (Matthews et al., 1990). The MACL also contains five items measuring
anger/frustration. These items load onto the hedonic tone sub-scale (Matthews et al.,
1990)9.
Self-Esteem
The Rosenberg Self-Esteem scale (RSES; Rosenberg, 1965) was used to
measure self-esteem in Study 2. The RSES is a widely used and validated self-
esteem measure consisting of 10 items (five positively worded and five negatively
worded) which are summed to give a total score (Aluja, Rolland, Garcia, & Rossier,
2007). The Rosenberg Self-Esteem Scale is responded to on a four-point scale (1:
Strongly agree; 4: Strongly disagree). Extensive reliability and validity data exist
for the RSES (Blascovich & Tomaka, 1991) and in Study 2 the internal consistency
of .86. The RSES was scored so that a higher score was indicative of higher self-
esteem.
Cyberball Appraisal Scale
An eight item scale was constructed to measure participants’ appraisals of
the Cyberball game (see Appendix E). Five items assessed how excluded
participants perceived themselves to be during the game (e.g. ‘during the game, to
9 There are two ways of scoring the Hedonic Tone subscale: 1) separate Hedonic Tone and Anger/Frustration scores can be calculated, 2) all 12 items can be summed giving a total Hedonic Tone score. Given the psychometric properties of the scale (Matthews et al., 1990) it was determined that calculating a total Hedonic Tone score was the most psychometrically sound method of obtaining a Hedonic Tone score
82
what extent did you feel left out’, α = .92) and three items measured perceived
control over the game (e.g. ‘to what extent did you feel you had control over the
game’, α = .88). Items are responded to on a four-point scale (1: Not at all; 4:
Extremely). These subscales will be referred to as Appraisal (Excluded) and
Appraisal (Control) respectively. A principal components analysis (using oblique
rotation) revealed that the two predicted components emerged clearly and accounted
for 79.02 percent of overall variance.
Coping Inventory for Task Stressors (Situational Version, CITS-S):
The CITS-S (Matthews & Campbell, 1998) is a questionnaire that has been
developed for immediate post-task assessment of coping (see Appendix F). It
consists of task-focused (e.g. ‘I concentrated hard on doing well’), emotion-focused
(e.g. ‘I wished I could change what was happening’), and avoidance (e.g. ‘I stayed
detached or distanced from the situation’) coping sub-scales. Each sub-scale
contains seven items which are responded to on a five-point scale (0: Not at all; 4:
Extremely). The internal consistencies for the task-focused, emotion-focused, and
avoidance sub-scales in Study 2 were 0.79, .76, and .77 respectively.
Procedure
After reading an information sheet and signing a consent form all
participants were randomly allocated to either the included or excluded condition.
Participants were informed that the current study aimed to assess the relationship
between personality and mental visualization skill. With their consent, the
83
participants’ photograph was taken, in order to identify them on screen during the
Cyberball game. Participants first completed the NEO-FFI and then played the
Cyberball game. After playing Cyberball, participants completed the MACL, RSES,
Appraisal scale, and CITS-S in that order. After completing all measures
participants were thoroughly debriefed as to the actual purpose of the experiment.
Results
A significant difference in NEO-FFI neuroticism score was obtained between
the low neuroticism (mean = 13.28) and high neuroticism (mean = 23.35) groups at
time of testing; F(1, 87) = 34.68, p < .001. However, an examination of the
distribution NEO-FFI neuroticism scores (obtained at time of testing) revealed that
rather than being bimodal, scores were normally distributed. Therefore, a median
split was conducted (using NEO-FFI scores, median = 18) in order to classify
individuals as low neuroticism (n = 42) or high neuroticism (n = 47). No difference
in mean NEO-FFI neuroticism score was obtained between the included and
excluded conditions, F(1, 87) = 1.93, p = .169. Given that gender differences in
neuroticism have been reported (Hankin & Abramson, 2001), gender was entered as
a covariate in all analyses10. A series of MANCOVAs were conducted in order to
examine Cyberball appraisal, mood and self-esteem, and coping as a function of
Condition (included vs. excluded) and Neuroticism Level (low vs. high). No
10 Removing gender as covariate did not change any of the reported results
84
significant multivariate or univariate effects of gender were obtained. Bonferroni
correction was implemented to set critical α level for all univariate analyses.
Table 3: Differences in appraisal (exclusion), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being included or excluded during the Cyberball game in Study 2
* Significant after Bonferroni correction
Cyberball Appraisal
After controlling for gender a significant multivariate effect of Condition was
obtained; Wilks’ Lambda (λ) = .58, F(2, 83) = 29.55, p < .001, partial η2 = .42. The
multivariate effect of Neuroticism Level; λ = .99, F(2, 83) = .44, p = .647, partial η2
= .01; and the multivariate interaction between Condition and Neuroticism Level; λ
Mean SD F df Error df p Partial η2 Appraised Exclusion: Included Excluded
9.75
16.07
3.77 4.76
47.36
1
84
< .001*
.36 Appraised Control: Included Excluded
12.00 9.07
2.48 1.97
40.45
1
84
< .001*
.33 Energetic Arousal: Included Excluded
21.61 19.69
4.45 4.32
4.55
1
84
.036
.05 Tense Arousal: Included Excluded
13.61 16.49
3.62 5.29
9.82
1
84
.002*
.11 Hedonic Tone: Included Excluded
34.70 33.04
1.89 2.95
8.79
1
84
.004*
.10 RSES: Included Excluded
29.70 26.13
3.84 4.91
19.23
1
84
< .001*
.19 T-F Coping: Included Excluded
13.14 13.98
4.62 4.17
.40
1
84
.530
.01 E-F Coping: Included Excluded
2.70 4.56
2.53 3.85
6.94
1
84
.010*
.08 Avoidance Coping: Included Excluded
6.54 7.96
4.30 4.32
2.26
1
84
.136
.03
85
= .96, F(2, 83) = 1.55, p = .217, partial η2 = .04; were non-significant. Critical α
level was set at .025 (after correction) for all univariate analyses. Ostracized
participants felt significantly more excluded and perceived themselves as having less
in control during Cyberball than individuals who were included during the game (see
Table 3). No significant effects of Neuroticism Level were obtained (see Table 4),
and neither of the univariate interactions between Condition and Neuroticism level
was statistically significant.
Mood and Self-Esteem
After controlling for gender significant multivariate effects of Condition; λ =
.78, F(4, 81) = 5.71, p < .001, partial η2 = .22; and Neuroticism Level; λ = .87, F(4,
81) = 3.13, p = .019, partial η2 = .13; were obtained. The multivariate interaction
between Condition and Neuroticism Level was not statistically significant; λ = .97,
F(4, 81) = .58, p = .681, partial η2 = .03. Critical α level was set at .012 (after
correction) for all univariate analyses. Ostracized individuals reported significantly
higher levels of tense arousal and significantly lower levels of hedonic tone (see
Table 3). Although approaching significance, the difference in energetic arousal
was not statistically significant after correction. No significant effects of
Neuroticism Level were obtained on any of the mood sub-scales (although
approaching significance, the difference in tense arousal between high and low
neuroticism participants was not significant after correction – see Table 4) and none
of the univariate interactions between Condition and Neuroticism Level were
statistically significant. With regard to self-esteem, excluded participants reported
86
significantly lower levels of self-esteem (see Table 3). High neuroticism participants
reported significantly lower levels of self-esteem, regardless of which condition they
were assigned to (see Table 4). The interaction between Condition and Neuroticism
Level in relation to self-esteem was not significant.
Table 4: Differences in appraisal (exclusion), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 2
* Significant after Bonferroni correction
Coping
After controlling for gender the multivariate effect of Condition was
approaching statistical significance; λ = .91, F(3, 82) = 2.53, p = .063, partial η2 =
Mean SD F df Error df p Partial η2 Appraised Exclusion: Low N High N
12.76 13.11
5.45 5.26
.87
1
84
.354
.01 Appraised Control: Low N High N
10.52 10.51
2.97 2.40
.311
1
84
.579
.00 Energetic Arousal: Low N High N
20.47 20.79
4.68 4.30
.03
1
84
.863
.00 Tense Arousal: Low N High N
14.19 15.85
4.78 4.61
3.88
1
84
.052
.04 Hedonic Tone: Low N High N
33.71 34.00
2.67 2.56
.09
1
84
.765
.00 RSES: Low N High N
29.19 26.74
4.95 4.27
9.54
1
84
< .001*
.10 T-F Coping: Low N High N
14.07 13.11
4.32 4.46
1.02
1
84
.317
.01 E-F Coping: Low N High N
3.76 3.53
3.70 3.10
.01
1
84
.784
.00 Avoidance Coping: Low N High N
7.48 7.06
4.49 4.25
.09
1
84
.760
.00
87
.09. The multivariate effect of Neuroticism Level; λ = .98, F(3, 82) = .50, p = .680,
partial η2 = .02; and the multivariate interaction between Condition and Neuroticism
Level; λ = .97, F(3, 82) = .83, p = .483, partial η2 = .03; were not statistically
significant. Critical α level was set at .016 (after correction) for all univariate
analyses. Excluded participants engaged in significantly more emotion-focused
coping than participants who were included during the Cyberball game (see Table
3). No differences in coping were obtained as a function of Neuroticism Level (see
Table 4) and none of the univariate interactions between Condition and Neuroticism
Level were statistically significant.
Mediational Analyses
Given that no neuroticism-related differences in either appraisal or coping
were obtained in Study 2, analyses examining whether stressor appraisals mediate
possible relationships between neuroticism and coping were not conducted.
Discussion
Study 2 aimed to 1) replicate previously reported negative ostracism effects
with widely-used and well-validated outcome measures, 2) determine whether being
excluded during Cyberball could evoke a coping response, 3) examine any
neuroticism-related differences in appraisal and coping during the Cyberball game,
and 4) determine whether stressor appraisals mediate any relationship(s) between
neuroticism and coping in the context of the Cyberball game.
88
Results revealed that the ostracism manipulation was successful. Ostracised
participants felt significantly more excluded and perceived themselves as having
significantly less control during the Cyberball game in comparison with individuals
who were included during Cyberball. Additionally, excluded individuals reported
significantly higher levels of tense arousal, and significantly lower levels of hedonic
tone and self-esteem. The findings from Study 2 therefore offer strong support for
Williams and colleagues contention that being excluded during Cyberball induces
feelings of social ostracism and that this ostracism has measurable negative
outcomes (e.g. Williams & Sommer, 1997; Zadro et al., 2004). Specifically, the
findings from Study 2 show that both mood and self-esteem are negatively affected
by being excluded during the Cyberball game. Importantly, this is the case even
when widely used scales, such as the MACL and the RSES are used as outcome
measures.
Study 2 also revealed that participants excluded during the Cyberball game
engaged in significantly more emotion-focused coping than included participants. It
therefore appears that being socially ostracized during the Cyberball game can evoke
a coping response. This finding reveals that Cyberball is a promising experimental
stressor with which to examine individual differences in coping with social stress.
Laboratory-stressors are useful because they overcome the reliance on dispositional
and retrospective methodologies that have limited much of the previous research in
this area (David & Suls, 1999). Moreover, presenting the same objective stressor to
all participants minimizes contextual confounds and allows firmer conclusions
regarding individual differences in the coping process to be made (Connor-Smith &
89
Flachsbart, 2007). From this perspective, Cyberball is likely to be a useful paradigm
for investigating individual differences in coping with social stress; as it is
exceptionally difficult to reproduce stressful social situations in a laboratory setting.
In Study 2, high neuroticism individuals reported significantly lower levels
of self-esteem regardless of which Cyberball condition they were assigned to. This
findings is consistent with numerous studies reporting that neuroticism is associated
with negative self-appraisals (Clark & Watson, 1991; Watson & Clark, 1984).
However, inconsistent with the findings of Study 1, no neuroticism-linked
differences in appraisals of the Cyberball game or self-reported coping were
obtained in Study 2. Given previous findings reporting relationships between
neuroticism, negative stressor appraisals, emotion-focused coping, and avoidance
coping when dispositional and retrospective coping measures are employed (e.g.
Bouchard, 2003; Carver et al., 1989; Endler & Parker, 1990b; R. Gomez et al.,
1999), as well as the results reported by Matthews and his colleagues (Matthews &
Campbell, 1998; Matthews et al., 2006) showing significant correlations between
neuroticism, emotion-focused, and avoidance coping in a variety of cognitive tasks,
the results of Study 2 were unexpected. However, before it can be concluded that
there are no neuroticism-linked differences in appraisal or coping during the
Cyberball game two methodological limitations of Study 2 need to be considered
and addressed.
Firstly, because any predicted effects should be small (due to the mild nature
of the stressor) it may be the case that the range of response options on the CITS-S
and Cyberball Appraisal Scale was not large enough to identify small neuroticism-
90
related differences in appraisal and coping. Indeed, Connor-Smith and Flachsbart
(2007) have argued that stressors which are limited in duration or scope provide less
room for individual differences to operate. Therefore, Study 3 (Chapter 6) increased
the range of response options in order to determine if small neuroticism-related
differences in coping could be obtained.
Secondly, no neuroticism-related differences in appraisals of the Cyberball
game were obtained in Study 2. This finding was unexpected as previous research
suggests that high neuroticism individuals are prone to make negative appraisals of
their environments (Engelhard, 2007; Gallagher, 1990; Mak et al., 2004; Schneider,
2004). Additionally, some research suggests that high neuroticism individuals are
particularly likely to appraise ambiguous stimuli in a threatening manner (e.g.
Watson & Clark, 1984). It is possible that the reason no differences in appraisal
were obtained in the current experiment was because the degree of social ostracism
experienced in the excluded condition was so unequivocal that it was no longer an
ambiguous situation. Neuroticism-linked differences in appraisals of the Cyberball
game may become apparent when the social exclusion experienced is more
ambiguous in nature. Study 4 (Chapter 7) expanded the range of response scales and
manipulated the extent to which participants were excluded during the Cyberball
game; in order to determine whether neuroticism-linked differences in appraisal and
coping could be obtained when the social exclusion experienced was more
ambiguous in nature.
91
CHAPTER 6: Study 3 – Increasing the Range of the Response Scales on
Outcome Measures
Aims and Hypotheses of Study 3
Given that the social stress experienced during the Cyberball game (which
lasts only approximately three or four minutes) is fairly mild in nature, large
individual differences in coping responses are unlikely (Connor-Smith & Flachsbart,
2007). Similarly, given that the social ostracism experienced in the excluded
condition is completely unequivocal in nature it is also unlikely that large individual
differences in appraisals of the Cyberball game would be obtained. It is therefore
possible that the reason no neuroticism-related differences in either appraisal or
coping were obtained in Study 2 was because the range of response options on the
questionnaires was not broad enough to detect small neuroticism-related differences
in appraisal and coping. Study 3 increased the range of response options on the
outcome measures in order to determine if small neuroticism-related differences in
both appraisal and coping could be obtained in the context of the Cyberball game.
Based on previous ostracism research (e.g. Zadro et al., 2004) as well as the
findings of Study 2, it was hypothesized that being excluded during Cyberball would
result in negative appraisals of the game. It was also hypothesized that ostracized
participants would report worse mood and lowered self-esteem when compared with
individuals who were included during the game. Specifically with regard to mood,
given the findings obtained using the MACL (Matthews et al., 1990) in Study 2, it
92
was predicted that being ostracized would result in significantly lower levels of
hedonic tone and significantly higher levels of tense arousal. Finally, based on the
findings of Study 2 it was predicted that being excluded during the game would be
associated with an emotion-focused coping response. Additionally, with the
implementation of expanded response scales an avoidance coping response was also
predicted to be associated with being excluded during the Cyberball game in the
current experiment.
With regard to neuroticism, it was predicted that the expanded response
scales would allow small neuroticism-linked differences in both appraisals of
Cyberball and coping to be obtained in Study 3. Therefore, in accordance with
previous research (e.g. Engelhard, 2007; Gallagher, 1990; Mak et al., 2004;
Schneider, 2004) and Study 1, high neuroticism participants were predicted to
appraise the Cyberball game more negatively than low-neuroticism participants.
With regard to coping, based on previous research using both retrospective and
dispositional methodologies, as well as experimental cognitive tasks (e.g. Bouchard,
2003; Carver et al., 1989; Endler & Parker, 1990b; R. Gomez et al., 1999; Matthews
& Campbell, 1998; Matthews et al., 2006), it was predicted that neuroticism-related
differences in emotion-focused and avoidance coping would be obtained in the
excluded Cyberball condition. Finally, if neuroticism-linked differences in both
appraisals and coping could be obtained, then Study 3 aimed to determine whether
the relationship between neuroticism and coping in the context of Cyberball is
mediated by appraisals of the game.
93
Method
Participants
Ninety-three participants aged between 17 and 26 years (mean = 20.91 years)
took part in Study 3. This study was conducted partly in conjunction with a group of
undergraduate students who were completing a research design and methodology
unit. Due to this, pre-screening participants on neuroticism was beyond the scope of
this study. Participants were all undergraduate psychology students and were friends
of the researchers. Participants were randomly allocated to either the included (n =
44) or excluded (n = 49) condition. The gender ratio of the sample reflected that of
the undergraduate psychology population (61 females; 32 males).
Measures
Personality
Given that this experiment was conducted in conjunction with undergraduate
students, financial constraints prohibited the use of the NEO-FFI. Therefore, the 10
item scale (see Appendix D) compiled from the International Personality Item Pool
(Goldberg et al., 2006, used as the screening measure in Study 2) was used to
measure neuroticism at time of testing. This neuroticism scale correlates very highly
with the NEO-FFI neuroticism scale (r = .84; Gow et al., 2005). Items were
responded to on a five-point scale (0: Very inaccurate; 4: Very accurate). The
neuroticism scale has an internal consistency of .86 (Goldberg et al., 2006).
94
Other Measures
Mood, self-esteem, Cyberball appraisal, and coping measures were the same
as those used in Study 2; however, the response scales of the MACL (0: Definitely;
10: Definitely Not), RSES (0: Strongly Agree; 10: Strongly Disagree), Cyberball
Appraisal Scale (0: Not at All; 10: Extremely), and CITS-S (0: Not at All; 10:
Extremely) were expanded11. The internal consistencies of all subscales (using
expanded response scales) obtained in the sample from Study 3 are summarised in
Table 23 (see Appendix G).
Procedure
After reading an information sheet and signing a consent form all
participants were randomly allocated to either the included or excluded condition.
Participants were informed that the experiment aimed to assess the relationship
between personality and mental visualization skill. With their consent, the
participants’ photograph was taken, in order to identify them on screen during the
Cyberball game. Participants first completed the neuroticism measure and then
played the Cyberball game. After playing Cyberball, participants completed the
MACL, RSES, Appraisal scale, and CITS-S in that order. After completing all
measures participants were thoroughly debriefed as to the actual purpose of the
experiment.
11 It should be noted that the disadvantage of expanding the response scales of the UMACL and CITS-S is that the data can no longer be compared directly with norms or with other related studies.
95
Results
A median split was conducted on the neuroticism scale in order to classify
participants as either low neuroticism or high neuroticism (median neuroticism score
= 20). Mean neuroticism scores did not differ significantly between the excluded
and included conditions, F(1, 91) = .09, p = .770. A series of MANCOVAs were
conducted in order to examine appraisal, mood and self-esteem, and coping as a
function of Condition (excluded vs. included) and Neuroticism Level (low vs. high).
Bonferroni correction was implemented to set critical α level for all univariate
analyses. Gender was again entered as a covariate in all analyses12 (Hankin &
Abramson, 2001).
Cyberball Appraisal
After controlling for gender a significant multivariate effect of Condition was
obtained; λ = .54, F(2, 87) = 36.81, p < .001, partial η2 = .46. The multivariate
effect of Neuroticism Level; λ = .99, F(2, 87) = .63, p = .537, partial η2 = .01; and
the multivariate interaction between Condition and Neuroticism Level; λ = 1.00,
F(2, 87) = .01, p = .999, partial η2 = .00, were non-significant. Critical α level was
set at .025 (after Bonferroni correction) for all univariate analyses. Ostracized
participants felt significantly more excluded and perceived themselves to have less
control during the game than included participants (see Table 5). No significant
12 A significant gender difference was obtained with regard to avoidance coping (see page 96); however, the neuroticism-related results were not changed if gender was removed as a covariate.
96
effects of Neuroticism Level were obtained (see Table 6) and neither of the
univariate interactions between Condition and Neuroticism Level was significant.
Table 5. Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being included or excluded during the Cyberball game in Study 3
* Significant after Bonferroni correction
Mood and Self-Esteem
After controlling for gender a significant multivariate effect of Condition was
obtained; λ = .77, F(4, 85) = 8.26, p < .001, partial η2 = .23. The multivariate effect
of Neuroticism Level was approaching statistical significance; λ = .91, F(4, 85) =
Mean SD F df Error df p Partial η2 Appraised Exclusion: Included Excluded
16.11 35.02
11.53 11.52
60.53
1
88
< .001*
.41 Appraised Control: Included Excluded
16.34 6.24
7.20 5.41
55.04
1
88
< .001*
.39 Energetic Arousal: Included Excluded
45.18 39.35
12.85 13.24
4.17
1
88
.044
.05 Tense Arousal: Included Excluded
24.14 34.10
13.29 16.66
9.44
1
88
.003*
.10 Hedonic Tone: Included Excluded
95.77 67.65
19.73 31.13
24.90
1
88
< .001*
.22 RSES: Included Excluded
78.27 67.06
15.30
21.21
11.18
1
88
.001*
.11 T-F Coping: Included Excluded
40.27 42.16
12.92 13.31
.61
1
88
.438
.01 E-F Coping: Included Excluded
22.18 29.08
14.95 16.14
5.08
1
88
.027
.06 Avoidance Coping: Included Excluded
28.70 33.04
7.89
10.72
2.92
1
88
.091
.03
97
2.05, p = .094, partial η2 = .09. The multivariate interaction between Condition and
Neuroticism Level was not statistically significant; λ = .95, F(4, 85) = 1.19, p =
.323, partial η2 = .05. Critical α level was set at .013 (after Bonferroni correction)
for all univariate analyses. With regard to the three mood subscales, excluded
individuals reported significantly higher levels of tense arousal and significantly
lower levels of hedonic tone (although approaching significance the difference in
energetic arousal was not significant after correction, see Table 5). No neuroticism-
related differences in mood were obtained (see Table 6) and none of the univariate
interactions between Condition and Neuroticism Level were statistically significant.
Additionally, individuals allocated to the excluded condition reported significantly
lower levels of self-esteem in comparison with individuals in the included condition
(see Table 5). After correction, no neuroticism-related difference in self-esteem was
obtained (although the difference was approaching statistical significance, see Table
6) and the univariate interaction between Condition and Neuroticism Level with
regard to self-esteem was not statistically significant.
Coping
A significant multivariate effect of gender was obtained λ = .88, F(3, 86) =
3.95, p = .011, partial η2 = .12. Critical α was set at .017 for all univariate analyses.
Males engaged in significantly more avoidance coping than females; F(1, 88) =
10.59, p = .002, partial η2 = .11. No gender differences in task-focused; F(1, 88) =
.08, p = .773, partial η2 = .00; or emotion-focused; F(1, 88) = .01, p = .944, partial
98
η2 = .00; coping were obtained. After controlling for gender the multivariate effects
of Condition; λ = .94, F(3, 86) = 2.00, p = .119, partial η2 = .07; and Neuroticism
Table 6. Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 3
* Significant after Bonferroni correction
Level; λ = .93, F(3, 86) = 2.29, p = .084, partial η2 = .07; were approaching
significance. The multivariate interaction between Condition and Neuroticism Level
was not significant; λ = .98, F(3, 86) = .62, p = .605, partial η2 = .02. No
significant differences in coping as a function of inclusion/exclusion status were
obtained (although the difference in both emotion-focused and avoidance coping
Mean SD F df Error df p Partial η2 Appraised Exclusion: Low N High N
26.78 25.37
14.05 15.81
.15
1
88
.705
.00 Appraised Control: Low N High N
10.00 12.07
8.32 7.77
1.19
1
88
.278
.01 Energetic Arousal: Low N High N
41.85 42.37
12.54 14.19
.01
1
88
.972
.00 Tense Arousal: Low N High N
30.15 28.61
15.51 16.39
.09
1
88
.760
.00 Hedonic Tone: Low N High N
82.79 79.09
29.64 30.13
.93
1
88
.337
.01 RSES: Low N High N
76.77 67.87
20.33 17.46
4.37
1
88
.040
.05 T-F Coping: Low N High N
39.98 42.59
13.30 12.89
.70
1
88
.405
.01 E-F Coping: Low N High N
22.11 29.61
13.97 16.95
5.99
1
88
.016*
.06 Avoidance Coping: Low N High N
30.40 31.59
10.06 9.37
3.03
1
88
.085
.03
99
was approaching significance, see Table 5). High neuroticism participants engaged
in significantly more emotion-focused coping than low neuroticism participants,
regardless of which condition they were assigned to (see Table 6). None of the
univariate interactions between Condition and Neuroticism Level were statistically
significant.
Mediational Analyses
Although a neuroticism-related difference in emotion-focused coping was
obtained, no neuroticism-related differences in appraisal of the Cyberball game were
found. Therefore analyses assessing whether stressor appraisals mediate the
relationships between neuroticism and emotion-focused coping were not conducted.
Discussion
Study 3 expanded the range of response scales in order to determine whether
small neuroticism-related differences in appraisal and coping could be obtained in
the context of the Cyberball game. With regard to being ostracised, results largely
replicated the findings of Study 2. Ostracized participants felt more excluded and
perceived themselves as having less control during the Cyberball game in
comparison with included participants. Furthermore, being ostracized was again
associated with significantly higher levels of tense arousal and significantly lower
levels of hedonic tone and self-esteem. As in Study 2, the difference in energetic
arousal as a function of inclusion/exclusion status was approaching significance.
100
These findings replicate those of Study 2, and provide additional support for
Williams and colleagues’ contention that being excluded during the Cyberball game
is associated with deleterious outcomes (Williams & Sommer, 1997; Zadro et al.,
2004). Again, this is the case even when widely-used and validated outcome
measures (such as the MACL and the RSES) are used. Importantly, the current
findings, in conjunction with the results of Study 2, reveal that tense arousal and
hedonic tone are the facets of mood (at least as measured by the MACL) that are
particularly influenced by the social ostracism experienced during the Cyberball
game. No significant differences in coping were obtained as a function of Cyberball
condition after Bonferroni correction was implemented. However, the effect of
Cyberball condition on both emotion-focused coping and avoidance coping was
approaching statistical significance (the p value for the effect of condition on
emotion-focused coping was .027; the p value for the effect of condition on
avoidance coping was .091). Tentatively, these results suggest that being excluded
during the Cyberball game may evoke both an emotion-focused coping response (as
previously reported in Study 2) as well as an avoidance coping response. However,
further studies are clearly needed to document this conclusively.
Even using expanded response scales in Study 3, no neuroticism-related
differences in appraisals of the Cyberball game were obtained. Again, this was
unexpected as previous research (including Study 1 in this thesis) suggests that
neuroticism is consistently associated with negative stressor appraisals (e.g.
Engelhard, 2007; Gallagher, 1990; Schneider, 2004). However, there is some
evidence that high neuroticism individuals are particularly likely to interpret
101
ambiguous stimuli in a negative or threatening manner (Watson & Clark, 1984).
The fact that no neuroticism-linked differences in appraisal were obtained in either
Study 2 or Study 3 may have been because the degree of social exclusion
experienced in the excluded condition was so unequivocal that it was not an
ambiguous situation. Similarly, there is no ambiguity in the included Cyberball
condition. Participants in the included condition experienced no social exclusion at
all. Therefore, in both Study 2 and Study 3, neither of the Cyberball conditions was
in the least bit ambiguous. It may be the case that neuroticism-related differences in
appraisals during the Cyberball game will be obtained when the social ostracism
experienced is more ambiguous in nature. Study 4 introduced an ambiguous
Cyberball condition in order to explore this idea in more detail.
A neuroticism-related difference in coping was obtained in Study 3.
Specifically, high neuroticism participants reported engaging in more emotion-
focused coping than low neuroticism participants; regardless of which Cyberball
condition they were assigned to. However, it is important to note that due to
financial constraints the neuroticism scale used in Study 3 was not the same as that
used in Study 2 (although the two scales are highly correlated; Gow et al., 2005).
Therefore, it is possible that the neuroticism-related difference in emotion-focused
coping obtained in Study 3 could be attributed to the different neuroticism measure,
rather than the expanded response scale on the CITS-S. Study 4 addresses this issue
by returning to the NEO-FFI and retaining the expanded response scales of the
outcome measures. However, bearing this measurement issue in mind, neuroticism
was associated with emotion-focused coping in the context of the Cyberball game.
102
This finding is consistent with previous research using dispositional and
retrospective coping measures, the results obtained in Study 1, as well as findings
examining coping in the context of experimental cognitive tasks (Bouchard, 2003;
Carver et al., 1989; Endler & Parker, 1990b; R. Gomez et al., 1999; Matthews &
Campbell, 1998; Matthews et al., 2006). Importantly, the findings of Study 2
overcome the reliance on dispositional and retrospective coping measurement that
has limited much of the previous research examining the relationships between
personality traits and coping. Study 3 extends previous research by revealing that
neuroticism is associated with emotion-focused coping in the context of an
experimental social stressor, in which all participants experience exactly the same
scenario. No neuroticism-related differences in avoidance coping were obtained in
Study 3 and this is inconsistent with previous findings (e.g. Carver et al., 1989;
Matthews & Campbell, 1998; Matthews et al., 2006; Watson & Hubbard, 1996).
However, the fact that participants were not pre-screened for neuroticism-level, and
a median split was used to classify individuals as high or low neuroticism, may have
limited the power (Irwin & McClelland, 2003; MacCallum, Zhang, Preacher, &
Rucker, 2002) of Study 3 to identify small differences in avoidance coping. Study 4
(Chapter 7) screened potential participants on neuroticism and only the top and
bottom quartiles were invited to participate in the study. Alternatively, it may be the
case that neuroticism is simply not associated with avoidance coping in the context
of the Cyberball game.
Finally, as with Study 2, neuroticism was not associated with stressor
appraisals in either of the Cyberball conditions. This suggests that the relationship
103
between neuroticism and emotion-focused coping in the context of Cyberball is not
mediated by appraisals of exclusion or controllability; at least when the degree of
social exclusion experienced is unambiguous. Study 4 manipulated the extent to
which participants were excluded during the Cyberball game, in order to determine
whether neuroticism-related differences in appraisal and coping could be obtained
using when the social ostracism experienced during the Cyberball game is more
ambiguous in nature. If neuroticism-linked differences in both appraisal and coping
could be obtained in a more ambiguous exclusion condition, then Study 4 aimed to
determine whether stressor appraisals mediate the relationship between neuroticism
and coping.
104
CHAPTER 7: Study 4 – Introducing an Ambiguous Cyberball Condition
Aims and Hypotheses of Study 4
The fact that no neuroticism-related differences in appraisal of the Cyberball
game were obtained in either Study 2 or Study 3 was unexpected. Previous research,
(including Study 1 in this thesis) has consistently linked neuroticism with negative
appraisals of both the self and the environment (e.g. Engelhard, 2007; Gallagher,
1990; Mak et al., 2004; Schneider, 2004; Watson & Clark, 1984). Importantly,
some researchers have suggested that high neuroticism individuals are particularly
likely to appraise ambiguous stimuli in a negative or threatening manner (Watson &
Clark, 1984). Additionally, there is a substantial body of research reporting that trait
anxiety (which is strongly correlated with neuroticism, e.g. Scheier, Carver, &
Bridges, 1994) is associated with a negative interpretive bias in the processing of
ambiguous information (e.g. Hadwin, Frost, French, & Richards, 1997; Taghavi,
Moradi, Neshat-Doost, Yule, & Dagleish, 2000; Waters, Craske, Bergman, &
Treanor, 2008). This interpretive bias has been obtained using both ambiguous
vignettes (Barrett, Rapee, Dadds, & Ryan, 1996; Bogels & Zigterman, 2000; Dineen
& Hadwin, 2004; Muris, Rapee, Meesters, Schouten, & Geers, 2003; Waters et al.,
2008) as well ambiguous homophones, which have both a neutral and a threat
interpretation (e.g. dye versus die; Hadwin et al., 1997; Taghavi et al., 2000). With
regard to Cyberball, it is possible that the degree of social exclusion experienced in
the excluded conditions of traditional games is so unequivocal that it does not
105
represent an ambiguous situation (at least with regard to the degree of exclusion
experienced, as everyone is completely excluded). The overarching aim of Study 4
was to determine whether neuroticism-related differences in appraisals and coping in
the context of the Cyberball game could be obtained when the social ostracism
experienced is more ambiguous in nature (see the method section for a description of
the ambiguous Cyberball condition).
Based on previous ostracism research, as well as the results obtained in both
Study 2 and Study 3, it was predicted that participants allocated to the completely
excluded Cyberball condition would appraise the game more negatively than
participants allocated to the ambiguous condition. Additionally, it was predicted
that individuals who were completely ostracized during the game would report
worse mood and lowered self-esteem. Specifically with regard to mood, given the
findings obtained using the MACL (Matthews et al., 1990) in both Study 2 and
Study 3, it was predicted that excluded participants would report significantly lower
levels of hedonic tone and significantly higher levels of tense arousal in comparison
with participants allocated to the ambiguous condition. Finally, based on the
findings of Study 2, it was again predicted that being excluded during the game
would be associated with an emotion-focused coping response. Additionally, given
the findings obtained in Study 3 (where the differences in both emotion-focused and
avoidance coping as a function of inclusion/exclusion status were approaching
statistical significance), it was predicted that being excluded during Cyberball would
also be associated with an avoidance coping response.
106
With regard to neuroticism, it was predicted that in the ambiguous Cyberball
condition high neuroticism participants would appraise the game more negatively
than low neuroticism participants. Based on the finding obtained in Study 3
(although bearing in mind that Study 3 did not use the NEO-FFI), which revealed
that neuroticism was associated with more emotion-focused coping regardless of
inclusion/exclusion status, it was predicted that high neuroticism participants would
engage in significantly more emotion-focused coping than low neuroticism
participants regardless of whether they were assigned to the excluded or ambiguous
Cyberball condition. Additionally, although no neuroticism-related difference in
avoidance coping was obtained in Study 3, given the substantial body of literature
which has reported that neuroticism to be associated with avoidance coping, a small
neuroticism-related difference in avoidance coping was also predicted. Finally, if
neuroticism-linked differences in both appraisal and coping could be obtained in the
context of the ambiguous Cyberball game, then Study 4 aimed to examine whether
the relationship between neuroticism and coping with social exclusion is mediated
by appraisals of the Cyberball game.
Method
Participants
Four hundred and fifty-seven undergraduate psychology students were
screened on neuroticism and the top and bottom quartiles were invited to participate
in Study 4. Eighty-seven participants (mean age = 19.42 years) took part in the
107
study. Approximately equal numbers of high (n = 47) and low neuroticism (n = 40)
participants were recruited. All participants were randomly allocated to either the
excluded (n = 42) or ambiguous (n = 45) condition. The gender ratio of the sample
reflected that of the undergraduate psychology population (62 females; 25 males).
Materials
Cyberball – Introducing an Ambiguous Condition
Cyberball was reprogrammed so that the probability of the computer-
generated ‘players’ throwing to the participant could be manipulated13. Given that
the game consists of a triad of players, a probability of .5 is analogous to the
complete inclusion condition used in traditional Cyberball research (the
computerized ‘player’ with the ball is equally likely to throw the ball the participant
or the other ‘player’). Alternatively, a probability of 0 (the computerized ‘player’
with the ball will never throw the ball to the participant) is largely analogous to the
complete exclusion condition used in previous Cyberball research; with the
exception that in traditional Cyberball games the excluded participant is thrown the
ball twice at the start of the game before being ostracised for the rest of the game.
Altering the probability of the participant receiving the ball allows the researcher to
manipulate the extent to which participants are excluded during the Cyberball game.
Four probability conditions were piloted-tested in order to determine the
most ambiguous level of social ostracism. The most ambiguous condition was
defined as the probability condition with the most variation in the degree to which
13 Acknowledgements must be made to Dr. Martyn Churcher for programming the Cyberball game to allow the level of exclusion to be manipulated.
108
participants perceived themselves to be excluded during the game. Seventy-six
participants (54 female; 22 male) were randomly assigned to one of four probability
conditions (.1, .2, .3, or .5). Participants were all undergraduate or postgraduate
psychology students at the University of Western Australia. Participants played the
Cyberball game, and then completed the Appraisal (Excluded) subscale of the
Cyberball Appraisal Scale used in Study 2 and Study 3 (the extended response scale
used in Study 3 was employed in the pilot testing). Variation in Appraisal
(Excluded) scores was examined in each probability condition and a clear peak in
variance was observed at a probability level of .3 (see Figure 2). A probability of .3
was therefore determined to be the most ambiguous Cyberball condition and was
used as the ambiguous condition in Study 4. In Study 4 Cyberball was set for a total
of 50 throws.
.1 .2 .3 .50
50
100
150
200
Condition
Var
ianc
e
Figure 2: Variance in Appraisal (Excluded) scores in piloted probability conditions
109
Neuroticism Screening
The 10-item scale (see Appendix D) compiled from the International
Personality Item Pool (Goldberg et al., 2006) was again used to screen potential
participants on neuroticism. The neuroticism scale has an internal consistency of .86
(Goldberg et al., 2006) and correlates highly with the NEO-FFI neuroticism scale
(Gow et al., 2005).
NEO-Five Factor Inventory (NEO-FFI)
The NEO-FFI (Costa & McCrae, 1992) was administered at time of testing.
As mentioned previously (see Chapters 3 and 5) internal consistencies of the five
subscales range between .68 (Agreeableness) and .86 (Neuroticism; Costa &
McCrae, 1992). The neuroticism subscale (α = .86) was of particular interest in
Study 4.
Mood, Self-Esteem, Appraisal, and Coping
Mood, self-esteem, appraisal, and coping measures were the same as those
used in Study 2 and Study 3; however, extended 11-point response scales for the
MACL (0: Definitely; 10: Definitely Not), RSES (0: Strongly Agree; 10: Strongly
Disagree), Appraisal Scale (0: Not at All; 10: Extremely), and CITS-S (0: Not at All;
10: Extremely) were again used in Study 4. The internal consistencies of all
subscales (using expanded response scales) obtained in the sample from Study 4 are
summarised in Table 27 (see Appendix G).
110
Procedure
After reading an information sheet and signing a consent form all
participants were randomly allocated to either the excluded or ambiguous Cyberball
condition. Participants were informed that the experiment aimed to assess the
relationship between personality and mental visualization skills. Participants first
completed the NEO-FFI and then played the Cyberball game. After playing
Cyberball, participants completed the MACL, RSES, Appraisal scale, and CITS-S in
that order. After completing all measures participants were thoroughly debriefed as
to the actual purpose of the experiment.
Results
A significant difference in NEO-FFI neuroticism score was obtained between
the low neuroticism (mean = 16.26) and high neuroticism (mean = 26.17) groups at
time of testing; F(1, 86) = 66.20, p < .001. However, an examination of the
distribution NEO-FFI neuroticism scores revealed that rather than being bimodal
scores were normally distributed. Therefore, a median split was conducted (using
NEO-FFI neuroticism scores, median = 20) to classify individuals as low
neuroticism (n = 45) or high neuroticism (n = 42). No difference in mean NEO-FFI
neuroticism score was obtained between the excluded and ambiguous conditions;
F(1, 86) = 1.51, p = .22. A series of MANCOVAs was conducted in order to
examine appraisal, mood and self-esteem, and coping as a function of Condition
(excluded vs. ambiguous) and Neuroticism Level (low vs. high). Bonferroni
111
correction was implemented to set critical α level for all univariate analyses. Gender
was again entered as a covariate in all analyses14 (Hankin & Abramson, 2001).
Cyberball Appraisal
After controlling for gender a significant multivariate effect of Condition was
obtained; λ = .58, F(2, 81) = 29.94, p < .001, partial η2 = .43. The multivariate
effect of Neuroticism Level; λ = .97, F(2, 81) = 1.38, p = .258, partial η2 = .03; and
the multivariate interaction between Condition and Neuroticism Level; λ = .95, F(2,
81) = 2.16, p = .121 partial η2 = .05; were not significant. Critical α level was set at
.025 for all univariate analyses. Completely ostracised participants felt significantly
more excluded and perceived themselves as having less control during the Cyberball
game than participants allocated to the ambiguous condition (see Table 7). No
significant effects of Neuroticism Level were obtained (see Table 8); however, after
Bonferroni correction the interaction between Neuroticism Level and Condition with
regard to Appraisal (Control) was approaching significance; F(1, 81) = 4.49, p =
.043, partial η2 = .05. This interaction is illustrated in Figure 3. One-way ANOVAs
revealed that the neuroticism-related difference in appraised controllability was
significant in the ambiguous condition; F(1, 41) = 6.59, p = .014; but not in the
complete exclusion condition; F(1, 40) = .01, p = .921.
14 Removing gender as covariate did not change any of the reported results
112
Excluded Ambiguous0
10
20
30Low NHigh N
Ap
pra
isa
l (C
on
tro
l) S
co
re
Figure 3: Mean Appraised Control scores (and Standard Error) for high and low Neuroticism participants in the excluded and ambiguous Cyberball conditions Mood and Self-Esteem
After controlling for gender, significant multivariate effects of Condition; λ =
.70, F(4, 79) = 8.49, p < .001, partial η2 = .30; and Neuroticism Level; λ = .88, F(4,
79) = 2.74, p = .034, partial η2 = .12; were obtained. The multivariate interaction
between Condition and Neuroticism Level was not significant; λ = .97, F(4, 79) =
.57, p = .689, partial η2 = .03. Critical α level was set at .013 for all univariate
analyses.
113
Table 7. Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being assigned to the excluded or ambiguous condition during the Cyberball game in Study 4
* Significant after Bonferroni correction
With regard to mood, excluded participants reported significantly higher
levels of tense arousal and lower levels of hedonic tone than participants allocated to
the ambiguous condition (see Table 7). High neuroticism participants reported
significantly lower levels of hedonic tone regardless of which condition they were
assigned to (see Table 8). Although approaching significance, the neuroticism-
related difference in energetic arousal was not significant after correction. With
regard to self-esteem, excluded participants reported significantly lower levels of
Mean SD F df Error df p Partial η2 Appraised Exclusion: Excluded Ambiguous
42.71 25.49
10.43 13.76
40.90
1
82
< .001*
.33 Appraised Control: Excluded Ambiguous
12.36 20.93
2.93 7.49
47.38
1
82
< .001*
.37 Energetic Arousal: Excluded Ambiguous
43.88 46.53
12.70 14.11
.58
1
82
.449
.01 Tense Arousal: Excluded Ambiguous
32.86 23.84
18.17 12.81
6.41
1
82
.013*
.07 Hedonic Tone: Excluded Ambiguous
67.36 96.58
28.95 21.17
28.47
1
82
< .001*
.26 RSES: Excluded Ambiguous
63.43 78.53
20.00 15.42
14.97
1
82
< .001*
.15 T-F Coping: Excluded Ambiguous
39.32 36.61
8.91
11.88
1.34
1
82
.251
.02 E-F Coping: Excluded Ambiguous
25.16 12.61
13.24 8.42
27.88
1
82
< .001*
.25 Avoidance Coping: Excluded Ambiguous
25.88 21.05
8.96 7.66
6.81
1
82
.011*
.08
114
self-esteem. High neuroticism participants reported significantly lower levels of
self-esteem regardless of which condition they were assigned to. None of the
univariate interactions between Neuroticism Level and Condition were statistically
significant.
Table 8. Differences in appraisal (excluded), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 4
* Significant after Bonferroni correction
Coping
After controlling for gender the multivariate effects of Condition, λ = .74,
F(3, 80) = 9.61, p < .001, partial η2 = .27, and Neuroticism Level, λ = .87, F(3, 80) =
Mean SD F df Error df p Partial η2 Appraised Exclusion: Low N High N
31.44 36.33
14.73 14.95
1.75
1
82
.189
.02 Appraised Control: Low N High N
18.00 15.50
7.76 6.31
2.29
1
82
.134
.03 Energetic Arousal: Low N High N
48.11 42.19
13.42 12.92
4.73
1
82
.032
.06 Tense Arousal: Low N High N
25.31 31.29
14.62 17.35
2.20
1
82
.142
.03 Hedonic Tone: Low N High N
89.96 74.45
28.68 27.60
7.15
1
82
.009*
.08 RSES: Low N High N
77.20 64.86
19.13 17.39
9.75
1
82
. 002*
.11 T-F Coping: Low N High N
37.59 38.32
10.09 11.10
.04
1
82
.846
.00 E-F Coping: Low N High N
14.64 23.09
11.02 12.96
11.75
1
82
.001*
.13 Avoidance Coping: Low N High N
21.91 25.00
9.41 7.54
3.13
1
82
.081
.04
115
3.92, p = .012, partial η2 = .13, were significant. The multivariate interaction
between Condition and Neuroticism Level was not significant; λ = .99, F(3, 80) =
.05, p = .986, partial η2 = .00. Critical α level was set at .017 for all univariate
analyses. Excluded participants engaged in significantly more emotion-focused and
avoidant coping than participants allocated to the ambiguous condition (see Table 7).
High neuroticism participants engaged in significantly more emotion-focused coping
than low neuroticism participants, regardless of which condition they were assigned
to (see Table 8). The neuroticism-related difference in avoidance coping was
approaching significance. None of the interactions were statistically significant.
However, an examination of partial correlations (controlling for gender) between
neuroticism and avoidance coping in both Cyberball conditions revealed that
neuroticism was significantly correlated with avoidance coping in the ambiguous
condition (r = .34, see Table 9).
Mediational Analyses
Given that a neuroticism-related difference appraised controllability was
obtained in the ambiguous condition, a neuroticism-related difference was obtained
in emotion-focused coping regardless of condition, and neuroticism was also
correlated with avoidance coping in the ambiguous condition, it is possible that
appraisals of the Cyberball game might mediate the relationship between
neuroticism and both emotion-focused and avoidance coping (particularly in the
ambiguous condition). A variable is a potential mediator (M) if it is significantly
associated with both the predictor variable (X) and the outcome variable (Y), and if
116
it significantly predicts Y when controlling for X. If M fully mediates the
relationship between X and Y, then the effect of X on Y should be reduced to non-
significance when controlling for M (Baron & Kenny, 1986). Partial correlations
(controlling for gender) between neuroticism, Appraisal (Excluded), Appraisal
(Control), task-focused coping, emotion-focused coping, and avoidance coping
obtained in each Cyberball condition are summarised in Tables 9 and 10.
Neuroticism was not significantly correlated with either of the appraisal subscales in
the excluded condition. In the ambiguous condition, the correlation between
neuroticism and Appraisal (Control) was approaching significance (p = .061).
However, as neuroticism was not significantly correlated with Appraisal (Excluded)
or Appraisal (Control), in either the excluded or the ambiguous Cyberball condition,
analyses examining the mediating effect of stressor appraisal on the relationship
between neuroticism and coping were not conducted.
Table 9. Partial correlations (controlling for gender) between neuroticism, appraisal (excluded), appraisal (control), task-focused (T-F), emotion-focused (E-F), and avoidance coping in the excluded Cyberball condition in Study 4
N Appraisal (Stress)
Appraisal (Control)
T-F Coping
E-F Coping
Avoidance Coping
N 1 .08 -.03 -.06 .43* .13 Appraisal (Excluded)
1
-.42*
-.03
.58*
.27*
Appraisal (Control)
1
.28*
-.08
-.14
T-F Coping
1
.05
-.26
E-F Coping
1
.46*
Avoidance Coping
1
* p < .05
117
Table 10. Partial correlations (controlling for gender) between neuroticism, appraisal (excluded), appraisal (control), task-focused (T-F), emotion-focused (E-F), and avoidance coping in the ambiguous Cyberball condition in Study 4
N Appraisal
(Stress) Appraisal (Control)
T-F Coping
E-F Coping
Avoidance Coping
N 1 .16 -.24 .09 .37* .34* Appraisal (Excluded)
1
-.52*
.03
.72*
.33*
Appraisal (Control)
1
.28*
-.43*
-.43*
T-F Coping
1
.08
-.22
E-F Coping
1
.46*
Avoidance Coping
1
* p < .05
Discussion
Study 4 introduced an ambiguous Cyberball condition in order to determine
if small neuroticism-related differences in appraisal and coping could be observed in
the context of the Cyberball game. It was predicted that neuroticism-related
differences in appraisal would be obtained when the social ostracism experienced
was more ambiguous in nature. With regard to being completely ostracized, results
replicated the findings of Study 2 and Study 3. Completely ostracized participants
felt significantly more excluded than individuals allocated to the ambiguous
Cyberball condition. Excluded individuals also perceived themselves as having less
control during the Cyberball game. As in both Study 2 and Study 3, being excluded
was associated with significantly higher levels of tense arousal, as well as
118
significantly lower levels of hedonic tone and self-esteem. However, in both of the
previous Cyberball studies, the difference in energetic arousal as a function of
inclusion/exclusion condition was approaching statistical significance. This was not
the case in Study 4, where the comparison group (i.e. individuals in the ambiguous
condition) also experienced some ostracism during the game. Taken together the
findings of Studies 2, 3 and 4 suggest that the facets of mood (as measured by the
MACL; Matthews et al., 1990) most strongly affected by the social exclusion
experienced during Cyberball are tense arousal and hedonic tone. The current series
of Cyberball studies add considerable weight to Williams and colleagues (Williams
& Sommer, 1997; Zadro et al., 2004) contention that being excluded during the
Cyberball game has measurable negative outcomes. Importantly, this effect appears
to be robust and emerges even when widely used outcome measures, such as the
MACL and the RSES, are used and the comparison group is partially excluded
during the game. This is consistent with previous research that reports the negative
effects of ostracism may increase when the social exclusion is obvious rather than
ambiguous (e.g. Snoek, 1962).
Individuals allocated to the excluded condition engaged in significantly more
emotion-focused and avoidance coping. The finding that being excluded resulted in
an emotion-focused coping response is consistent with Study 2. However, the
results of Study 4 revealed that being excluded during the Cyberball game is also
associated with an avoidance coping response. Additionally, in Study 3 the effect of
both emotion-focused and avoidance coping as a function of inclusion/exclusion
status was approaching statistical significance. Taken together, the results of Studies
119
2, 3 and 4 reveal that being excluded during the Cyberball game can evoke both an
emotion-focused and an avoidance coping response. The fact that Cyberball is
consistently associated with worsened mood and lowered self-esteem, as well as
evoking an emotion-focused and avoidance coping response, is strong evidence for
its utility as a laboratory-stressor with which to examine individual differences in
coping with social stress. As mentioned previously, identifying useful laboratory-
stressors is important as they overcome many of the contextual confounds associated
with dispositional and retrospective methodologies, and allow for greater control of
the variables under investigation (Connor-Smith & Flachsbart, 2007). Given the
immense difficulties in controlling variables in real-world social settings, the current
findings highlight the utility of the Cyberball game in this context.
Interestingly, the interaction between Neuroticism Level and Condition in
relation to appraised controllability was approaching statistical significance (after
Bonferonni correction). More specifically, as predicted a neuroticism-related
difference in appraised controllability was only observed in the ambiguous
Cyberball condition. This finding is consistent with research that suggests that
individuals high in neuroticism (and related traits such as negative affectivity and
trait anxiety) appraise ambiguous stimuli in a negative or threatening manner
(Watson & Clark, 1984). This result may have important implications for Cyberball
research, and social ostracism research in general. Previous research using the
Cyberball game has identified no individual difference variables or experimental
manipulations that moderate the effects of being ostracised during traditional
Cyberball games (e.g. Eisenberger et al., 2003; Gonsalkorale & Williams, 2006;
120
Zadro et al., 2004). Ambiguous Cyberball conditions offer exciting opportunities to
examine individual differences in sensitivity to social ostracism.
With regard to coping, results from Study 4 revealed that high neuroticism
participants engaged in significantly more emotion-focused coping than low
neuroticism participants; regardless of which Cyberball condition they were
assigned to. Importantly, taken together the findings of Study 3 and Study 4 suggest
that neuroticism-related differences in emotion-focused coping can be obtained
using both the NEO-FFI and the neuroticism scale compiled from the International
Personality Item Pool (Goldberg et al., 2006). Although no interaction between
Condition and Neuroticism Level (with regard to avoidance coping) was obtained in
Study 4, neuroticism scores were significantly correlated with avoidance coping in
the ambiguous condition. This finding extends the results of Study 3. Taken
together, the results of Study 3 and 4 reveal that neuroticism is associated with an
emotion-focused coping response; regardless of whether participants are completely
excluded, partially excluded (i.e. in the ambiguous condition) or completely
included during the Cyberball game. Neuroticism also appears to be associated with
avoidance coping when the social exclusion experienced is ambiguous in nature,
although more research is needed to document this conclusively. These
relationships should be explored using laboratory-stressors in other stress domains.
Overall, the findings of Study 4 are consistent with Study 1, and with
previous research using dispositional and retrospective coping measures reporting
that neuroticism is associated with both emotion-focused and avoidance coping
strategies (e.g. Vollrath & Torgersen, 2000). Similarly, these results are consistent
121
with research reporting a relationship between neuroticism and both emotion-
focused and avoidance coping in the context of experimental cognitive tasks
(Matthews & Campbell, 1998; Matthews et al., 2006). Importantly, the findings of
Study 3 and Study 4 extend the research literature and reveal that neuroticism is
associated with emotion-focused and avoidance coping in the context of an
experimental social stressor. No significant correlations between neuroticism and
appraisals of the Cyberball game were obtained in Study 4, and therefore no
meditational analyses were conducted. However, it should be noted that the negative
correlation between neuroticism and appraised controllability was approaching
significance in the ambiguous condition (consistent with the interaction obtained in
the MANCOVA). This finding warrants further investigation and future research
should explore the potential mediating influence of stressor appraisals (in particular
controllability appraisals) on the relationship between neuroticism and both
emotion-focused and avoidance coping in the context of stressful situations in which
control appraisals can be manipulated. Study 5 (Chapter 9) employs an anagram-
solving task that has previously been used to evaluate the goodness of fit hypothesis
(Endler, Macrodimitris, & Kocovski, 2000; Endler, Speer, Johnson, & Flett, 2000),
as an experimental stressor with which to explore this issue. Chapter 8 summarises
the findings obtained in Studies 2, 3, and 4 using the Cyberball game.
122
CHAPTER 8: Summary of Cyberball Studies
As mentioned in Chapters 2 and 3, much of the previous research examining
the relationship between personality-related variables and coping has been limited
by reliance on dispositional and retrospective coping measures. David and Suls
(1999) argue that obtaining coping reports immediately after a stressful event has
occurred would allow for firmer conclusions regarding the relationship between
personality traits and coping to be made. Additionally, Connor-Smith and
Flachsbart (2007) have recently concluded that the impact of personality on coping
could best be identified through the use of standardised laboratory-stressors. For a
variety of reasons (see Chapter 4 for a discussion of these reasons), the Cyberball
game (Williams & Jarvis, 2006) appeared a promising laboratory-stressor with
which to examine individual differences in coping with a novel social stressor.
The current series of Cyberball studies had three major aims. Firstly, they
aimed to replicate the previously reported negative effects of being socially excluded
during the Cyberball game (e.g. Williams & Sommer, 1997; Zadro et al., 2004)
using well-known and well-validated outcome measures. Secondly, the current
experiments aimed to determine whether the ball-throwing game could evoke a
coping response. If being excluded during the Cyberball game can evoke both a
negative affective response and a coping response, then Cyberball is likely to be a
valuable laboratory-based stressor with which to examine individual differences in
coping with social exclusion. Finally, Studies 2, 3 and 4 all aimed to determine
whether neuroticism-related differences in appraisal and coping could be obtained in
123
the context of the Cyberball game. Importantly, as all participants experienced
exactly the same social stressor, and self-reports could be obtained immediately after
the conclusion of the Cyberball game, it was thought that this would allow firmer
conclusions to be made regarding neuroticism-related differences in appraisal and
coping.
Aim 1: The Cyberball Game, Appraisals, Mood and Self-Esteem
In all of the current Cyberball experiments, participants who were ostracised
during the game felt significantly more excluded than individuals allocated to either
the included or ambiguous conditions. Additionally, excluded participants perceived
themselves as have\ing less control during the Cyberball game than participants
allocated to the included or ambiguous condition. These results reveal that in all of
the current Cyberball studies the ostracism manipulations were successful.
Importantly, the finding that being excluded during the game is associated with
negative appraisals is completely consistent with previous ostracism research using
the ball-throwing paradigm, and the Cyberball game in particular (e.g. Williams &
Sommer, 1997; Zadro et al., 2004). The current findings therefore confirm that
being excluded during the Cyberball game does evoke feelings of social ostracism.
In relation to mood, the results obtained in the current Cyberball studies
revealed that being excluded during the game was consistently associated with
worsened mood. Specifically, in all three Cyberball studies excluded participants
reported significantly lower levels of hedonic tone as well as significantly higher
124
levels of tense arousal in comparison with participants allocated to either the
included or ambiguous conditions. Importantly, these findings were obtained using
the MACL (Matthews et al., 1990), which is a widely used and well validated mood
measure. The current findings therefore extend previous research that reports a
negative impact of social ostracism on general mood (e.g. Williams & Sommer,
1997). Specifically, the current findings reveal that hedonic tone and tense arousal
are the facets of mood (as measured by the MACL) that are most strongly impacted
by the social exclusion experienced during the Cyberball game. Future research
should examine the effect of social exclusion on mood using other
conceptualisations and measures of mood; for example the distinction between
Positive Affect and Negative Affect (Watson & Clark, 1988).
With regard to self-esteem, being excluded during the Cyberball game was
consistently associated with lower levels of self-esteem. Importantly, these findings
were obtained using the RSES (Rosenberg, 1965). The finding that social exclusion
has a negative impact on self-esteem is robust and fully consistent with previous
ostracism research using a short, three-item self-esteem scale designed by Williams
and his colleagues (e.g. Williams & Sommer, 1997; Zadro et al., 2004). Studies 2,
3, and 4 all confirm that being excluded during the Cyberball game results in
reductions in self-esteem; even when a widely-used and well validated self-esteem
measure is employed. Taken together the findings obtained in the current series of
Cyberball studies add considerable weight to Williams and colleagues (Williams &
Sommer, 1997; Zadro et al., 2004) contention that being excluded during the
Cyberball game has measurable negative outcomes. Importantly, this negative
125
ostracism effect appears to be robust and emerges even when widely used outcome
measures (such as the MACL and the RSES) are used and the comparison group is
also partially excluded during the ball-throwing game.
Aim 2: Cyberball as a Laboratory-Stressor
For Cyberball to be considered a useful laboratory-stressor it needs to be
documented that, as well as producing a negative affective response, being excluded
during the game can also evoke a coping response. In Studies 2 and 4 participants
who were excluded during the Cyberball game engaged in significantly more
emotion-focused coping than participants in the included or ambiguous conditions.
In Study 4 excluded participants also engaged in significantly more avoidance
coping than participants in the included or ambiguous conditions. Additionally, in
Study 3 the effect of inclusion/exclusion status on both emotion-focused coping and
avoidance coping was approaching statistical significance (after Bonferroni
correction). Taken together, the results of Studies 2, 3, and 4 therefore reveal that
being excluded during the game can produce both an emotion-focused and an
avoidance coping response.
The fact that Cyberball is consistently associated with worsened mood and
lowered self-esteem, and can evoke both an emotion-focused and avoidance coping
response, is strong evidence for its utility as a laboratory-stressor. Additionally, the
fact that all participants experience exactly the same scenario and can complete all
measures immediately after playing the game makes Cyberball a useful stressor with
126
which to examine individual differences in coping with social stress. Given the
immense difficulties in controlling variables in real-world social settings, the current
findings highlight the utility of the Cyberball as an experimental social stressor.
Aim 3: Neuroticism, Appraisal, and Coping in the Context of Cyberball
No neuroticism-related differences in appraisal of the Cyberball game were
obtained in either the excluded or included conditions. This was unexpected as
previous research has consistently linked neuroticism with negative appraisals of
both the self and the environment (e.g. Engelhard, 2007; Gallagher, 1990; Mak et
al., 2004; Schneider, 2004; Watson & Clark, 1984). However, when the social
ostracism experienced during the game was more ambiguous in nature, high
neuroticism participants tended to perceive themselves as having less control over
the game than low neuroticism participants. This finding is consistent with research
suggesting that individuals who score highly in neuroticism (and related variables
such as negative affectivity and trait anxiety) appraise ambiguous information or
situations in a negative or threatening manner (Watson & Clark, 1984). This finding
has implications for general social ostracism research, as previous research using the
traditional Cyberball game has identified no individual difference variables or
experimental manipulations that moderate the effects of being ostracised in
traditional Cyberball games (Zadro et al., 2004). An ambiguous Cyberball condition
offers exciting opportunities to examine individual differences in sensitivity to social
ostracism, as well as the impact of experimental manipulations (for example
127
manipulating the ingroup/outgroup status of the computer-generated ‘players’) on
the outcomes associated with being socially excluded during the game.
When expanded response scales were implemented in Studies 3 and 4, high
neuroticism participants engaged in significantly more emotion-focused coping than
low neuroticism participants (these findings were obtained using two different
neuroticism measures). This was the case regardless of whether participants were
completely excluded, partially excluded (i.e. in the ambiguous condition), or
completely included during the Cyberball game. Additionally, in the ambiguous
Cyberball condition neuroticism was significantly correlated with avoidance coping.
These findings are consistent with Study 1 and previous research reporting that
neuroticism is positively correlated with avoidance coping and associated with a
general reliance on emotion-focused coping (e.g. Matthews & Campbell, 1998;
Matthews et al., 2006; Vollrath & Torgersen, 2000). Importantly, the current
findings overcome the limitations associated with dispositional and retrospective
coping measurement. However, it should be noted that the relevance of personality
traits to stress responses and coping may vary substantially across different types of
stressors. These findings should be replicated in other stress domains (see Study 5).
The current Cyberball studies found no evidence that the relationship
between neuroticism and emotion-focused coping is mediated by appraisals of the
Cyberball game, although future research should investigate the possibility that
controllability appraisals mediate the relationship between neuroticism and both
emotion-focused and avoidance coping in ambiguous situations. This issue should
be explored using laboratory tasks in which the level of control can be
128
experimentally manipulated. It was also concluded that the relationship between
neuroticism, appraisal, and both emotion-focused and avoidance coping should be
examined using experimental stressors in different stress domains. Study 5 (Chapter
9) employed an anagram-solving task in order to examine the relationships between
neuroticism, appraisal, and coping in the context of a task that is more cognitive in
nature.
Conclusions from the Cyberball Studies
The results obtained in the current series of studies reveal that being
excluded during the Cyberball game is consistently associated with increased tense
arousal, as well as lowered hedonic tone and self-esteem. Moreover, when
expanded response scales are used, being excluded during the Cyberball game
appears to evoke both an emotion-focused and avoidance coping response. Taken
together, these findings suggest that Cyberball is a useful laboratory-stressor with
which to examine individual differences in coping with social stress. Additionally,
the current findings revealed that neuroticism is consistently associated with more
emotion-focused coping during the Cyberball game. This is the case regardless of
whether participants are completely excluded, partially excluded, or completely
included during the game. Neuroticism is also associated with avoidance coping
when the ostracism experienced is ambiguous. Importantly, the current studies
overcome the reliance on dispositional and retrospective coping measures that has
limited much of the previous research in this area. These results are consistent with
129
previous research that reports neuroticism is correlated with avoidance coping and
emotion-focused coping. However, these findings should be replicated in other
stress domains.
130
CHAPTER 9: Study 5 – Neuroticism and Coping in the Context of an
Anagram-solving Task
Aims and Hypotheses of Study 5
Some researchers have suggested that neuroticism is associated with a
general reliance on emotion-focused and avoidance coping (e.g. Vollrath &
Torgersen, 2000; Vollrath et al., 1995); however, these conclusions have been
limited by a reliance on dispositional or retrospective coping measurement. The
results obtained in Study 3 and Study 4, which used Cyberball as an experimental
social stressor, revealed that high neuroticism individuals do report engaging in
more emotion-focused coping than low neuroticism participants. Moreover, this is
the case regardless of whether participants are completely excluded, partially
excluded, or completely included during the Cyberball game. Additionally,
neuroticism scores were associated with avoidance coping in the ambiguous
Cyberball condition. These findings are consistent with previous research
suggesting that neuroticism is correlated with both emotion-focused and avoidance
coping. Importantly, Cyberball is an analogue of a stressful social situation. There
is some experimental evidence that neuroticism is also associated with both
emotion-focused and avoidance coping in the context of a variety of other tasks
(Matthews & Campbell, 1998; Matthews et al., 2006; Penley & Tomaka, 2002).
However, these tasks are not widely known or used in the coping literature.
131
In a series of experiments examining task-induced stress and individual
differences in coping, Matthews and colleagues (1998) reported significant
correlations between neuroticism and both emotion-focused and avoidant coping
with a number of specific activities. These activities included rapid information
processing, mental arithmetic, working memory, and driving simulation. More
recently, Matthews and colleagues (2006) replicated these findings using mental
arithmetic, driving simulation, and impossible anagrams as experimental stressors.
Similarly, Penley and Tomaka (2002) reported correlations between neuroticism and
emotion-focused coping in the context of performing a speech task. These data
support previous findings obtained using dispositional and retrospective coping
measures, and appear to be a promising start in understanding the link between
neuroticism and coping with experimental cognitive stressors.
Recently Endler and his colleagues (Endler, Macrodimitris et al., 2000;
Endler, Speer et al., 2000) have used an anagram-solving task (although, in this task
all anagrams are solvable) to examine stressor appraisals, situation-specific coping,
and the goodness of fit hypothesis. As yet, this task has not been used to examine
any personality-related differences in coping. An anagram-solving task appears
ideal as a laboratory-based stressor for examining neuroticism-linked differences in
appraisal and coping for a number of reasons. Firstly, the anagram-solving task has
been used successfully in previous coping research examining the goodness of fit
hypothesis. Importantly, this research has revealed that manipulating aspects of the
task can influence appraised controllability (Endler, Macrodimitris et al., 2000).
Secondly, using an anagram-solving task minimizes differences in experience or
132
knowledge (Endler, Macrodimitris et al., 2000). Thirdly, the anagram-solving task
represents a controlled stressful situation, in which all participants are faced with
exactly the same scenario. This should provide clarity in interpreting individual
differences in coping. Finally, by responding to all measures immediately after
completing the anagram-solving task, the limitations of dispositional and
retrospective self-reporting are minimized.
Study 5 had three specific aims. The first aim was to empirically evaluate an
anagram-solving task as an experimental stressor. In Study 5, the difficulty of the
anagrams and participants’ level of control were manipulated across two
experimental conditions. It was predicted that, if the anagram-solving task is a
useful experimental stressor, then individuals allocated to the High Stress anagram
condition (see method section for a description of the Mild/High Stress anagram
conditions) should appraise the task more negatively than individuals allocated to
the Mild Stress condition. Additionally, it was predicted that individuals allocated to
the High Stress condition would experience a negative affective response.
Specifically, based on the Cyberball findings it was predicted that individuals in the
High Stress condition would report worse mood and lowered self-esteem in
comparison with individuals in the Mild Stress condition. Finally, if the anagram-
solving task is an effective experimental stressor, then being assigned to the High
Stress condition should evoke a bigger coping response.
With the proviso that the anagram-task was determined to be a useful
laboratory-based stressor, the second aim of Study 5 was to determine if
neuroticism-linked differences in appraisal and situation-specific coping could be
133
obtained. In accordance with Study 1, Study 4, and previous research (e.g.
Engelhard, 2007; Gallagher, 1990; Mak et al., 2004; Schneider, 2004), high
neuroticism participants were predicted to appraise the anagram-solving task more
negatively than low-neuroticism participants. With regard to coping, based on
previous research using retrospective and dispositional coping measures (Bouchard,
2003; Carver et al., 1989; R. Gomez et al., 1999), the laboratory-based findings of
Matthews and colleagues (Matthews & Campbell, 1998; Matthews et al., 2006), and
the results obtained in Study 1, Study 3 and Study 4, it was predicted that
neuroticism-related differences in emotion-focused and avoidance coping would be
obtained.
According to the goodness of fit and matching hypotheses (see Chapter 1)
stressor appraisals influence the relationship between the stressful situation and
coping strategy choice. Given that neuroticism is associated with negative
appraisals of the environment (e.g. Engelhard, 2007; Gallagher, 1990; Mak et al.,
2004; Schneider, 2004), it may be the case that the relationship between neuroticism
and emotion-focused/avoidance coping is at least partially mediated by stressor
appraisals. If neuroticism-related differences in both appraisals of the anagram-
solving task and coping strategy are obtained, then the final aim of Study 5 was to
determine whether stressor appraisals mediate the relationship between neuroticism
and self-reported coping.
134
Method
Participants
Eighty-one participants aged between 18 and 31 years (mean = 21.32 years)
took part in Study 5. This study was conducted partly in conjunction with a group of
undergraduate students who were completing a research design and methodology
unit. Due to this, pre-screening participants for neuroticism level was beyond the
scope of this study. Participants were undergraduate students. Approximately equal
numbers of male (n = 42) and female (n = 39) participants were recruited. All
participants were randomly allocated to either a Mild (n = 40) or High (n = 41)
Stress condition.
Materials
Anagrams
Anagrams were sourced from Tresselt and Mayzner (1966), who provide
normative solution times for a sample of 134 solution words and 378 associated
anagrams. Study 5 had two experimental conditions. Individuals allocated to the
Mild Stress condition attempted six anagrams (see Table 11) with median solve
times of 31 seconds or less. In an attempt to maximise perceived control
participants in this condition were given as long as they needed to solve the
anagrams, were provided with a pen and paper to assist them in the task, and did not
have to complete the anagrams in the order provided. Individuals allocated to the
High Stress condition attempted six anagrams with median solve times of 120
135
seconds or more (see Table 15). In order to minimise perceived control, participants
in this condition were given only 30 seconds to solve the anagrams, were not
allowed to use pen or paper to assist them with the task, and had to complete the
anagrams in the order provided. All anagrams had only a single correct solution and
in the High Stress condition participants were shown the solution before proceeding
to the next anagram. In the Mild Stress condition participants were shown the
solutions after completing all anagrams.
Table 11. Anagrams and solutions (with median solve times taken from Tresselt & Mayzner, 1966) used in the Mild and High Stress conditions of Study 5
Anagram Solution Median Solve Time (seconds)
Mild Stress
iuftr fruit 15 oewrp power 22 gawno wagon 26 uimcs rhtib
music birth
27 28
ciotn tonic 31 High Stress
oapnr apron 132 speua pause 143 dtuai audit 159 aebrl blare 225 nrcui incur > 240 glaei agile > 240
Neuroticism Measure
Given that this experiment was conducted in conjunction with undergraduate
students, financial constraints again prohibited the use of the NEO-FFI. Therefore,
136
the 10 item scale (see Appendix D) compiled from the International Personality Item
Pool (Goldberg et al., 2006) was used to measure neuroticism at time of testing.
This neuroticism scale correlates very highly with the NEO-FFI neuroticism scale (r
= .84; Gow et al., 2005). Additionally, results obtained in Study 3 and Study 4
revealed that neuroticism-related differences in emotion-focused coping can be
obtained when either the NEO-FFI or the current neuroticism scale is used. Items
were responded to on a five-point scale (0: Very inaccurate; 4: Very accurate). The
neuroticism scale has an internal consistency of .86 (Goldberg et al., 2006).
Anagram Appraisal Scale
A six item scale was constructed to measure participants’ appraisals of the
anagram-solving task (see Appendix H). Three items assessed participants stress
appraisals (e.g. ‘How stressed did you feel during the anagram task’, α = .93) and
three items measured perceived control (e.g. ‘How much influence do you think you
had in the anagram task’, internal consistency α = .71) during the anagram task.
Items were responded to on an 11-point scale (0: Not at all; 10: Extremely).
Other Measures
The MACL, RSES, and CITS-S (with 11-point response scales) were again
used to measure mood, self-esteem, and coping. The internal consistencies of all
subscales, based on the sample from Study 5, are summarised in Table 23 (see
Appendix G).
137
Procedure
After reading an information sheet and signing a consent form all
participants were randomly allocated to either the Mild or High Stress condition.
Participants were informed that the experiment aimed to assess the relationship
between personality and language ability. Participants first completed the
personality measure and then attempted the anagram-solving task. After finishing
the anagrams, participants completed the MACL, RSES, Anagram Task Appraisal
scale, and CITS-S in that order. After completing all measures participants were
thoroughly debriefed as to the actual purpose of the experiment.
Results
Neuroticism scores were normally distributed. A median split was
conducted on the neuroticism scale (median = 20) in order to classify individuals as
low neuroticism (n = 37) or high neuroticism (n = 44). No difference in mean
neuroticism scores was obtained between individuals assigned to the Mild and High
Stress conditions, F(1, 79) = .22, p = .637. Given that gender differences in
neuroticism have been reported (Hankin & Abramson, 2001), gender was entered as
a covariate in all analyses15. An ANCOVA was conducted in order to examine the
number of anagrams solved correctly as a function of Condition (mild vs. high
stress) and Neuroticism Level (low vs. high). A series of MANCOVAs were
conducted in order to examine appraisal, mood and self-esteem, and coping as a
15 Removing gender as covariate did not change any of the reported results
138
function of Condition and Neuroticism Level. Bonferroni correction was
implemented to set critical α level for all univariate analyses.
Number of Anagrams Solved Correctly
Descriptive statistics summarising the total number of anagrams solved by
high and low neuroticism participants in both anagram conditions are presented in
Table 12. Missing data mean that the degrees of freedom in these analyses do not
reflect those in all other analyses (n = 73 for this analysis). After controlling for
gender participants assigned to the Mild Stress condition solved significantly more
anagrams than participants assigned to the High Stress condition; F(1, 68) = 81.97, p
< .001; partial η2 = .55. The between-groups effect of Neuroticism Level; F(1, 68) =
.99, p = .323; partial η2 = .01; and the interaction between Condition and
Neuroticism Level; F(1, 68) = .23, p = .634; partial η2 = .003; were not statistically
significant.
Table 12. Mean number of anagrams solved (and standard deviations) as a function of Condition and Neuroticism (N) Level
Mean Number of
Anagrams Solved Standard Deviation
Mild Stress Low N 5.33 1.50 High N 5.89 .76
High Stress Low N 2.11 1.91 High N 2.37 1.71
139
Appraisal
A MANCOVA was conducted in order to examine the two appraisal sub-
scales [Appraisal (Stress); Appraisal (Control)] as a function of Condition (mild vs.
high stress) and Neuroticism Level (low vs. high). No significant multivariate or
univariate effects of gender were obtained. After controlling for gender significant
multivariate effects of Condition; λ = .71, F(2, 75) = 15.56, p < .001, partial η2 =
.29; and Neuroticism Level; λ = .85, F(2, 75) = 6.85, p = .002, partial η2 = .16; were
obtained. The multivariate interaction between Condition and Neuroticism Level
was not statistically significant; λ = .99, F(2, 75) = .50, p = .610, partial η2 = .01.
Critical α level was set at .025 (after Bonferonni correction) for all univariate
analyses. In comparison with participants in the Mild Stress condition, participants
in the High Stress condition perceived the anagram task to be significantly more
stressful and perceived themselves as having significantly less control during the
anagram task (see Table 13). High neuroticism participants perceived the anagram
task to be significantly more stressful, and perceived themselves as having
significantly less control during the anagram task than low neuroticism participants,
regardless of whether they were allocated to the Mild or High stress condition (see
Table 14). Neither of the univariate interactions between Condition and Neuroticism
Level was statistically significant.
140
Table 13: Differences in appraisal (stress), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of being allocated to the Mild or High stress Condition in Study 5
* Significant after Bonferroni correction
Mood and Self-Esteem
A MANCOVA was conducted in order to examine the three mood sub-scales
and self-esteem as a function of Condition (mild vs. high stress) and Neuroticism
Level (low vs. high). No significant multivariate or univariate effects of gender
were obtained. After controlling for gender significant multivariate effects of
Condition; λ = .64, F(4, 73) = 10.21, p < .001, partial η2 = .36; and Neuroticism
Level; λ = .84, F(4, 73) = 3.56, p = .010, partial η2 = .16; were obtained. The
multivariate interaction between Condition and Neuroticism Level was not
Mean SD F df Error df p Partial η2 Appraised Stress: Easy Hard
12.53 18.39
8.38 6.42
12.75
1
76
.001*
.14 Appraised Control: Easy Hard
19.73 13.07
5.29 5.52
31.10
1
76
< .001*
.29 Energetic Arousal: Easy Hard
49.13 36.83
10.42 14.44
18.71
1
76
< .001*
.20 Tense Arousal: Easy Hard
36.00 48.22
17.10 13.86
14.05
1
76
< .001*
.16 Hedonic Tone: Easy Hard
80.75 58.24
27.45 26.04
13.20
1
76
.001*
.15 RSES: Easy Hard
71.05 52.71
16.95 14.35
28.82
1
76
< .001*
.28 T-F Coping: Easy Hard
44.48 41.24
9.97
11.53
2.68
1
76
.105
.03 E-F Coping: Easy Hard
25.40 34.95
17.95 14.39
6.50
1
76
.013*
.08 Avoidance Coping: Easy Hard
27.88 29.61
7.80 8.88
.30
1
76
.586
.00
141
statistically significant; λ = .97, F(4, 73) = .49, p = .745, partial η2 = .03. Critical α
level was set at .012 (after Bonferroni correction) for all univariate analyses. With
regard to the three mood subscales, individuals allocated to the High Stress
condition reported significantly higher levels of tense arousal, and significantly
lower levels of energetic arousal and hedonic tone than participants in the Mild
Stress condition (see Table 13). After correction, no neuroticism-related differences
in mood were obtained (although the difference in tense arousal was approaching
statistical significance, see Table 14). Additionally, individuals allocated to the
High Stress condition reported significantly lower levels of self-esteem than
individuals in the Mild Stress condition (see Table 13). High neuroticism
participants reported significantly lower levels of self-esteem, regardless of which
condition they were assigned to (see Table 14). None of the univariate interactions
between Condition and Neuroticism Level were statistically significant.
Coping
A MANCOVA was conducted in order to examine the three coping sub-
scales as a function of Condition (mild vs. high stress) and Neuroticism Level (low
vs. high). No significant multivariate or univariate effects of gender were obtained.
After controlling for gender significant multivariate effects of Condition; λ = .87,
F(3, 74) = 3.79, p = .014, partial η2 = .13; and Neuroticism Level; λ = .84, F(3, 74) =
4.63, p = .005, partial η2 = .16; were obtained. The multivariate interaction between
Condition and Neuroticism Level was not statistically significant; λ = .98, F(3, 74) =
.54, p = .657, partial η2 = .02. Critical α level was set at .017 for all univariate
142
analyses. Individuals allocated to the High Stress condition reported engaging in
significantly more emotion-focused coping than participants in the Mild Stress
condition (see Table 13). High neuroticism participants reported engaging in
significantly more emotion-focused and avoidance coping than low neuroticism
participants, regardless of which condition they were assigned to (see Table 14).
None of the univariate interactions between Condition and Neuroticism Level were
statistically significant.
Table 14: Differences in appraisal (stress), appraisal (control), energetic arousal, tense arousal, hedonic tone, RSES, task-focused (T-F), emotion-focused (E-F), and avoidance coping as a function of Neuroticism Level in Study 5
* Significant after Bonferroni correction
Mean SD F df Error df p Partial η2 Appraised Stress: Low N High N
12.54 17.98
7.96 7.17
12.92
1
76
.001*
.15 Appraised Control: Low N High N
18.03 14.95
6.63 5.77
7.71
1
76
.007*
.09 Energetic Arousal: Low N High N
45.32 40.86
13.68 14.06
2.82
1
76
.097
.04 Tense Arousal: Low N High N
38.14 45.59
17.97 14.76
5.04
1
76
.028
.06 Hedonic Tone: Low N High N
74.97 64.64
30.23 27.16
3.59
1
76
.062
.05 RSES: Low N High N
67.70 56.77
17.43 17.34
11.91
1
76
.001*
.14 T-F Coping: Low N High N
43.35 42.41
11.89 10.00
.14
1
76
.706
.00 E-F Coping: Low N High N
24.00 35.48
15.21 16.51
11.97
1
76
< .001*
.14 Avoidance Coping: Low N High N
26.24 30.86
8.65 7.57
7.37
1
76
.008*
.09
143
Mediational Analyses
Given that neuroticism-related differences in both appraisal and coping were
obtained, the potential mediating effect of appraisal on the relationship between
neuroticism and coping was examined16. Partial correlations (controlling for gender)
between neuroticism, stress appraisals, controllability appraisals, emotion-focused
coping, and avoidance coping obtained in each anagram condition are summarised in
Tables 15 and 16. In both anagram conditions neuroticism was significantly
correlated with stress and controllability appraisals, as well as emotion-focused
coping. Both appraisal subscales were also significantly correlated with emotion-
focused and avoidance coping. In the High Stress condition neuroticism was also
significantly correlated with avoidance coping. These correlations suggested that, in
the Mild Stress condition, appraised stress and appraised controllability were both
potential mediators of the relationship between neuroticism and emotion-focused
coping. In the High Stress condition, appraised stress and appraised controllability
were potential mediators of the relationship between neuroticism and both emotion-
focused and avoidance coping. In order to determine whether the relationship
between neuroticism and both emotion focused and avoidance coping is mediated by
stress and controllability appraisals, a series of hierarchical multiple regressions
were conducted.
16 Given the experimental manipulation was specifically designed to influence appraisals of the task (the potential mediating variable), mediational analyses were conducted separately for each anagram condition.
144
Table 15. Partial correlations (controlling for gender) between neuroticism, stress appraisal, appraised controllability, emotion-focused coping, and avoidance coping in the mild stress condition
Neuroticism Appraisal
(Stress) Appraisal (Control)
E-F Coping Avoidance Coping
Neuroticism 1 .39* -.40* .40* .20 Appraisal (Stress)
1
-.68*
.81*
.48*
Appraisal (Control)
1
-.74*
-.30
E-F Coping 1 .53* Avoidance Coping
1
* p < .05
Table 16. Partial correlations (controlling for gender) between neuroticism, stress appraisal, appraised controllability, emotion-focused coping, and avoidance coping in the high stress condition
Neuroticism Appraisal
(Stress) Appriaisal (Control)
E-F Coping Avoidance Coping
Neuroticism 1 .52* .33* .70* .32* Appraisal (Stress)
1
-.50*
.70*
.60*
Appraisal (Control)
1
-.43*
-.49*
E-F Coping 1 .43* Avoidance Coping
1
* p < .05
In the Mild Stress condition both stress and controllability appraisals
significantly predicted emotion-focused coping, while controlling for neuroticism
(see Step 3 in Table 17). In the High Stress condition only stress appraisals
significantly predicted emotion-focused and avoidance coping while controlling for
neuroticism (see Step 3 in Table 18 and Table 19). Therefore, the potential
145
mediating effect of stress appraisal on the relationship between neuroticism and
emotion-focused coping was examined in both conditions. In the hard condition, the
potential mediating effect of stress appraisal on the relationship between neuroticism
and avoidance coping was also examined. The potential mediating effect of
appraised controllability was only examined in the easy anagram condition. In order
to control for any gender effects, gender was entered as the first step in all analyses.
Mild Stress Condition
Neuroticism, Appraisal, and Emotion-Focused Coping
Emotion-focused coping was entered as the outcome variable. Gender was
entered in the first step of the analysis. Neuroticism was entered in the second step
of the analysis. Appraisal (Stress) and Appraisal (Control) were entered in the third
step of the analysis. The results of the hierarchical regression are summarized in
Table 17. The final step of the model accounted for 73 percent of the variance in
emotion-focused coping; R2 = .73, F(2, 37) = 23.68, p < .001; and together
Appraisal (Stress) and Appraisal (Control) fully mediated the relationship between
neuroticism and emotion-focused coping, as the effect of neuroticism on emotion-
focused coping became non-significant when both appraisal scores were entered into
the analysis. The Sobel test is a formal test of the significance of the mediated
effect, where the mediated effect is divided by the square root of its standard error
term and compared to a normal distribution (Sobel, 1982). The Sobel test confirmed
that Appraisal (Stress) was a significant mediator of the relationship between
neuroticism and emotion-focused coping (z = 2.25, p = .025). The mediating effect
146
of Appraisal (Control) on the relationship between neuroticism and emotion-focused
coping was approaching significance (z = 1.89, p = .059).
Table 17. Results of the hierarchical regression testing the mediating effect of stressor appraisals on the relationship between neuroticism and emotion-focused coping in the mild stress condition
B Standard Error
β t p
Step 1
Gender 1.66 5.81 .05 .29 .777 Step 2 Gender 3.40 5.44 .10 .63 .536 Neuroticism .89 .34 .40 2.64 .012 Step 3 Gender 1.37 3.21 .04 .43 .672 Neuroticism .09 .22 .04 .41 .682 Appraisal (Stress)
1.22
.26
.57
4.67
< .001
Appraisal (Control)
1.15
.42
-.34
-2.76
.009
Note: outcome variable is emotion-focused coping
High Stress Condition
Neuroticism, Appraisal, and Emotion-Focused Coping
Emotion-focused coping was entered as the outcome variable. Gender was
entered in the first step of the analysis. Neuroticism was entered in the second step
of the analysis. Appraisal (Stress) and Appraisal (Control) were entered in the third
step of the analysis. The results of the hierarchical regression are summarized in
Table 18. The final step of the model accounted for 67 percent of the variance in
emotion-focused coping; R2 = .67, F(2, 38) = 17.92, p < .001; and as neuroticism
remained a significant predictor of emotion-focused coping, stressor appraisals did
147
not fully mediate the relationship between neuroticism and emotion-focused coping.
However, the Sobel test revealed that Appraisal (Stress) was a partial mediator of the
relationship between neuroticism and emotion-focused coping (z = 2.05, p = .040).
Table 18. Results of the hierarchical regression testing the mediating effect of stressor appraisals on the relationship between neuroticism and emotion-focused Coping in the high stress condition
B Standard
Error β t p
Step 1 Gender 6.89 4.53 .24 1.52 .137 Step 2 Gender 5.86 3.30 .20 1.78 .084 Neuroticism 1.53 .26 .68 5.98 < .001 Step 3 Gender 3.47 2.90 .12 1.19 .240 Neuroticism .99 .26 .44 3.86 < .001 Appraisal (Stress)
.96
.28
.43
3.42
.002
Appraisal (Control)
-.18
.30
-.07
-.59
.559
Note: outcome variable is emotion-focused coping Neuroticism, Appraisal, and Avoidance Coping
Avoidance coping was entered as the outcome variable. Gender was entered
in the first step of the analysis. Neuroticism was entered in the second step of the
analysis. Appraisal (Stress) and Appraisal (Control) were entered in the third step of
the analysis. The results of the hierarchical regression are summarized in Table 19.
The final step of the model accounted for 47 percent of the variance in avoidance
coping; R2 = .47, F(2, 38) = 7.99, p < .001. In the final step of the analysis
neuroticism became a non-significant predictor of avoidance coping. The Sobel test
148
confirmed that Appraisal (Stress) was a significant mediator of the relationship
between neuroticism and avoidance coping (z = 2.27, p = .023).
Table 19. Results of the hierarchical regression testing the mediating effect of stressor appraisals on the relationship between neuroticism and avoidance coping in the high stress condition
B Standard Error
β t p
Step 1 Gender 5.82 2.72 .32 2.14 .039 Step 2 Gender 5.53 2.62 .31 2.11 .041 Neuroticism .43 .20 .31 2.10 .043 Step 2 Gender 3.26 2.26 .18 1.45 .157 Neuroticism -.01 .20 -.01 -.05 .958 Appraisal (Stress)
.64
.22
.46
2.92
.006
Appraisal (Control)
-.39
.23
-.25
-1.69
.100
Note: outcome variable is avoidance coping
Discussion
The specific aims of Study 5 were to 1) assess the utility of an anagram-
solving task as a laboratory-based stressor, 2) determine if neuroticism-linked
differences in appraisal and coping could be obtained in the context of the anagram-
solving task, and 3) determine whether stressor appraisals mediate the relationship
between neuroticism and self-reported coping in the context of the anagram-solving
task.
149
The Anagram-Solving Task as an Experimental Stressor
Individuals in the Mild Stress condition solved significantly more anagrams
than individuals in the High Stress condition. This suggests that the anagram
difficulty manipulation was successful. Additionally, participants allocated to the
High Stress condition appraised the task more negatively than individuals allocated
to the Mild Stress condition. Specifically, individuals in the High Stress condition
reported feeling significantly more stressed, and perceived themselves as having
significantly less control during the task. Moreover, participants in the High Stress
condition reported significantly higher levels of tense arousal, as well as
significantly lower levels of energetic arousal, hedonic tone, and self-esteem.
Finally, individuals allocated to the High Stress condition reported engaging in
significantly more emotion-focused coping than participants in the Mild Stress
anagram condition. It therefore appears that the experimental manipulation evoked
both negative appraisals of the task, as well as a negative affective response.
Additionally, manipulating the difficulty of the task can also elicit an emotion-
focused coping response. Taken together these results suggest that the anagram task
is a useful laboratory-stressor.
Neuroticism, Appraisal, and Coping in the Context of the Anagram-Solving Task
Regardless of which anagram condition participants were assigned to, high
neuroticism individuals appraised the task as being significantly more stressful and
perceived themselves as having significantly less control when compared with low
neuroticism participants. These appraisal differences do not reflect differences in
150
task performance as high and low neuroticism participants solved the same number
of anagrams in both conditions. This is consistent with previous research (including
Study 1 in this thesis) revealing that neuroticism is associated with negative
subjective stressor appraisals (Engelhard, 2007; Gallagher, 1990; Mak et al., 2004;
Schneider, 2004). Consistent with the findings obtained in the previous Cyberball
studies, high neuroticism participants reported engaging in significantly more
emotion-focused coping than low neuroticism participants. This was the case
regardless of which anagram condition participants were allocated to. Additionally,
as predicted a neuroticism-linked difference in avoidance coping was also obtained.
Specifically, high neuroticism participants reported engaging in significantly more
avoidance coping than low neuroticism participants. Again, this was the case
regardless of anagram condition.
The findings of Study 5 are consistent with previous research using
retrospective and dispositional coping measures, which report correlations between
neuroticism and both emotion-focused and avoidance coping (Bouchard, 2003;
Carver et al., 1989; R. Gomez et al., 1999). Additionally, the findings of Study 5 are
consistent with those of Matthews and his colleagues (Matthews & Campbell, 1998;
Matthews et al., 2006), as well as Penley and Tomaka (2002), who all report that
neuroticism is correlated with emotion-focused and avoidance coping in a variety of
laboratory-based cognitive tasks. Importantly, Study 5 extends the findings of
Matthews and his colleagues by demonstrating that neuroticism-related differences
in emotion-focused and avoidance coping can be obtained in the context of an
151
anagram-solving task more well-known in the coping literature (e.g. Endler,
Macrodimitris et al., 2000; Endler, Speer et al., 2000).
In the previous Cyberball experiments, as well as the current anagram-
solving study, neuroticism-linked differences in emotion-focused coping (and
avoidance coping in the anagram-solving study) were obtained regardless of which
condition participants were assigned to. Differences were obtained irrespective of
whether participants were completely excluded, partially excluded, completely
included, allocated to the Mild Stress condition, or allocated to the High Stress
condition. These findings suggest that neuroticism may be associated with a general
reliance on emotion-focused (and possibly avoidance) coping as has been suggested
previously (e.g. Vollrath & Torgersen, 2000; Vollrath et al., 1995). Alternatively,
the fact that neuroticism-related differences in emotion-focused coping emerge even
in the mild or no stress conditions, raises the possibility that confounds between
neuroticism and emotion-focused coping measures may be resulting in a spurious
relationship between the two constructs. In particular, content overlap between
items on neuroticism and coping scales (specifically emotion-focused coping items
which are often negatively worded) may have confounded both previous and current
findings.
Item overlap has previously been posited to confound the relationship
between emotion-focused coping and maladaptive outcomes such as depression and
anxiety (Austenfeld & Stanton, 2004; Costa & McCrae, 1990; Coyne & Racioppo,
2000; Stanton, Danoff-Burg, Cameron, & Ellis, 1994; Stanton, Kirk, Cameron, &
Danoff-Burg, 2000). Similarly, neuroticism is widely documented to be associated
152
with exaggerated reporting of symptomatology (Feldman, Cohen, Doyle, Skoner, &
Gwaltney, 1999; Watson & Pennebaker, 1989); however, the possibility of
confounds between neuroticism and emotion-focused coping measures has not been
investigated. If content overlap does account for the relationship between
neuroticism and emotion-focused coping, this will have important implications for
research examining the role of neuroticism in the coping process. Study 6 (Chapter
10) addresses this issue in more detail.
Finally, according to the goodness of fit hypothesis (see Chapter 1), stressor
appraisals influence the relationship between stressful situations and coping strategy
choice. Specifically, when stressors are appraised as uncontrollable individuals
should engage in more emotion-focused and avoidance coping (e.g. Conway &
Terry, 1994; Park et al., 2004). Given that neuroticism is associated with a negative
appraisal style, it is possible that stressor appraisals may mediate the relationship
between neuroticism and emotion-focused and avoidance coping. In the Mild
Stress condition the mediating effect of stress appraisals on the relationship between
neuroticism and emotion-focused was significant. Additionally, the mediating effect
of appraised controllability on the relationship between neuroticism and emotion-
focused coping was approaching significance. Given that the Sobel test is
notoriously conservative (MacKinnon, Warsi, & Dwyer, 1995) this finding warrants
further investigation.
In the High Stress condition, stress appraisals partially mediated the
relationship between neuroticism and emotion-focused coping. Additionally, stress
appraisal was a significant mediator of the relationship between neuroticism and
153
avoidance coping in the High Stress condition. However, controllability appraisals
(which are vital according to the goodness of fit hypothesis) did not mediate
relationship between neuroticism and either emotion-focused or avoidance coping.
It is possible that in the High Stress condition controllability was so experimentally
constrained that there was less opportunity for individual differences to influence
control appraisals. In contrast, in the Mild Stress condition, where there was more
scope for individual differences to influence control appraisals, the mediating effect
of appraised controllability on the relationship between neuroticism and emotion-
focused coping was approaching significance. In a sense, this finding is analogous
to the distinction made between ambiguous and unambiguous Cyberball conditions
in Study 4. Taken together, the interaction between neuroticism level and condition
with regard to appraised controllability (obtained when using an ambiguous
Cyberball condition in Study 4), along with the findings obtained in Study 5, suggest
that the mediating influence of appraised controllability may only operate when
stressor characteristics are sufficiently ambiguous to allow an interpretive bias to
operate. This possibility warrants further investigation.
The findings obtained in Study 5 offer preliminary evidence that stressor
appraisals may mediate the relationship between neuroticism and both emotion-
focused and avoidance coping. Importantly, these results were obtained in the
context of a laboratory-based stressor in which all participants experienced exactly
the same scenario. Future research should explore this relationship in more detail as
it is likely to have important implications for coping intervention programs.
Specifically, the current findings highlight of the importance of ensuring individuals
154
make realistic appraisals of their environments; as these appraisals are likely to
influence their choice of coping strategy.
Conclusion
The results obtained in Study 5 confirmed the utility of an anagram-solving
task as an experimental stressor for examining individual differences in coping.
Results were consistent with the findings from Study 3 and Study 4, which reported
a neuroticism-linked difference in emotion-focused coping during the Cyberball
game (as well as a relationship between neuroticism and avoidance coping when
ostracism was more ambiguous). The fact that neuroticism-linked differences in
emotion-focused coping have been obtained regardless of which condition
participants are assigned to (in both the Cyberball game and the anagram-solving
task) suggests that neuroticism may be associated with a general reliance on
emotion-focused coping. Alternatively, content overlap between neuroticism and
emotion-focused coping items may have confounded previous findings. Study 6
(Chapter 10) aims to address this issue in more detail. Finally, results from Study 5
revealed that stressor appraisals can mediate the relationship between neuroticism
and both emotion-focused and avoidance coping in the context of an experimental
stressor. Exploring this relationship is likely to be a fruitful avenue for future
research.
155
CHAPTER 10: Study 6 – Neuroticism and Emotion-Focused Coping Measures:
Content Overlap and Item Cross-Loading?
Aims of Study 6
In Studies 3, 4, and 5 neuroticism-linked differences in emotion-focused
coping were obtained regardless of whether participants were completely included,
partially excluded, or completely excluded during the Cyberball game, and
regardless of whether they were assigned to the easy or hard condition in the
anagram-solving task. As mentioned previously, these findings are consistent with
the argument that neuroticism may be associated with a general reliance on emotion-
focused coping (Vollrath & Torgersen, 2000; Vollrath et al., 1995). However, the
fact that a neuroticism-linked difference emerged even in the no stress condition (i.e.
the completely included Cyberball condition) raised a concern that content overlap
between items on the neuroticism and emotion-focused coping measures may have
confounded the results of the Cyberball and anagram-solving experiments.
Item confounds have previously been posited to account for the relationship
between emotion-focused coping and maladaptive outcomes such as depression and
anxiety (Austenfeld & Stanton, 2004; Costa & McCrae, 1990; Coyne & Racioppo,
2000; Stanton et al., 1994; Stanton et al., 2000). Indeed, many emotion-focused
coping items contain distress-laden content and others merge distress with emotional
expression (Stanton et al., 1994). Similarly, many emotion-focused coping items
contain self-deprecatory content nearly identical to that found on inventories
156
assessing psychopathology (Stanton et al., 1994). Given the apparent confounding
of emotional expression and distress (as well as self-deprecation) it is perhaps
unsurprising that the emotion-focused subscales of current coping measures are
reported to be correlated with measures of distress or psychopathology (Stanton et
al., 1994). Indeed, it has even been suggested that researchers interested in coping
are at high risk of giving substantive interpretation to spurious relationships between
coping and adaptational outcome (Coyne & Racioppo, 2000). Importantly,
neuroticism is substantially correlated with many of the criteria used to gauge coping
effectiveness, including measures of subjective distress and symptomatology
(Feldman et al., 1999; Saklofske, Kelly, & Janzen, 1995; Watson & Hubbard, 1996;
Watson & Pennebaker, 1989); however, to the best of our knowledge the possibility
of item confounds between neuroticism and emotion-focused coping measures has
not yet been investigated empirically. Study 6 aimed to address this important issue.
If confounds between neuroticism and emotion-focused coping measures are
identified this will have serious implications for both previous and future research
examining the role of neuroticism in the coping process.
Participants in Studies 3, 4, and 5 all completed the neuroticism scale
compiled from the International Personality Item Pool (Appendix D; Goldberg et al.,
2006). This neuroticism scale was completed either in pre-screening or during the
experimental testing session. Additionally, after playing Cyberball or completing
the anagram-task all participants completed the CITS-S (which contains an emotion-
focused coping subscale; Matthews & Campbell, 1998). Study 6 used Confirmatory
Factor Analysis (CFA) to examine whether neuroticism and emotion-focused coping
157
items loaded onto separate factors; or whether any cross-loading of neuroticism and
emotion-focused coping items was evident. If item cross-loadings must be permitted
in order to achieve satisfactory model fit, then this would provide evidence that
some items are simultaneous indicators of both neuroticism and emotion-focused
coping. Therefore, the hypothesis of measurement confounds could not be rejected.
A second (and stronger) test of this hypothesis would be to examine whether the
addition of any cross-loading pathways (assuming any cross-loading items are
identified) results in improved model fit.
Method
Participants
The sample consisted of 287 individuals (109 male; 178 female) who had
participated in Studies 3, 4, and 5. Data from Study 2 (the first Cyberball
experiment) were not included as subsequently to this experiment the response
scales of all outcome measures were extended. Participant characteristics and
methodological procedures have been described previously (Chapters 6, 7, and 9).
Participants were grouped into one of two subsamples. Individuals who had been
excluded during Cyberball (in Study 3 and 4); as well as individuals assigned to the
hard anagram condition (in Study 5), where classified as High Stress (n = 147).
Participants who had been partially excluded or included during Cyberball (in
Studies 3 and 4), as well as individuals assigned to the easy anagram condition (in
Study 5), were classified as Mild Stress (n = 140).
158
Procedure
All participants completed the 10 item neuroticism scale compiled from the
International Personality Item Pool (Goldberg et al., 2006). Neuroticism items are
presented in Table 21. Participants in Study 4 where pre-screened with the
neuroticism scale before being invited to participate in the actual experiment.
Participants in Study 3 and Study 5 completed the neuroticism scale during testing
sessions. In Study 3 and Study 5 the neuroticism scale was completed before
participants played the Cyberball game or attempted the anagram-solving task. All
participants completed the CITS-S (Matthews & Campbell, 1998) after playing
Cyberball or completing the anagram-solving task. Of particular interest in Study 6
was the emotion-focused coping subscale which contains seven items. Emotion-
focused coping items are also presented in Table 21.
Analysis
Neuroticism items were responded to on a five-point scale (0: Very
inaccurate; 4: Very accurate). Emotion-focused coping items were responded to on
an 11-point scale (0: Not at all; 10: Extremely). The CFA was conducted using
AMOS 17. Covariance matrices were analysed using maximum likelihood methods.
Four goodness of fit indicators were considered in evaluating the model; however, it
should be noted that in this evolving field of research there is no definitive set of
criteria for defining model fit (Vandenberg & Lance, 2000). The value of Chi
squared (χ2) may be used as an index of fit; however, χ2 divided by its degrees of
freedom (χ2/DF) is less sensitive to sample size. The minimum acceptable value of
159
χ2/DF is three (Kline, 1998), although a χ2/DF of less than two is preferred (Byrne,
1989). The root mean square error of approximation (RMSEA) tests the fit of the
model to the covariance matrix. As a guideline an RMSEA of less than .05 indicates
good fit and an RMSEA of less than .08 indicates acceptable fit (Browne & Cudeck,
1993). The comparative fit index (CFI) has a range between zero and one (with one
indicating perfect fit). A CFI of .95 is considered good fit whilst a CFI of .90 is
considered adequate (Blunch, 2008). The standardised root mean square residual
(SRMR) is the average distance between predicted and observed variances and
covariances in the model, based on standardised residuals. The smaller the SRMR
the better the model fit. An SRMR of zero indicates perfect fit. A value of less than
.05 is widely considered good fit while a value of less than .08 is considered to be
adequate fit (Blunch, 2008).
Measurement invariance assesses the extent to which a model remains
constant across time or groups. Multiple-groups CFA allows for increasing levels of
stringency to be imposed on a model when it is tested in a new sample as a test of
measurement invariance. As with model fit statistics, there are no definitive criteria
for determining measurement invariance (Vandenberg & Lance, 2000). Study 6
assessed measurement invariance using four steps. Firstly, the unconstrained
(baseline) model against which further invariance constraints were tested was
assessed. In the second step the measurement weights (factor loadings) in the two
samples (High and Mild stress) were constrained to be equal. In the third step the
structural covariances (covariances between the factors) were also constrained to be
equal. In the final step the measurement errors were also constrained. Changes in
160
CFI and χ2 were used to assess measurement invariance If CFI is reduced by .01 or
more the hypothesis of between-group invariance can be rejected (Cheung &
Rensvold, 2002). A statistically significant change in χ2 can also be used to reject
the hypothesis of between-group invariance.
Results
Independent CFAs conducted in the High and Mild Stress samples
A two-factor model which specified that all neuroticism items loaded onto a
neuroticism factor and all emotion-focused coping items loaded onto an emotion-
focused coping factor was tested in both the High and Mild Stress samples. The
model also specified that the neuroticism and emotion-focused coping factors should
co-vary. Error terms for individual neuroticism and emotion-focused coping items
were only permitted to be correlated within measures. In the High Stress sample all
fit statistics were in the good range, with the exception of SRMR which was
acceptable. In the Mild Stress sample χ2/DF and CFI were good whilst RMSEA and
SRMR were in the acceptable range (see Table 20). Factor loadings for individual
items and factor covariances are summarised in Table 21. Examination of
modification indices can identify pathways that if released would result in improved
model fit. In AMOS, any modification index above 3.84 (generally rounded to 4 in
the AMOS display options) would improve model fit if released (Blunch, 2008). In
both the High and Mild stress samples no cross-loading pathways were identified
that, if released, improve model fit.
161
Table 20. Model fit indices obtained in the High Stress and Mild Stress samples
* p < .05
Multi-group CFA and measurement invariance
Having shown that the two factor model provided adequate fit in both the
High Stress and Mild Stress samples, increasing levels of constraint were applied to
assess measurement invariance. The fit indices (and change in both χ2 and CFI)
obtained in each step of the analysis are summarised in Table 22. CFI for the
combined High and Mild Stress data was .972 using the unconstrained two factor
model, which is in the good range. All other fit statistics were in the acceptable to
good range, and are the sum (in the case of χ2) or average of indices obtained when
the model was tested on High and Mild Stress models independently. When the
factor loadings were constrained to equality, CFI was reduced to .971 (a change of
.001 from baseline). The change in χ2 was not statistically significant. When the
factor covariances were constrained to equality, CFI increased to .972 (no change
from baseline). Again, the change in χ2 was not statistically significant. Finally,
when the measurement errors were constrained to equality, CFI was reduced to .964
(a change of .012 from baseline). Additionally, the change in χ2 was statistically
significant. Therefore, it was concluded that the factor loadings and covariances
High Stress Mild Stress χ2 120.87 158.35* χ2/DF 1.14 1.49 RMSEA .031 .060 CFI .986 .960 SRMR .052 .058
162
were invariant across the High Stress and Mild Stress samples but item error was not
invariant.
Table 21. Item loadings on the neuroticism (N) and emotion-focused coping (EFC) factors (and standardised covariance between N and EFC factors) in the High Stress and Mild Stress samples
High Stress
Mild Stress
Neuroticism Items:
N EFC N EFC
I am relaxed most of the time .59 .59 I get stressed out easily .70 .77 I seldom feel blue .58 .52 I worry about things .62 .68 I am easily disturbed .40 .58 I get upset easily .74 .76 I change my mood a lot .67 .70 I have frequent mood swings .66 .75 I get irritated easily .59 .69 I often feel blue .73 .70 Emotion-Focused Coping Items: Worried about what I would do next .62 .62 Blamed myself for not doing better .74 .83 Became preoccupied with my problem .59 .65 Wished I could change what’s happening .55 .61 Worried about my inadequacies .80 .89 Blamed myself for not knowing what to do .92 .86 Blamed myself for becoming too emotional .56 .79 Covariance (Standardised) between N and EFC .34* .47* Note: All item loadings are significant at p < .05 * p < .05.
163
Table 22. Fit indices (and change in χ2 and CFI) obtained in each step of the measurement invariance analysis
Unconstrained Model
Factor Loadings
Constrained
Factor Covariances Constrained
Measurement Errors
Constrained χ2 279.23* 294.63* 296.84* 345.23* χ2/DF 1.32 1.30 1.29 1.33 ∆χ2 15.40 2.21 48.39* RMSEA .033 .032 .032 .034 SRMR .052 .055 .061 .064 CFI .972 .971 .972 .964 ∆CFI .001 .000 .012 ∆χ2 = change in χ2, ∆CFI = change in CFI (compared to unconstrained model) * p < .05
Discussion
Study 6 used CFA to assess the possibility that confounds between item
content on the neuroticism and emotion-focused coping measures may have resulted
in spurious relationships between neuroticism and emotion-focused coping in the
previous Cyberball and anagram-solving experiments. Item confounds have
previously been posited to account for the relationship between emotion-focused
coping and maladaptive outcomes, such as depression and anxiety (Costa &
McCrae, 1990; Coyne & Racioppo, 2000; Stanton et al., 1994). Importantly,
neuroticism is substantially correlated with many of the criteria used to gauge coping
effectiveness, including measures of subjective distress and symptomatology
(Feldman et al., 1999; Saklofske et al., 1995; Watson & Hubbard, 1996; Watson &
Pennebaker, 1989); however, the possibility of item confounds between neuroticism
and emotion-focused coping measures has not yet been investigated empirically.
164
A two-factor model, specifying that all neuroticism items loaded onto a
neuroticism factor and all emotion-focused coping items loaded onto an emotion-
focused coping factor was tested on data collected in Studies 3, 4, and 5. In both the
High Stress and Mild Stress samples this model provided acceptable to good fit to
the actual data. Furthermore, the measurement invariance analysis established that
individual factor loadings as well as the covariance between the neuroticism and
emotion-focused coping factors in each sample could be constrained to equality
without substantially impacting model fit. Importantly, in both the High Stress and
Mild Stress samples there was no evidence for any cross-loading of neuroticism or
emotion-focused coping items. Furthermore, examination of the modification
indices revealed no cross-loading pathways that, if released, would improve model
fit. These results provide preliminary evidence that the neuroticism-linked
differences in emotion-focused coping obtained in the previous Cyberball and
anagram-solving experiments were not merely an artefact of item cross-loading.
However, it is important to note that Study 6 has several limitations. These
limitations stem largely from the fact that this study was not planned to be conducted
as part of this thesis from the outset. Rather, the fact that all participants in Studies
3, 4, and 5 had completed the same measures provided the opportunity to
empirically address this important issue.
The first major limitation of Study 6 was that in order to obtain samples large
enough to conduct the CFA (and admittedly these samples are still relatively small
for CFA), groups had to be created by collapsing across Cyberball and anagram
conditions. Ideally, this model should be independently tested in complete
165
exclusion, partial exclusion, and complete inclusion Cyberball conditions; as well as
in hard and easy anagram conditions. A second, and perhaps more important
limitation, is that the neuroticism and emotion-focused coping measures were
responded to on different scales. The neuroticism items were responded to on a
five-point scale whilst the emotion-focused coping items were responded to on an
11-point scale. The fact that neuroticism and emotion-focused coping items were
responded to on different scales means that these items yielded different
distributions. Therefore, it may be the case that neuroticism and emotion-focused
coping items loaded onto separate factors simply because their underlying
distributions were different. The current findings should be replicated using
neuroticism and emotion-focused coping measures that are responded to on similar
scales in order to clarify this issue. Additionally, in the current study the
neuroticism scale focused on dispositional behaviour whilst the coping scale focused
on situation-specific behaviour. It is possible that using a dispositional coping
measure would reveal more overlap. Future research should use the dispositional
version of the CITS (CITS-D) to explore this issue.
Conclusion
The results of the CFA and measurement invariance analysis conducted in
Study 6 indicated that a two-factor model provided acceptable fit for the data
collected in Studies 3-5. Importantly, factor loadings and covariances were invariant
across the two stress samples, and there was no evidence for any cross-loading of
166
neuroticism and emotion-focused coping items. Therefore, these findings offer
preliminary evidence that the neuroticism-linked differences in emotion-focused
coping reported previously in this thesis are not merely an artefact of item cross-
loading between neuroticism and emotion-focused coping measures. However, the
limitations associated with Study 6 prohibit the drawing of firm conclusions
regarding possible confounds between neuroticism and emotion-focused coping
items. These limitations need to be addressed, as this is an important issue that may
have serious implications for both previous and future research examining the role of
neuroticism in the coping process. CFA is an excellent tool for hypothesis testing
that imposes a level of discipline by requiring that models be specified a priori
(French & Tait, 2004), and is likely to be extremely useful in addressing this issue
empirically.
167
CHAPTER 11: General Discussion
This thesis investigated relationships between neuroticism, extraversion, and
situational coping. Chapter 1 provided a brief review of the coping literature and
Chapter 2 provided a more detailed summary of research examining relationships
between personality-related variables (including the big five factors) and coping.
Neuroticism and extraversion were identified as the personality dimensions that have
received the most attention in the research literature and therefore became the
research focus. Chapter 2 concluded that previous research in this area has been
limited by a reliance on retrospective and dispositional coping measures.
Specifically, these measures are argued to overestimate the relationship between
personality-related variables and coping strategy use in any given stressful situation
(e.g. Coyne & Gottlieb, 1996; David & Suls, 1999). It was concluded that
measuring coping in the context of situation-specific stressors (that allow coping
questionnaires to be administered concurrently or immediately after the event)
would allow firmer conclusions regarding the relationships between personality-
traits and coping responses to be made (Bolger, 1990). The primary aim of this
thesis was to explore the relationships between neuroticism, extraversion, and
coping in the context of situation-specific stressors; in order to overcome the
reliance on dispositional and retrospective coping measures that has limited previous
research.
Study 1 (Chapter 3) empirically tested the argument that dispositional coping
measurement overestimates the relationship between personality and coping (David
168
& Suls, 1999). In this study participants were assigned to either a dispositional
condition (completed a dispositional coping measure) or a situational condition
(completed a coping measure with regard to approaching university examinations).
Correlations between neuroticism, extraversion, and coping obtained in each group
were compared. Results offered support for the contention that relationships
between personality-traits and coping are larger when coping is measured
dispositionally. Specifically, results revealed that relationships between extraversion
and coping strategy use were overestimated when coping was measured
dispositionally. In contrast, the predicted correlations between neuroticism and
passive/maladaptive coping were obtained regardless of whether coping was
measured situationally or dispositionally. Based on these results, neuroticism
became the focus of the remainder of the thesis. Additionally, results obtained in
Study 1 raised the possibility that stressor appraisals mediate the relationship
between neuroticism and coping strategy use. A secondary aim of this thesis was to
explore this possibility. A number of methodological limitations associated with
using examinations as a situational stressor were identified and it was concluded that
the relationship between neuroticism and coping could best be examined in the
context of laboratory-based stressors. Chapter 4 introduced a ball-throwing
paradigm (Cyberball; Williams & Jarvis, 2006; Zadro et al., 2004), which has been
used successfully in ostracism research, and outlined its potential as a laboratory-
stressor for exploring individual differences in coping.
Studies 2-4 (Chapters 5-7) examined the relationship between neuroticism
and coping in the context of Cyberball (an experimental social stressor). This series
169
of experiments established that 1) Cyberball is a useful laboratory-stressor for
examining individual differences in coping, and 2) neuroticism is associated with
emotion-focused coping; regardless of whether participants are completely excluded,
partially excluded, or completely included during the game. Additionally,
neuroticism is correlated with avoidance coping when social ostracism is ambiguous
in nature. Importantly, these findings were obtained using a methodology that
overcomes the limitations of dispositional and retrospective coping measurement.
Results from this series of experiments also revealed that stressor appraisals do not
mediate the relationship between neuroticism and emotion-focused coping (at least
in the context of the Cyberball game). However, neuroticism was negatively
associated with controllability appraisals when the ostracism experienced was
ambiguous in nature. More research using ambiguous Cyberball conditions and
larger samples is needed to clarify whether appraisals of the game can mediate the
relationship between neuroticism and emotion-focused coping in the context of
Cyberball. To date, no individual difference variables (or experimental
manipulations) have been identified that moderate the negative impact of being
excluded during traditional Cyberball games (Gonsalkorale & Williams, 2006;
Zadro et al., 2004). Using ambiguous Cyberball conditions may shed light on
individual differences in sensitivity to social ostracism, and are likely to be a useful
context for examining the impact of experimental manipulations (such as
in/outgroup manipulations, e.g. Gonsalkorale & Williams, 2006) on perceptions of
social ostracism. Future Cyberball research should explore these possibilities.
170
Study 5 (Chapter 9) examined the relationship between neuroticism and
coping in the context of an anagram-solving task (Endler, Macrodimitris et al.,
2000). This experiment established that 1) the anagram-solving task is also a useful
laboratory-stressor for examining individual differences in coping, and 2)
neuroticism is associated with both emotion-focused and avoidance coping;
regardless of whether participants were assigned to the easy or hard anagram
condition. Again, this methodology minimises the limitations of
dispositional/retrospective coping measurement. Additionally, Study 5 offered
preliminary evidence that stressor appraisals can mediate the relationship between
neuroticism and both emotion-focused and avoidance coping.
Taken together the findings from Studies 3-5 clearly reveal that neuroticism
is associated with emotion-focused coping in the context of laboratory-based
stressors (as well as avoidance coping in the context of the ambiguous Cyberball
condition and the anagram-solving task). Indeed, the fact that neuroticism is
associated with emotion-focused coping regardless of experimental conditions (as
well as avoidance coping in both the Mild and High Stress anagram conditions)
suggests that neuroticism may be associated with a general reliance on emotion-
focused and to a lesser extent avoidant coping strategies (however, it should be
noted that the reliability of the avoidance coping subscale was low in all studies
using the expanded response scale, and findings should be interpreted with caution –
see Table 23). The findings obtained in Studies 3-5 are consistent with previous
research (e.g. Matthews & Campbell, 1998; Matthews et al., 2006; Penley &
Tomaka, 2002; Vollrath & Torgersen, 2000; Vollrath et al., 1995); however, the
171
current findings overcome the reliance on retrospective and dispositional coping
measurement that have limited previous research in this area. Moreover, Study 5
provided preliminary evidence that the relationships between neuroticism, emotion-
focused, and avoidance coping may be mediated by stressor appraisals. This is
consistent with the goodness of fit hypothesis outlined in Chapter 1. It is also
consistent with research reporting an interpretive information processing bias (e.g.
Hadwin et al., 1997; Taghavi et al., 2000) associated with high levels of trait
anxiety, which is strongly correlated with neuroticism (e.g. Scheier et al., 1994).
These findings offer a possible explanation for why neuroticism is associated with
passive and maladaptive ways of coping. Specifically, it appears that high
neuroticism individuals may appraise/interpret stressful situations more negatively
than low neuroticism individuals, and therefore choose to use emotion-focused and
avoidant coping strategies. However, this relationship was only obtained in Study 5
and the results are preliminary in nature. More research exploring the mediating
effect of appraisal on the relationship between neuroticism and coping is clearly
needed. Additionally, future research should address the possibility that personality
traits may interact to influence appraisals, stress-responses, and coping. Similarly,
research exploring the relationships between personality and coping, using both
broad dimensions (eg. neuroticism) and specific personality facets (eg. worry) is
likely to be enlightening.
Establishing that stressor appraisals can mediate the relationship between
neuroticism and coping may have important implications for practitioners interested
in coping intervention. Importantly, this may help bridge the gap between
172
theoretical coping research and clinical research (Coyne & Racioppo, 2000). The
mediating influence of stressor appraisal on the relationship between neuroticism
and coping fits conceptually within a cognitive-behavioural therapy framework, and
coping interventions using this framework to target both stressor appraisals and
coping behaviours/responses may be effective. However, future research is needed
to establish this empirically. Future research should also examine exactly what
situational/stressor characteristics influence the mediating effect of appraisals on the
relationship between neuroticism and coping. The goodness of fit hypothesis
emphasises the influence of stressor controllability on choice of coping strategy (e.g.
Endler, Macrodimitris et al., 2000; Endler & Parker, 1990a; Zakowski, Hall, Klein,
& Baum, 2001). Consistent with this, Study 5 provided evidence that appraised
controllability can mediate the relationship between neuroticism and coping in the
context of a laboratory-stressor. However, the issue of stressor ambiguity also
appears to warrant further research. The current findings suggest that neuroticism-
linked differences in appraisal are most likely to be obtained when the stressful
situation is ambiguous in nature. Laboratory-based stressors, which allow
characteristics of the stressful situation (e.g. controllability and ambiguity) to be
experimentally manipulated, are likely to be particularly useful in clarifying exactly
what situational characteristics determine whether or not individual differences in
appraisal will mediate relationships between personality-related variables and
coping.
It should be noted that the samples of the laboratory-based experiments
reported in this thesis consisted of undergraduate psychology students. The
173
undergraduate psychology population is not representative of the community
population as a whole. Indeed, there is some evidence that neuroticism scores are
slightly elevated in college samples (Costa & McCrae, 1992). Future research
should explore the relationships between neuroticism, appraisal, and coping obtained
in laboratory-based stressors using more representative community samples.
Furthermore, although the rationale for using laboratory-stressors to examine the
relationship between neuroticism and coping was clearly established (see Chapters 3
and 4), there are a number of limitations associated with laboratory experiments.
Most importantly, the generalisation of findings obtained in laboratory settings to the
‘real-world’ can be problematic (Tajfel, 1972). Specifically in this thesis, ethical
considerations meant that participants could only be exposed to very mild stressors
in the laboratory setting. The way in which individuals cope with mild laboratory-
stressors may be different to how they cope with stressful situations that have real-
world importance and implications. Now that the relationship between neuroticism
and emotion-focused coping (and avoidance coping to a lesser degree) has been
clearly established under controlled laboratory conditions, future research should
replicate these findings using situation-specific real-world stressors.
Real-world stressors are inherently ambiguous and should allow scope for
individual differences in appraisal and coping to be exhibited, and would also allow
the mediating effect of stressor appraisals to be examined. However, in order for
clear conclusions regarding individual differences to be made, it is imperative that
all participants report the coping strategies used when dealing with the same
situational stressor. For example, obtaining immediate coping reports from
174
individuals undergoing the same medical or dental procedures, individuals being
interviewed for the same jobs, and individuals giving formal presentations (or
engaging in other forms of public speaking) may minimise the limitations of
dispositional and retrospective coping measurement while maintaining some ‘real-
world’ validity. Further real-world research that simply examines correlations
between personality and scores obtained using dispositional and/or retrospective
coping measures is unlikely to be enlightening (Connor-Smith & Flachsbart, 2007).
It should also be noted that all of the variables of interest in this thesis were
measured with self-report questionnaires. Self-report methodologies are associated
with numerous limitations, including demand biases and social desirability effects
(see Coyne & Gottlieb, 1996, for an excellent review of the limitations specifically
associated with coping checklists). Although the measurement of coping
(particularly emotion-focused coping) is likely to remain reliant on self-report
measures, physiological and behavioural measures may be useful when exploring
the impact and outcomes associated with being exposed to laboratory-stressors.
There is a substantial literature reporting that stress is linked to physiological arousal
(see Chida & Hamer, 2008, for a meta-analysis of 729 studies). It would be
interesting to know whether physiological responses (such as heart rate and skin
conductance) are influenced by being excluded during Cyberball, or being asked to
solve anagrams. This information would be useful in determining what laboratory-
tasks are the most effective in evoking a stress response. Moreover, an
understanding of whether or not physiological measures correlate with self-reported
175
arousal levels (as measured by the MACL) would be useful in assessing the
accuracy of self-reports.
Similarly, identifying whether or not laboratory-stressors can affect
performance on behavioural tasks known to be influenced by stress level would also
provide useful information for researchers. For example, Zakowski and colleagues
(2001) describe a proof-reading task in which performance (identifying grammatical
errors) has been shown to be inversely related to current stress level (e.g. S. Cohen,
1980; Glass & Singer, 1992; Zakowski et al., 2001). Supplementing self-report data
with physiological and behavioural observations wherever possible would allow for
comparisons between physiological, behavioural, and self-report data to be made
and would further our understanding of the outcomes associated with being exposed
to laboratory-stressors.
The fact that in Studies 3-5 neuroticism linked differences in emotion-
focused coping were obtained regardless of experimental conditions is consistent
with the contention that neuroticism is associated with a general reliance on
emotion-focused coping strategies (Vollrath & Torgersen, 2000). However, these
findings also raised the possibility that confounds between the neuroticism and
emotion-focused coping items could be resulting in a spurious relationship between
measures of these constructs. This issue was addressed in Study 6 (Chapter 10),
which used CFA on data collected in Studies 3-5 to determine whether content
overlap between items on the neuroticism and emotion-focused coping measures had
confounded results. Results of the CFA and invariance analyses revealed that a two
factor solution (Neuroticism and Emotion-Focused Coping) provided adequate to
176
good fit to the data, and there was no evidence of any item cross-loading. However,
limitations associated with Study 6 (see Chapter 10 for a discussion of these
limitations) mean these results are not conclusive. The question of measurement
confounds remains an important one, and more research (particularly using
neuroticism and emotion-focused coping items responded to on scales with similar
response ranges) is needed to clarify this issue. Until this question is clarified,
researchers in this field are at risk of making substantive conclusions about a
possibly spurious relationship between neuroticism and emotion-focused coping.
CFA is likely to be extremely useful statistical tool for addressing this question
empirically.
Additionally, emotion-focused coping items on generic coping checklists
have been criticised for being confounded with outcomes (Austenfeld & Stanton,
2004; Costa & McCrae, 1990; Coyne & Racioppo, 2000; Stanton et al., 1994;
Stanton et al., 2000). Similarly, the emotion-focused items on the CITS-S (which
was used in Studies 2-5) contain references to emotional reactions, such as worry
and self-blame (see Table 25 in Chapter 10 for a list of actual items). These items
may be better conceptualised as tapping unconscious or automatic emotional
reactions to (or outcomes of) stressful situations, rather than coping as defined by
Lazarus and Folkman (1984), which is viewed as a conscious and effortful process17.
This is a problem inherent with many coping scales (Coyne & Gottlieb, 1996; Coyne
& Racioppo, 2000). Perhaps emotion-focused coping should be reconceptualised to
17 Although it must be acknowledged that the CITS-S instructions emphasize the use deliberately chosen strategies to deal with stress
177
include both emotional reactivity and emotional regulation components18. Emotion-
focused subscales of current coping measures (such as the CITS-S) appear to do a
good job of capturing emotional reactivity during stressful situations, but do not
necessarily capture the strategies individuals might use to regulate these emotional
reactions. There is now a substantial literature focused specifically on emotion
regulation as a psychological construct (see Gross, 2007, for an excellent handbook),
and this should be used to inform future coping research.
Integrating the coping and emotional regulation literatures is likely to be
beneficial for coping researchers. Specifically, Gross’s (2001) process model of
emotion-regulation is likely to provide a useful framework for examining individual
differences in emotion-regulation in the context of specific stressful situations. This
model outlines five families of emotion regulation processes: situation selection,
situation modification, attentional deployment, cognitive change, and response
modulation (Gross, 2001; Gross & John, 2003). For researchers interested in
individual differences in situational coping, the concepts of attentional deployment,
cognitive change, and response modulation are likely to be particularly useful. This
framework could be used to develop emotion-focused coping items tapping emotion-
regulation rather than emotional reactivity.
Integrating the coping and emotion regulation literatures is a logical next step
for researchers interested in individual differences in emotion-focused coping, and
may reinvigorate theoretical interest in coping as a psychological construct.
18 The work of Stanton and colleagues has begun to address this issue and has revealed that automatic emotional reactivity and controlled emotion regulation load onto separate factors (see Austenfeld & Stanton, 2004; Stanton et al., 2000; Stanton, Parsa, & Austenfeld, 2002)
178
Additionally, identifying the strategies employed to regulate emotion within specific
contexts, rather than relying on generic measures, would provide useful information
that may be of applied benefit. Specifically, understanding exactly what emotion-
regulation strategies are effective in given contexts would provide practitioners
interested in intervention with useful information. Again, this should help bridge the
gap between coping research and clinical intervention research identified by Coyne
and Racioppo (2000). However, obtaining this information would require the
development of task/stressor specific measures. The process of designing these
scales would need to be monitored closely in order to allow meaningful comparisons
to be made across studies and stressful situations, and prevent the conceptual
limitations that have been associated with much of the coping literature to date. The
framework outlined by Ajzen (2006) for constructing theory of planned behaviour
questionnaires may prove useful in this context.
Conclusion
The primary focus of this thesis was on neuroticism-related differences in
situational coping. The research reported in this thesis established the utility of two
laboratory-based tasks as experimental stressors. It also established that neuroticism
is associated with situational coping in the context of both these laboratory-stressors,
as well as in the context of a real-life stressor (approaching university examinations).
Across laboratory-based studies neuroticism was consistently associated with
emotion-focused coping, as well as avoidance coping in the ambiguous Cyberball
179
condition and the anagram-solving task. In the examination study (Study 1)
neuroticism was associated with passive and maladaptive emotion-focused and
avoidant strategies (such as denial and venting emotions). Importantly, the
methodologies employed in this thesis overcome the reliance on dispositional and
retrospective coping measurement that has limited much previous research in this
area. The current research extends the literature by showing that predicted
relationships between neuroticism and coping can be obtained within the constraints
of controlled laboratory tasks, in which all participants experience exactly the same
scenario.
Secondly, the current research extends the literature by establishing that
stressor appraisals may mediate the relationship between neuroticism and both
emotion-focused and avoidance coping. This finding offers a possible explanation
for why neuroticism is associated with the use of passive and maladaptive coping
strategies, and is consistent with more general explanations of situational variation in
coping strategy use (e.g. the goodness of fit hypothesis). Further exploration of the
mediating influence of appraisals on the relationship between neuroticism and
coping is identified as an area that warrants further investigation. In particular, the
exploration of individual differences in the appraisal of ambiguous stressors, and the
subsequent choice of coping strategies, may prove fruitful. A greater understanding
these relationships may help bridge the gap between theoretical coping research and
clinical intervention research. On a related note, the use of ambiguous Cyberball
conditions may extend our understanding of the negative impact of social ostracism,
180
and help identify whether there are any individual differences in sensitivity to being
socially excluded.
Finally, this thesis established that item confounds between neuroticism and
emotion-focused coping do not to account for the relationship between neuroticism
and emotion-focused coping (at least in the current samples); however, further
research addressing this important issue is recommended. CFA was identified as a
useful tool for empirically evaluating the possibility of confounds between
neuroticism and emotion-focused coping measures. Additionally, it was concluded
that reconceptualising emotion-focused coping as encompassing both emotional-
reaction and emotion-regulation components is likely to be beneficial. From this
perspective, integrating the coping and emotion-regulation literatures is likely to
further our understanding of coping as a psychological construct and suggest
directions for future research.
181
REFERENCES
Ajzen, I. (2006). Constructing a TpB Questionnaire: Conceptual and Methodological
Considerations. from http://people.umass.edu/aizen/pdf/tpb.measurement.pdf
Aldridge, A. A., & Roesch, S. C. (2008). Developing coping typologies of minority
adolescents: A latent profile analysis. Journal of Adolescence, 31, 499-517.
Aluja, A., Rolland, J., Garcia, L., & Rossier, J. (2007). Dimensionality of the
Rosenberg self-esteem scale and its relationship with the three- and five-
factor personality models. Journal of Personality Assessment, 88, 246-249.
Amirkhan, J. H., Risinger, R. T., & Swickert, R. J. (1995). Extraversion: A "hidden"
personality factor in coping. Journal of Personality, 63, 189-212.
Austenfeld, J. L., & Stanton, A. L. (2004). Coping through emotional approach: A
new look at emotion, coping, and health-related outcomes. Journal of
Personality, 72, 1335-1363.
Band, E. B., & Weisz, J. R. (1988). How to feel better when it feels bad: Children's
perspectives on coping with everyday stress. Developmental Psychology, 24,
247-253.
Band, E. B., & Weisz, J. R. (1990). Developmental differences in primary and
secondary control coping and adjustment to juvenile diabetes. Journal of
Clinical Child Psychology, 19, 150-158.
Bandura, A. (1995). Self-efficacy in changing societies. New York: Cambridge
University Press.
182
Barbaranelli, C., Caprara, G. V., Rabasca, A., & Pastorelli, C. (2003). A
questionnaire for measuring the Big Five in late childhood. Personality and
Individual Differences, 34, 645-664.
Barnett, P. A., & Gottlib, L. H. (1988). Psychosocial functioning and depression:
Distinguishing among antecedents, concomitants, and consequences.
Psychological Bulletin, 117, 497-529.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction
in social psychological research: Conceptual, strategic, and statistical
considerations. Journal of Personality and Social Psychology, 51, 1173-
1182.
Barrett, P. M., Rapee, R. M., Dadds, M. R., & Ryan, S. M. (1996). Family
enhancement of cognitive style in anxious and aggressive children. Journal
of Abnormal Child Psychology, 24, 187-203.
Baumeister, R. F., & 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., & Tice, D. M. (1990). Anxiety and social exclusion. Journal of
Social and Clinical Psychology, 9, 165-195.
Billings, A. G., & Moos, R. H. (1981). The role of coping responses and social
resources in attenuating the stress of life events. Journal of Behavioral
Medicine, 4, 139-157.
183
Blalock, J. A., & Joiner, T. E. (2000). Interaction of cognitive avoidance coping and
stress in predicting depression/anxiety. Cognitive Therapy and Research, 24,
47-65.
Blascovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. P. Robinson, P.
R. Shaver & L. S. Wrightsman (Eds.), Measures of personality and
psychological attitudes, Volume 1. San Diega, CA: Academic Press.
Blunch, N. J. (2008). Introduction to structural equation modelling using SPSS and
AMOS. London: Sage.
Boehm, C. (1986). Capital punishment in tribal Montenegro: Implications for law,
biology, and theory of social control. Ethology and Sociobiology, 7, 305-320.
Bogels, S. M., & Zigterman, D. (2000). Dysfunctional cognitions in children with
social phobia, seperation anxiety disorder, and generalised anxiety disorder.
Journal of Abnormal Child Psychology, 28, 205-211.
Bolger, N. (1990). Coping as a personality process: A prospective study. Journal of
Personality and Social Psychology, 59, 525-537.
Borgatta, E. F. (1964). The structure of personality charcteristics. Behavioral
Science, 9, 8-17.
Bouchard, G. (2003). Cognitive appraisals, neuroticism, and openness as correlates
of coping strategies: An integrative model of adaptation to marital
difficulties. Canadian Journal of Behavioural Science, 35, 1-12.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit.
Sociological Methods and Research, 21, 230-258.
184
Buss, D. M. (1990). The evolution of anxiety and social exclusion. Journal of Social
and Clinical Psychology, 9, 196-201.
Buss, D. M., & Kenrick, D. (1998). Evolutionary social psychology. In D. T.
Gilbert, S. T. Fiske & G. Lindzey (Eds.), The handbook of social psychology
(pp. 982-1026). Boston: McGraw-Hill.
Byrne, B. M. (1989). A primer of LISREL. Basic applications and programming for
confirmatory factor analytic models. New York: Springer-Verlag.
Cacioppo, J. T., & Hawkley, L. C. (2003). Social isolation and health, with an
emphasis on underlying mechanisms. Perspectives in Biology and Medicine
46, 39-52.
Carver, C. S., & Scheier, M. F. (1994). Situational coping and coping dispositions in
a stressful transaction. Journal of Personality and Social Psychology, 66,
184-195.
Carver, C. S., Scheier, M. F., & Pozo, C. (1992). Conceptualizing the process of
coping with health problems. In H. S. Friedman (Ed.), Hostility, coping, and
health (pp. 167-199). Washington, DC: American Psychological Association.
Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping
strategies: A theoretically based approach. Journal of Personality and Social
Psychology, 56, 267-283.
Case, R. B., Moss, A. J., Case, N., McDermott, M., & Elberly, S. (1992). Living
alone after myocardial infarction: impact on prognosis. Journal of the
American Medical Association, 267, 515-519.
185
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for
testing measurement invariance. Structural Equation Modeling, 9, 233-255.
Chida, Y., & Hamer, M. (2008). Chronic psychosocial factors and acute
physiological responses to laboratory-induced stress in healthy populations:
A quantitative review of 30 years of investigations. Psychological Bulletin,
134, 829-855.
Christensen, A. J., Benotch, E. G., Wiebe, J. S., & Lawton, W. J. (1995). Coping
with treatment-related stress: Effects on patient adherence in hemodialysis.
Journal of Consulting and Clinical Psychology, 63, 454-459.
Clark, L., & Watson, D. (1991). General affective dispositions in physical and
psychological health. In C. Snyder & D. R. Forsyth (Eds.), Handbook of
social and clinical psychology. New York: Permagon Press.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for
the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum Associates.
Cohen, S. (1980). Aftereffects of stress on human performance and social behavior:
A review of research and theory. Psychological Bulletin, 88, 82-108.
Compas, B. E., & Boyer, M. C. (2001). Coping and attention: implications for child
health and pediatric conditions. Journal of Developmental and Behavioral
Pediatrics, 22, 323-333.
Compas, B. E., Connor-Smith, J. K., Saltzman, H., Harding Thomsen, A., &
Wadsworth, M. E. (2001). Coping with stress during childhood and
adolescence: Problems, progress, and potential in theory and research.
Psychological Bulletin, 127, 87-127.
186
Compas, B. E., Worsham, N. L., Ey, S., & Howell, D. C. (1996). When mom or dad
has cancer II. Coping, cognitive appraisals, and psychological distress in
children of cancer patients. Health Psychology, 15, 167-175.
Connor-Smith, J. K., & Compas, B. E. (2002). Vulnerability to social stress: Coping
as a mediator or moderator of sociotropy and symptoms of anxiety and
depression. Cognitive Therapy and Research, 26, 39-55.
Connor-Smith, J. K., & Compas, B. E. (2004). Coping as a moderator of relations
between reactivity to interpersonal stress, health status, and internalizing
problems. Cognitive Therapy and Research, 28, 347-368.
Connor-Smith, J. K., Compas, B. E., Wadsworth, M. E., Harding Thomsen, A., &
Saltzman, H. (2000). Responses to stress in adolescence: measurement of
coping and involuntary stress responses. Journal of Consulting and Clinical
Psychology, 68, 976-992.
Connor-Smith, J. K., & Flachsbart, C. (2007). Relations between personality and
coping: A meta-analysis. Journal of Personality and Social Psychology, 93,
1080-1107.
Conway, V. J., & Terry, D. J. (1994). Appraised controllability as a moderator of the
effectiveness of different coping strategies: A test of the goodness of fit
hypothesis. Australian Journal of Psychology, 44, 1-7.
Costa, P. T., Jr., & McCrae, R. R. (1985). The NEO Personality Inventory manual.
In. Odessa: Psychological Assessment Resources.
Costa, P. T., Jr., & McCrae, R. R. (1987). Neuroticism, somatic complaints, and
disease: Is the bark worse than the bite? Journal of Personality, 55, 299-316.
187
Costa, P. T., Jr., & McCrae, R. R. (1988). From catalog to classification: Murray's
needs and the five-factor model. Journal of Personality and Social
Psychology, 55, 258-265.
Costa, P. T., Jr., & McCrae, R. R. (1990). Personality: Another "hidden factor" in
stress research. Psychological Inquiry, 1, 22-24.
Costa, P. T., Jr., & McCrae, R. R. (1992). Revised NEO Personality Inventory
(NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional
manual. Odessa: Psychological Assessment Resources.
Coyne, J. C., & Gottlieb, B. H. (1996). The mismeasure of coping by checklist.
Journal of Personality, 64, 959-991.
Coyne, J. C., & Racioppo, M. W. (2000). Never the twain shall meet? Closing the
gap between coping research and clinical intervention research. American
Psychologist, 55, 655-664.
Craighead, W. E., Kimball, W. H., & Rehak, P. J. (1979). Mood changes,
physiological responses, and self-statements during social rejection imagery.
Journal of Consulting and Clinical Psychology, 47, 385-396.
Crawford, J. R., & Henry, J. D. (2003). The Depression Anxiety Stress Scale
(DASS): Normative data and latent structure in a large non-clinical sample.
British Journal of Clinical Psychology, 42, 111-131.
David, J. P., & Suls, J. (1999). Coping efforts in daily life: Role of big five traits and
problem appraisal. Journal of Personality, 67, 265-294.
Davis, B. O., Jr. (1991). Benjamin O. Davis, Jr., American: An autobiography.
Washington, DC: Smithsonian Institution Press
188
DeLongis, A., & Holtzman, S. (2005). Coping in context: The role of stress, social
support, and personality in coping. Journal of Personality, 73, 1-24.
Diener, E., & Larsen, R. J. (1984). Temporal stability and cross-situational
consistency of affective, behavioral, and cognitive responses. Journal of
Personality and Social Psychology, 47, 871-883.
Dineen, K. A., & Hadwin, J. A. (2004). Anxious and depressive symptoms and
children's judgements of their own and others' interpretations of ambiguous
social scenarios Journal of Anxiety Disorders, 18, 499-513.
Ebata, A. T., & Moos, R. H. (1991). Coping and adjustment in distressed and
healthy adolescents. Journal of Applied Developmental Psychology, 12, 33-
54.
Eisenberger, N. I., & Lieberman, M. D. (2004). Why rejection hurts: A common
neural alarm system for physical and social pain. Trends in Cognitive
Sciences, 8, 294-300.
Eisenberger, N. I., & Lieberman, M. D. (2005). When it hurts to be left out: The
neurocognitive overlap between physical and social pain. In K. D. Williams,
J. Forgas & W. Von Hippel (Eds.), The social outcast: Ostracism, social
exclusion, rejection, and bullying (pp. 109-130). New York: Psychology
Press.
Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection
hurt? An fMRI study of social exclusion. Science, 302, 290-292.
189
Endler, N. S., Macrodimitris, S. D., & Kocovski, N. L. (2000). Controllability in
cognitive and interpersonal tasks: Is control good for you? Personality and
Individual Differences, 29, 951-962.
Endler, N. S., & Parker, J. D. A. (1990a). Coping inventory for stressful situations
manual. Toronto: Multi Health Systems.
Endler, N. S., & Parker, J. D. A. (1990b). Multidimensional assessment of coping: A
critical evaluation. Journal of Personality and Social Psychology, 58, 844-
854.
Endler, N. S., & Parker, J. D. A. (1990c). State and trait anxiety, depression, and
coping styles. Australian Journal of Psychology, 42, 207-220.
Endler, N. S., Speer, R. L., Johnson, J. M., & Flett, G. L. (2000). Controllabilty,
coping, efficacy, and distress. European Journal of Personality, 14, 245-264.
Engelhard, I. M. (2007). Preexisting neuroticism, subjective stressor severity, and
posttraumatic stress disorder in soldiers deployed to Iraq. Canadian Journal
of Psychiatry, 52, 505-509.
Faulkner, S. L., & Williams, K. D. (1999). After the whistle is blown: The aversive
impact of ostracism. University of Toledo.
Feldman, P. J., Cohen, S., Doyle, W. J., Skoner, D. P., & Gwaltney, J. M., Jr.
(1999). The impact of personality on the reporting of unfounded symptom
and illness. Journal of Personality and Individual Differences, 77, 370-378.
Fickova, E. (2001). Personality regulators of coping behavior in adolescents. Studia
Psychologica, 43, 219-226.
190
Fields, L., & Prinz, R. J. (1997). Coping and adjustment during childhood and
adolescence. Clinical Psychology Review, 17, 937-976.
Folkman, S. (1984). Personal control and stress in the coping process: A theoretical
analysis. Journal of Personality and Social Psychology, 46, 839-852.
Folkman, S. (1992). Making the case for coping. In B. N. Carpenter (Ed.), Personal
coping: Theory, research, and application. Westport: Praeger Publishers.
Folkman, S., Lazarus, R. S., Dunkel-Schetter, C., DeLongis, A., & Gruen, R. J.
(1986). The dynamics of a stressful encounter. Journal of Personality and
Social Psychology, 50, 992-1003.
Folkman, S., Lazarus, R. S., Gruen, R. J., & DeLongis, A. (1986). Appraisal, coping,
health status, and psychological symptoms. Journal of Personality and
Social Psychology, 50, 571-579.
Folkman, S., & Moskowitz, J. T. (2004). Coping: Pitfalls and promise. Annual
Review of Psychology, 55, 745-774.
Fournier, M., de Ridder, D., & Bensing, J. (2002). Optimism and adaptation to
chronic disease: The role of optimism in relation to self-care options of type
1 diabetes mellitus, rheumatoid arthritis and multiple sclerosis. British
Journal of Health Psychology, 7, 409-432.
French, D. J., & Tait, R. J. (2004). Measurement invariance in the General Health
Questionnaire-12 in young Australian adolescents. European Child and
Adolescent Psychiatry, 13, 1-7.
Freud, S. (1933). New introductory lectures on psychoanalysis. New York: Norton.
191
Gallagher, D. J. (1990). Extraversion, neuroticism and appraisal of stressful
academic events. Personality and Individual Differences, 11, 1053-1057.
Geller, D. M., Goodstein, L., Silver, M., & Sternberg, W. C. (1974). On being
ignored: The effects of violation of implicit rules of social interaction.
Sociometry, 37, 541-556.
Glass, D. C., & Singer, J. E. (1992). Urban stress. New York: Academic Press.
Goldberg, L. R. (1990). An alternative "description of personality": The Big-Five
factor structure. Journal of Personality and Social Psychology, 59, 1216-
1229.
Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C., Cloninger,
C. R., et al. (2006). The international personality item pool and the future of
public domain personality measures. Journal of Research in Personality, 40,
84-96.
Gomez, A. (1997). Locus of control and Type A behaviour pattern as predictors of
coping styles among adolescents. Personality and Individual Differences, 23,
391-398.
Gomez, R. (1998a). Impatience-aggression, competitiveness and avoidant coping:
Direct and moderating effects on maladjustment among adolescents.
Personality and Individual Differences, 25, 649-661.
Gomez, R. (1998b). Locus of control and avoidant coping: Direct, interactional and
mediational effects on maladjustment in adolescents. Personality and
Individual Differences, 24, 325-334.
192
Gomez, R., Holmberg, K., Bounds, J., Fullarton, C., & Gomez, A. (1999).
Neuroticism and extraversion as predictors of coping styles during early
adolescence. Personality and Individual Differences, 27, 3-17.
Gonsalkorale, K., & Williams, K. D. (2006). The KKK won't let me play: ostracism
even by a despised outgroup hurts. European Journal of Social Psychology,
37, 1176-1186.
Goodall, J. (1986). Social rejection, exclusion, and shunning among the Gombe
chimpanzees. Ethology and Sociobiology, 7, 227-236.
Gow, A. J., Whiteman, M. C., Pattie, A., & Deary, I. J. (2005). Goldberg’s ‘IPIP’
Big-Five factor markers: Internal consistency and concurrent validation in
Scotland. Personality and Individual Differences, 39, 317-329.
Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The causes and consequences
of the need for self-esteem: A terror management theory. In R. F. Baumeister
(Ed.), Public self and private self (pp. 189-212). New York: Springer-Verlag.
Gross, J. J. (2001). Emotion regulation in adulthood: Timing is everything. Current
Directions in Psychological Science, 10, 214-219.
Gross, J. J. (Ed.). (2007). Handbook of emotion regulation. New York: Guildford
Publications, Inc.
Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation
processes: Implications for affect, relationships, and well-being. Journal of
Personality and Social Psychology, 85, 348-362.
Gruter, M. (1986). Ostracism on trial: The limits of individual rights. Ethology and
Sociobiology, 7, 271-279.
193
Haan, N. (1969). A tripartite model of ego-functioning: Values and clinical research
applications. Journal of Nervous and Mental Disease, 148, 14-30.
Hadwin, J. A., Frost, S., French, C. C., & Richards, A. (1997). Cognitive processing
and trait anxiety in typically developing children: Evidence for an
interpretation bias. Journal of Abnormal Psychology, 106, 486-490.
Hankin, B. L., & Abramson, L. Y. (2001). Development of gender differences in
depression: An elaborated cognitive vulnerability-transactional stress theory.
Psychological Bulletin, 127, 773-796.
Hart, K. E., & Hittner, J. B. (1995). Optimism and pessimism: Associations to
coping and anger-reactivity. Personality and Individual Differences, 19, 827-
839.
Headey, B., & Wearing, A. (1989). Personality, life events, and subjective well-
being: Toward a dynamic equilibrium model. Journal of Personality and
Social Psychology, 57, 731-739.
Heron, T. E. (1987). Timeout from positive reinforcement. In I. O. Cooper, T. E.
Heron & H. Merrill (Eds.), Appled behavior analysis. Columbus, OH:
Charles E. Merrill.
Irwin, J. R., & McClelland, G. H. (2003). Negative consequences of dichotomizing
continuous predictor variables. Journal of Marketing Research, 40, 366-371.
James, W. (1890). The principles of psychology New York: Holt.
John, O. P. (1990). The "Big Five" factor taxonomy: Dimensions of personality in
the natural language and in questionnaires. In L. A. Pervin (Ed.), Handbook
of personality: Theory and research. New York: The Guilford Press.
194
John, O. P., Goldberg, L. R., & Angleitner, A. (1984). Better than the alphabet:
Taxonomies of personality-descriptive terms in English, Dutch, and German.
In H. C. J. Bonarius, G. L. M. van Heck & N. G. Smids (Eds.), Personality
psychology in Europe: Theoretical and empirical developments (Vol. 1, pp.
83-100). Lisse: Swets and Zeitlinger.
Kline, R. B. (1998). Principles and practice of structural equation modeling. New
York: Guildford Press.
Lancaster, J. B. (1986). Primate social behavior and ostracism. Ethology and
Sociobiology, 7, 215-225.
Lazarus, R. S. (1966). Psychological stress and the coping process. New York:
McGraw-Hill.
Lazarus, R. S. (1993). Coping theory and research: Past, present, and future.
Psychosomatic Medicine, 55, 234-247.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York:
Springer Publishing Company.
Leary, M. R. (1990). Responses to social exclusion: Social anxiety, jealousy,
loneliness, depression, and low self-esteem. Journal of Social and Clinical
Psychology, 9, 221-229.
Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs, D. L. (1995). Self-esteem as
an interpersonal monitor: The sociometer hypothesis. Journal of Personality
and Social Psychology, 68, 518-530.
195
Lee, C., Ashford, S. J., & Jamieson, L. F. (1993). The effects of type A behavior
dimensions and optimism on coping strategy, health, and performance.
Journal of Organizational Behavior, 14, 143-157.
Levenson, J. L., Kay, R., Monteferrante, J., & Herman, M. V. (1984). Denial
predicts favorable outcome in unstable angina pectoris. Psychosomatic
Medicine, 46, 109-117.
Levine, J., Warrenburg, S., Kerns, R., Schwartz, G., Delaney, R., Fontana, A., et al.
(1987). The role of denial in recovery from coronary heart disease.
Psychosomatic Medicine, 49, 109-117.
Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the Depression Anxiety
Stress Scales. Sydney: The Psychology Foundation of Australia, Inc.
MacCallum, R. C., Zhang, S., Preacher, K. J., & Rucker, D. D. (2002). On the
practice of dichotomization of quantitative variables. Psychological Methods,
7, 19-40.
MacDonald, G., Kingsbury, R., & Shaw, S. (2005). Adding insult to injury: Social
pain theory and response to social exclusion. In K. D. Williams, J. Forgas &
W. von Hippel (Eds.), The social outcast: Ostracism, social exclusion,
rejection, and bullying (pp. 77-90). New York: Psychology Press.
MacKinnon, D. P., Warsi, G., & Dwyer, J. H. (1995). A simulation study of
mediated effect measures. Multivariate Behavioral Research, 30, 41-62.
Magnus, K., Diener, E., Fujita, F., & Pavot, W. (1993). Extraversion and
neuroticism as predictors of objective life events: A longitudinal analysis.
Journal of Personality and Social Psychology, 65, 1046-1053.
196
Mahdi, N. Q. (1986). Pukhtunwali: Ostracism and honor among the Pathan Hill
tribes. Ethology and Sociobiology, 7, 295-304.
Mak, A. S., Blewitt, K., & Heaven, P. C. L. (2004). Gender and personality
influences in adolescent threat and challenge appraisals and depressive
symptoms. Personality and Individual Differences, 36, 1483-1496.
Matthews, G., & Campbell, S. E. (1998). Task-induced stress and individual
differences in coping. In Proceedings of the Human Factors and Ergonomics
Society 42nd Annual Meeting (pp. 821-825). Santa Monica: Human Factors
and Ergonomics Society.
Matthews, G., Emo, A. K., Funke, G., Zeidner, M., Roberts, R. D., Costa, P. T., Jr.,
et al. (2006). Emotional intelligence, personality, and task-induced stress.
Emotion, 12, 96-107.
Matthews, G., Jones, D. M., & Chamberlain, A. G. (1990). Refining the
measurement of mood: The UWIST Mood Adjective Checklist. British
Journal of Psychology, 81, 17-42.
Matthews, G., Schwean, V. L., Campbell, S. E., Saklofske, D., & Mohamed, A. A.
R. (2000). Personality, self-regulation, and adaptation: A cognitive-social
framework. In M. Boekaerts, P. R. Pintrich & M. Zeidner (Eds.), Handbook
of self-regulation (pp. 171-207). San Diego: Academic Press.
McCrae, R. R., & Costa, P. T., Jr. (1985). Updating Norman's "adequate taxonomy":
Intelligence and personality dimensions in natural language and in
questionnaires. Journal of Personality and Social Psychology, 49, 710-721.
197
McCrae, R. R., & Costa, P. T., Jr. (1986). Personality, coping, and coping
effectiveness in an adult sample. Journal of Personality, 54, 385-405.
McCrae, R. R., & Costa, P. T., Jr. (1987). Validation of a five-factor model of
personality across instruments and observers. Journal of Personality and
Social Psychology, 52, 81-90.
McCrae, R. R., & John, O. P. (1992). An introduction to the Five-Factor model and
its applications. Journal of Personality, 60, 175-215.
McInnis, S., & Williams, K. D. (1999). Social ostracism and the invisible worker.
Paper presented at the Society for Australasian Social Psychology, Coolum.
McWilliams, L. A., Cox, B. J., & Enns, M. W. (2003). Use of the Coping Inventory
for Stressful Situations in a clinically depressed sample: Factor structure,
personality correlates, and prediction of distress. Journal of Clinical
Psychology, 59, 1371-1385.
Mead, G. H. (1934). Mind, self, and society. Chicago: University of Chicago Press.
Menninger, K. (1963). The vital balance: The life process in mental health and
illness. New York: Viking.
Monat, A., & Lazarus, R. S. (1991). Stress and coping: An anthology. New York:
Columbia University Press.
Moore, B. S., Sherrod, D. R., Liv, T. J., & Underwood, B. (1979). The dispositional
shift in attribution over time. Journal of Experimental Social Psychology, 15,
553-569.
Morling, B., & Evered, S. (2006). Secondary control reviewed and defined.
Psychological Bulletin, 132, 269-296.
198
Morling, B., & Evered, S. (2007). The construct formerly known as secondary
control: Reply to Skinner (2007). Psychological Bulletin, 133, 917-919.
Moskowitz, J. T., Hult, J. R., Bussolari, C., & Acree, M. (2009). What works in
coping with HIV? A meta-analysis with implications for coping with serious
illness Psychological Bulletin, 135, 121-141.
Muris, P., Meesters, C., & Diederen, R. (2005). Psychometric properties of the Big
Five Questionnaire for Children (BFQ-C) in a Dutch sample of young
adolescents. Personality and Individual Differences, 38, 1757-1769.
Muris, P., Rapee, R. M., Meesters, C., Schouten, E., & Geers, M. (2003). Threat
perception abnormalities in children: The role of anxiety disorders,
symptoms, chronic anxiety, and state anxiety. Journal of Anxiety Disorders,
17, 271-287.
Nezlek, J., Kowalski, R., Leary, M. R., Blevins, T., & Holgate, S. (1997).
Personality moderators of reactions to interpersonal rejection: Depression
and trait self-esteem. Personality and Social Psychology Bulletin, 23, 1235-
1244.
Nezu, A. M., & Carnevale, G. J. (1987). Interpersonal problem solving and coping
reactions of Vietnam veterans with post-traumatic stress disorder. Journal of
Applied Psychology, 96, 155-157.
Norman, W. T. (1963). Toward an adequatetaxonomy of personality attributes:
Replicated factor structure in peer nomination personality ratings Journal of
Abnormal and Social Psychology, 66, 574-583.
199
O'Brien, T. B., & DeLongis, A. (1996). The interactional context of problem-,
emotion-, and relationship-focused coping: The role of the big five
personality factors. Journal of Personality, 64, 775-813.
Ormel, J., & Wohlfarth, T. (1991). How neuroticism, long-term difficulties, and life
situation change influence psychological distress: A longitudinal model.
Journal of Personality and Social Psychology, 60, 744-755.
Park, C. L., Armeli, S., & Tennen, H. (2004). Appraisal-coping goodness of fit: A
daily internet study. Personality and Social Psychology Bulletin, 30, 558-
569.
Park, C. L., Folkman, S., & Bostrom, A. (2001). Appraisals of controllability and
coping in caregivers and HIV+ men: Testing the goodness-of-fit hypothesis.
Journal of Consulting and Clinical Psychology, 69, 481-488.
Parker, J. D. A., & Endler, N. S. (1992). Coping with coping assessment: A critical
review. European Journal of Personality, 6, 321-344.
Parkes, K. R. (1986). Coping in stressful episodes: The role of individual
differences, environmental factors, and situational characteristics. Journal of
Personality and Social Psychology, 51, 1277-1292.
Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and
Social Behavior, 19, 2-21.
Penley, J. A., & Tomaka, J. (2002). Associations among the Big Five, emotional
responses and coping with acute stress. Personality and Individual
Differences, 32, 1215-1128.
200
Pepitone, A., & Wilpizeski, C. (1960). Some consequences of social rejection.
Journal of Abnormal and Social Psychology, 60, 359-364.
Price, D. D. (2000). Psychological and neural mechanisms of the affective
dimension of pain. Science, 288, 1769-1722.
Ptacek, J. T., Pierce, G. R., & Elliott, T. R. (2003). Issues in the study of stress and
coping in rehabilitation settings. Rehabilitation Psychology, 48, 113-124.
Ptacek, J. T., Pierce, G. R., & Thompson, E. L. (2006). Finding evidence of
dispositional coping. Journal of Research in Personality.
Ptacek, J. T., Smith, R. E., Espe, K. A., & Rafferty, B. (1994). Limited
correspondence between daily coping reports and retospective coping recall
Psychological Assessment, 6, 41-49.
Rim, Y. (1987). A comparitive study of two taxonomies of coping styles,
personality, and sex. Personality and Individual Differences, 4, 521-526.
Roesch, S. C., & Weiner, B. (2001). A meta-analytic review of coping with illness:
Do causal attributions matter? . Journal of Psychosomatic Research, 50, 205-
219.
Rosenberg, M. (1965). Society and adolescent child. Princeton, NJ: Princeton
University Press.
Ross, M. (1989). Relation of implicit theories to the construction of personal
histories. Psychological Review, 96, 341-357.
Rothbaum, F., Weisz, J. R., & Snyder, S. S. (1982). Changing the world and
changing the self: A two-process model of perceived control. Journal of
Personality and Social Psychology, 42, 5-37.
201
Rudolph, K. D., Dennig, M. D., & Weisz, J. R. (1995). Determinants and
consequences of childrens coping in the medical setting: Conceptualization,
review, and critique. Psychological Bulletin, 118, 328-357.
Saklofske, D., Kelly, I. W., & Janzen, B. L. (1995). Neuroticism, depression, and
depression proneness. Personality and Individual Differences, 18, 27-31.
Saklofske, D., & Yackulic, R. A. (1989). Personality predictors of loneliness.
Personality and Individual Differences, 10, 467-472.
Sandler, I. N., Tein, J. Y., & West, S. G. (1994). Coping, stress, and the
psychological symptoms of children of divorce: A cross-sectional and
longitudinal study. Child Development, 65, 1744-1763.
Schacter, S. (1951). Deviation, rejection, and communication. Journal of Abnormal
and Social Psychology, 46, 190-207
Schacter, S. (1959). The psychology of affiliation. Stanford, CA: Stanford University
Press.
Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguishing optimism
from neuroticism (and trait anxiety, self-mastery, and self-esteem): A
reevaluation of the life orientation test Journal of Personality and Social
Psychology, 67, 1063-1078.
Scheier, M. F., Weintraub, J. K., & Carver, C. S. (1986). Coping with stress:
Divergent strategies of optimists and pessimists. Journal of Personality and
Social Psychology, 51, 1257-1264.
Schneider, T. R. (2004). The role of neuroticism on psychological and physiological
stress responses. Journal of Experimental Social Psychology, 40, 795-804.
202
Seligman, M. E. P. (1975). Helplessness: On depression, development, and death.
San Francisco: Freeman and Company.
Skinner, E. (1996). A guide to constructs of control. Journal of Personality and
Social Psychology, 71, 549-570.
Skinner, E., Edge, K., Altman, J., & Sherwood, H. (2003). Searching for the
structure of coping: A review and critique of category systems for classifying
ways of coping. Psychological Bulletin, 129, 216-269.
Snoek, J. D. (1962). Some effects of of rejection upon attraction. Journal of
Abnormal and Social Psychology, 64, 175-182.
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in
structural equation models. Sociological Methodology, 13, 290-312.
Solomon, Z., Avitzur, M., & Mikulincer, M. (1989). Coping resources and social
functioning following combat stress reactions: A longitudinal study. Journal
of Social and Clinical Psychology, 8, 87-96.
Solomon, Z., Mikulincer, M., & Avitzur, M. (1988). Coping, locus of control, social
support, and combat-related post-traumatic stress disorder Journal of
Personality and Social Psychology, 55, 279-285.
Solomon, Z., Mikulincer, M., & Benbenishty, R. (1989). Locus of control and
combat related post-traumatic stress disorder: The intervening role of battle
intensity threat appraisal and coping. British Journal of Clinical Psychology,
28, 131-144.
203
Somerfield, M. R., & McCrae, R. R. (2000). Stress and coping research:
Methodological challenges, theoretical advances, and clinical applications.
American Psychologist, 55, 620-625.
Sommer, K. L., Williams, K. D., Ciarocco, N. J., & Baumeister, R. F. (2001). When
silence speaks louder than words: Explorations into the intrapsychic and
interpersonal consequences of social ostracism. Basic and Applied Social
Psychology, 23, 225-243.
Stanton, A. L., Danoff-Burg, S., Cameron, C. L., & Ellis, A. P. (1994). Coping
through emotional approach: Problems of conceptualization and
confounding. Journal of Personality and Social Psychology, 66, 350-362.
Stanton, A. L., Kirk, S. B., Cameron, C. L., & Danoff-Burg, S. (2000). Coping
through emotional approach: Scale construction and verification. Journal of
Personality and Social Psychology, 78, 1150-1169.
Stanton, A. L., Parsa, A., & Austenfeld, J. L. (2002). The adaptive potential of
coping through emotional approach. In C. R. Snyder & S. J. Lopez (Eds.),
Handbook of positive psychology (pp. 148-158). New York: Oxford
University Press.
Stone, A. A., Kessler, R. C., & Haythornthwaite, J. A. (1991). Measuring daily
events and experiences: Decisions for researchers. Journal of Personality,
59, 575-607.
Stone, A. A., & Neale, J. (1984). New measure of daily coping: Development and
preliminary results. Journal of Personality and Social Psychology, 46, 892-
906.
204
Strutton, D., & Lumpkin, J. (1992). Relationship between optimism and coping
strategies in the work environment. Psychological Reports, 71, 1179-1186.
Suls, J., David, J. P., & Harvey, J. H. (1996). Personality and coping: Three
generations of research. Journal of Personality, 64, 711-735.
Taghavi, M. R., Moradi, A. R., Neshat-Doost, H. T., Yule, W., & Dagleish, T.
(2000). Interpretation of ambiguous emotional information in clinically
anxious children and adolescents. Cognition & Emotion, 14, 809-822.
Tajfel, H. (1972). Experiments in a vacuum. In J. Israel & H. Tajfel (Eds.), The
context of social psychology: A critical assessment. London: Academic
Press.
Tennen, H., Afleck, G., Armeli, S., & Carney, M. A. (2000). A daily process
approach to coping: Linking theory, research, and practice. American
Psychologist, 55, 626-636.
Terry, D. J. (1991). Coping resources and situational appraisals as predictors of
coping behavior. Personality and Individual Differences, 12, 1031-1047.
Terry, D. J. (1994). Determinants of coping: The role of stable and situational
factors. Journal of Personality and Social Psychology, 66, 895-910.
Terry, D. J., & Hynes, G. J. (1998). Adjustment to a low-control situation:
Reexamining the role of coping responses. Journal of Personality and Social
Psychology, 74, 1078-1092.
Tresselt, M. E., & Mayzner, M. S. (1966). Normative solution times for a sample of
134 solution words and 378 associated anagrams. Psychonomic Monograph
Supplements, 1, 293-298.
205
Tupes, E. C., & Christal, R. E. (1992). Recurrent personality factors based on trait
ratings. Journal of Personality, 60, 225-251.
Twenge, J. M., Baumeister, R. F., Tice, D. M., & Stucke, T. S. (2001). Social
exclusion causes of self-defeating behavior. Journal of Personality and
Social Psychology, 83, 606-615.
Uchino, B. N., Cacioppo, J. T., & Kiecolt-Glaser, J. K. (1996). The relationship
between social support and physiological processes: A review with emphasis
on underlying mechanisms and implications for health. Psychological
Bulletin, 119, 488-531.
Vaillant, G. E. (1977). Adaption to life Boston, MA: Little, Brown.
Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the
measurement invariance literature: Suggestions, practices, and
recommendations for organizational research. Organizational Research
Methods, 3, 4-69.
Vollrath, M. (2001). Personality and stress. Scandinavian Journal of Psychology, 42,
335-347.
Vollrath, M., & Torgersen, S. (2000). Personality types and coping. Personality and
Individual Differences, 29, 367-378.
Vollrath, M., Torgersen, S., & Alnaes, R. (1995). Personality as long-term predictor
of coping. Personality and Individual Differences, 18, 117-125.
Waters, A. M., Craske, M. G., Bergman, R. L., & Treanor, M. (2008). Threat
interpretation bias as a vulnerability factor in childhood anxiety disorders.
Behaviour, Research, and Therapy, 46, 39-47.
206
Watson, D., & Clark, L. (1988). Development and validation of brief measures of
positive and negative affect: The PANAS scales Journal of Personality and
Social Psychology, 54, 1063-1070.
Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to
experience aversive emotional states. Psychological Bulletin, 96, 465-490.
Watson, D., & Hubbard, B. (1996). Adaptational style and dispositional structure:
Coping in the context of the five-factor model. Journal of Personality, 64,
737-774.
Watson, D., & Pennebaker, J. (1989). Health complaints, stress, and distress:
Exploring the central role of negative affectivity. Psychological Review, 96,
234-254.
Williams, K. D. (1997). Social ostracism. In R. Kowalski (Ed.), Aversive
interpersonal behaviors (pp. 133-170). New York: Plenum Press.
Williams, K. D. (2001). Ostracism: The power of silence. New York: Guilford
Publications.
Williams, K. D., Bernieri, F. J., Faulkner, S. J., Gada-Jain, N., & Grahe, J. E. (2000).
The Scarlet Letter Study: Five days of social ostracism. Journal of Personal
and Interpersonal Loss, 5, 19-63.
Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). CyberOstracism: Effects of
being ignored over the internet. Journal of Personality and Social
Psychology, 79, 748-762.
207
Williams, K. D., Govan, C. L., Croker, V., Tynan, D., Cruickshank, M., & Lam, A.
(2002). Investigations into differences between social- and cyberostracism.
Group Dynamics: Theory, Research, and Practice, 6, 65-77.
Williams, K. D., & Jarvis, B. (2006). Cyberball: A program for use in research on
interpersonal ostracism and acceptance. Behavior Research Methods, 38,
174-180.
Williams, K. D., & Sommer, K. L. (1997). Social Ostracism by coworkers: Does
rejection lead to loafing or compensation. Personality and Social Psychology
Bulletin, 23, 693-706.
Williams, K. D., Wheeler, L., & Harvey, J. (2001). Inside the social mind of the
ostracizer. In J. Forgas, K. D. Williams & L. Wheeler (Eds.), The social
mind: Cognitive and motivational aspects of interpersonal behavior (pp.
294-320). New York: Cambridge University Press.
Williams, K. D., & Zadro, L. (2001). Ostracism: On being ignored, excluded, and
rejected. In M. R. Leary (Ed.), Interpersonal rejection (pp. 21-53). New
York: Oxford University Press.
Woods, D. (1978). Biko. New York: Paddington Press.
Zadro, L., & Williams, K. D. (1999). Long-term effects of giving or receiving the
silent treatment. Paper presented at the Society for Australasian Social
Psychology, Coolum.
Zadro, L., Williams, K. D., & Richardson, R. (2004). How low can you go?
Ostracism by computer is sufficient to lower self-reported levels of
208
belonging, control, self-esteem, and meaningful existence. Journal of
Experimental Social Psychology, 40, 560-567.
Zakowski, S. G., Hall, M. H., Klein, L. C., & Baum, A. (2001). Appraised control,
coping, and stress in a community sample. Annals of Behavioral Medicine,
23, 158-165.
Zeidner, M. (1994). Personal and contextual determinants of coping and anxiety in
an evaluative situation: A prospective study. Personality and Individual
Differences, 16, 899-918.
Zeidner, M., & Saklofske, D. (1996). Adaptive and maladaptive coping. In M.
Zeidner & N. S. Endler (Eds.), Handbook of coping: Theory, research,
applications. New York: John Wiley and Sons, Inc.
209
APPENDICES
210
Appendix A: Exam Appraisal Scale
Identification Number ___________________ Age _______________________________ Gender _______________________________ Year of University Study _____________ Faculty/Course ___________________ This questionnaire aims to assess your perception of a stressful event – in this case the upcoming examinations. Please answer each item in relation to your examinations. Please answer every item. The response scale is 1: not at all; 2: somewhat; 3: moderately; 4: extremely 1. To what extent do you find the upcoming event
stressful? 1 2 3 4 2. How challenging do you perceive the upcoming
event to be? 1 2 3 4 3. To what extent do you feel you have control over the
outcome? 1 2 3 4 4. To what extent can you predict how well you will
perform? 1 2 3 4 5. To what extent are you worried about the upcoming
event? 1 2 3 4 6. How confident are you about how you will perform? 1 2 3 4 7. How threatening do you perceive the upcoming event to be? 1 2 3 4 8. How important do you perceive the upcoming event
to be? 1 2 3 4 Subscales: Stress: Items 1, 2, 5, 7, and 8. Control: Items 3, 4, and 6.
211
Appendix B: COPE (Dispositional Version)
We are interested in how people respond when they confront difficult or stressful events in their lives. There are lots of ways to try to deal with stress. This questionnaire asks you to indicate what you generally do and feel when you experience stressful events. Obviously, different events bring out somewhat different responses, but think about what you usually do when you are under a lot of stress. Then respond to each of the following items by choosing one number for each, using the response choices listed below. 1: I usually don’t do this at all 2: I usually do this a little bit 3: I usually do this a medium amount 4: I usually do this a lot Please try to respond to each item separately in your mind from each other item. Choose your answers thoughtfully, and make answers that are as true FOR YOU as you can. Please answer every item. There are no ‘right’ or ‘wrong’ answers’, so choose the most accurate answer for YOU – not what you think ‘most people’ would say or do. Indicate what YOU usually do when you experience a stressful event. 1. I try to grow as a person as a result of the experience 1 2 3 4
2. I turn to work or other substitute activities to take my mind off things 1 2 3 4
3. I get upset and let my emotions out 1 2 3 4
4. I try to get advice from someone about what to do 1 2 3 4
5. I concentrate my efforts on doing something about it 1 2 3 4
6. I say to myself “this isn’t real” 1 2 3 4
7. I put my trust in God 1 2 3 4
8. I laugh about the situation 1 2 3 4
9. I admit to myself that I can’t deal with it, and give
up trying 1 2 3 4
10. I restrain myself from doing anything too quickly 1 2 3 4
11. I discuss my feelings with someone 1 2 3 4
12. I use alcohol or drugs to make myself feel better 1 2 3 4
13. I get used to the idea that this is happening 1 2 3 4
14. I talk to someone to find out more about the situation 1 2 3 4
15. I keep myself from getting distracted by other thoughts or activities 1 2 3 4
212
16. I daydream about things other than this event 1 2 3 4
17. I get upset, and am really aware of it 1 2 3 4
18. I seek God’s help 1 2 3 4
19. I make a plan of action 1 2 3 4
20. I make jokes about it 1 2 3 4
21. I accept that this will happen and that it can’t be changed 1 2 3 4
22. I hold off doing anything about it until the Situation permits 1 2 3 4
23. I try to get emotional support from friends and relatives 1 2 3 4
24. I just give up trying to reach my goal 1 2 3 4
25. I take additional action to try and get rid of the problem 1 2 3 4
26. I lose myself for a while by drinking alcohol or taking drugs 1 2 3 4
27. I refuse to believe that this is happening 1 2 3 4
28. I let my feelings out 1 2 3 4
29. I try to see it in a different light, to make it seem more positive 1 2 3 4
30. I talk to someone who could do something concrete about the problem 1 2 3 4
31. I sleep more than usual 1 2 3 4
32. I try to come up with a strategy about what to do 1 2 3 4
33. I focus on dealing with this problem and, if necessary let other things slide a little 1 2 3 4
34. I get sympathy and understanding from someone 1 2 3 4
35. I drink alcohol or take drugs, in order to think about it less 1 2 3 4
36. I kid around about it 1 2 3 4
37. I give up the attempt to get what I want 1 2 3 4
38. I look for something good in what is happening 1 2 3 4
39. I think about how I might best handle the situation 1 2 3 4
40. I pretend that it isn’t really happening 1 2 3 4
41. I make sure not to make matters worse by acting to soon 1 2 3 4
213
42. I try hard to prevent other things from interfering with my efforts at dealing with this 1 2 3 4
43. I go to the movies or watch television, to think about it less 1 2 3 4
44. I accept the reality of the fact that it is happening 1 2 3 4
45. I ask people who have had similar experiences what they did 1 2 3 4
46. I feel a lot of emotional distress and I find myself expressing those feelings a lot 1 2 3 4
47. I take direct action to get around the problem 1 2 3 4
48. I try to find comfort in my religion 1 2 3 4
49. I force myself to wait for the right time to do something 1 2 3 4
50. I make fun of the situation 1 2 3 4
51. I reduce the amount of time I put into solving the problem 1 2 3 4
52. I talk to someone about how I feel 1 2 3 4
53. I use drugs or alcohol to help me get through it 1 2 3 4
54. I learn to live with it 1 2 3 4
55. I put aside other activities in order to concentrate on this 1 2 3 4
56. I think hard about what steps to take 1 2 3 4
57. I act as though it is not happening 1 2 3 4
58. I do what has to be done, one step at a time 1 2 3 4
59. I learn something from the experience 1 2 3 4
60. I pray more than usual 1 2 3 4
214
Appendix C: COPE (Situational Version)
We are interested in how people respond when they confront difficult or stressful events in their lives. There are lots of ways to try to deal with stress. This questionnaire asks you what you do and feel when you experience a specific stressful event – in this case the upcoming exams. Please respond to each item by indicating the extent to which you are using it to cope with the stress associated with the exams. Please try to respond to each item separately. Choose your answers thoughtfully, and make answers that are as true FOR YOU as you can. Please answer every item. There are no ‘right’ or ‘wrong’ answers’, so choose the most accurate answer for YOU – not what you think ‘most people’ would say or do. Please indicate what YOU are doing to cope with the upcoming exams. The response scale is – 1: I am not doing this at all 2: I am doing this a little bit 3: I am doing this a medium amount 4: I am doing this a lot 1. I try to grow as a person as a result of the experience 1 2 3 4
2. I turn to work or other substitute activities to take my mind off things 1 2 3 4
3. I get upset and let my emotions out 1 2 3 4
4. I try to get advice from someone about what to do 1 2 3 4
5. I concentrate my efforts on doing something about it 1 2 3 4
6. I say to myself “this isn’t real” 1 2 3 4
7. I put my trust in God 1 2 3 4
8. I laugh about the situation 1 2 3 4
9. I admit to myself that I can’t deal with it, and give up trying 1 2 3 4
10. I restrain myself from doing anything too quickly 1 2 3 4
11. I discuss my feelings with someone 1 2 3 4
12. I use alcohol or drugs to make myself feel better 1 2 3 4
13. I get used to the idea that this is happening 1 2 3 4
14. I talk to someone to find out more about the situation 1 2 3 4
15. I keep myself from getting distracted by other thoughts or activities 1 2 3 4
16. I daydream about things other than this event 1 2 3 4
17. I get upset, and am really aware of it 1 2 3 4
215
18. I seek God’s help 1 2 3 4
19. I make a plan of action 1 2 3 4
20. I make jokes about it 1 2 3 4
21. I accept that this will happen and that it can’t be changed 1 2 3 4
22. I hold off doing anything about it until the Situation permits 1 2 3 4
23. I try to get emotional support from friends and relatives 1 2 3 4
24. I just give up trying to reach my goal 1 2 3 4
25. I take additional action to try and get rid of the problem 1 2 3 4
26. I lose myself for a while by drinking alcohol or taking drugs 1 2 3 4
27. I refuse to believe that this is happening 1 2 3 4
28. I let my feelings out 1 2 3 4
29. I try to see it in a different light, to make it seem more positive 1 2 3 4
30. I talk to someone who could do something concrete about the problem 1 2 3 4
31. I sleep more than usual 1 2 3 4
32. I try to come up with a strategy about what to do 1 2 3 4
33. I focus on dealing with this problem and, if necessary let other things slide a little 1 2 3 4
34. I get sympathy and understanding from someone 1 2 3 4
35. I drink alcohol or take drugs, in order to think about it less 1 2 3 4
36. I kid around about it 1 2 3 4
37. I give up the attempt to get what I want 1 2 3 4
38. I look for something good in what is happening 1 2 3 4
39. I think about how I might best handle the situation 1 2 3 4
40. I pretend that it isn’t really happening 1 2 3 4
41. I make sure not to make matters worse by acting to soon 1 2 3 4
42. I try hard to prevent other things from interfering with my efforts at dealing with this 1 2 3 4
216
43. I go to the movies or watch television, to think about it less 1 2 3 4
44. I accept the reality of the fact that it is happening 1 2 3 4
45. I ask people who have had similar experiences what they did 1 2 3 4
46. I feel a lot of emotional distress and I find myself expressing those feelings a lot 1 2 3 4
47. I take direct action to get around the problem 1 2 3 4
48. I try to find comfort in my religion 1 2 3 4
49. I force myself to wait for the right time to do something 1 2 3 4
50. I make fun of the situation 1 2 3 4
51. I reduce the amount of time I put into solving the problem 1 2 3 4
52. I talk to someone about how I feel 1 2 3 4
53. I use drugs or alcohol to help me get through it 1 2 3 4
54. I learn to live with it 1 2 3 4
55. I put aside other activities in order to concentrate on this 1 2 3 4
56. I think hard about what steps to take 1 2 3 4
57. I act as though it is not happening 1 2 3 4
58. I do what has to be done, one step at a time 1 2 3 4
59. I learn something from the experience 1 2 3 4
60. I pray more than usual 1 2 3 4
217
Appendix D: Neuroticism Scale
On this page, there are phrases describing people's behaviours. Please use the rating scale below to describe how accurately each statement describes you. Describe yourself as you generally are now, not as you wish to be in the future. Describe yourself as you honestly see yourself, in relation to other people you know of the same sex as you are, and roughly your same age. So that you can describe yourself in an honest manner, your responses will be kept in absolute confidence. Please read each statement carefully, and then circle the number that corresponds to the description on the scale.
Response Options
1: Very Inaccurate 2: Moderately Inaccurate 3: Neither Inaccurate nor Accurate 4: Moderately Accurate 5: Very Accurate
I …
1) Am relaxed most of the time 1 2 3 4 5
2) Get stressed out easily 1 2 3 4 5
3) Seldom feel blue 1 2 3 4 5
4) Worry about things 1 2 3 4 5
5) Am easily disturbed 1 2 3 4 5
6) Get upset easily 1 2 3 4 5
7) Change my mood a lot 1 2 3 4 5
8) Have frequent mood swings 1 2 3 4 5
9) Get irritated easily 1 2 3 4 5
10) Often feel blue 1 2 3 4 5
218
Appendix E: Cyberball Appraisal Scale
This questionnaire aims to assess your experiences during the cyberball game. Please answer each question specifically in relation to your experiences during the cyberball game. The response scale is – 1) not at all; 2) somewhat; 3) moderately; 4) extremely
1) To what extent did you feel you had control over the game? 1 2 3 4 2) During the game, to what extent did you feel left out? 1 2 3 4 3) During the game, to what extent did you feel frustrated? 1 2 3 4 4) To what extent did you feel you had influence over the game? 1 2 3 4 5) During the game, to what extent did you feel excluded? 1 2 3 4 6) To what extent did you feel you had an impact on the game? 1 2 3 4 7) During the game, to what extent did you feel rejected? 1 2 3 4 8) During the game, to what extent did you feel annoyed 1 2 3 4
Subscales: Excluded: Items 2, 3, 5, 7, and 8. Control: Items 1, 4, and 6.
219
Appendix F: CITS (Situational Version)
CITS-S
Think about how you dealt with any difficulties or problems that arose while you were performing the task you have just performed. Below are listed some options for dealing with problems such as poor performance or negative reactions to doing the task. Please indicate how much you used each option, specifically as a deliberately chosen way of dealing with problems. To answer circle one of the following answers: Extremely = 4 Very much = 3 Somewhat = 2 A little bit = 1 Not at all = 0
I ...
1. Worked out a strategy for successful performance 0 1 2 3 4
2. Worried about what I would do next 0 1 2 3 4
3. Stayed detached or distanced from the situation 0 1 2 3 4
4. Decided to save my efforts for something more worthwhile 0 1 2 3 4
5. Blamed myself for not doing better 0 1 2 3 4
6. Became preoccupied with my problems 0 1 2 3 4
7. Concentrated hard on doing well 0 1 2 3 4
8. Focused my attention on the most important parts of the task 0 1 2 3 4
9. Acted as though the task wasn't important 0 1 2 3 4
10. Didn't take the task too seriously 0 1 2 3 4
11. Wished that I could change what was happening 0 1 2 3 4
12. Blamed myself for not knowing what to do 0 1 2 3 4
13. Worried about my inadequacies 0 1 2 3 4
14. Made every effort to achieve my goals 0 1 2 3 4
15. Blamed myself for becoming too emotional 0 1 2 3 4
16. Was single-minded and determined in my efforts to overcome any problems 0 1 2 3 4
17. Gave up the attempt to do well 0 1 2 3 4
18. Told myself it wasn't worth getting upset 0 1 2 3 4
19. Was careful to avoid mistakes 0 1 2 3 4
20. Did my best to follow the instructions for the task 0 1 2 3 4
21. Decided there was no point in trying to do well 0 1 2 3 4
220
Scoring Instructions
Summate item scores as follows:
Task-focus = 1 + 7 + 8 + 14 + 16 + 19 + 20
Emotion-focus = 2 + 5 + 6 + 11 + 12 + 13 + 15
Avoidance = 3 + 4 + 9 + 10 +17 + 18 +21
221
Appendix G: Reliabilities of Measures in Studies 3-5
Table 23. Internal consistencies of the subscales of the MACL, the RSES, the subscales of the Cyberball Appraisal Scale, and the subscales of the CITS-S obtained in Studies 3, 4, and 5
Cronbach’s α Study 3 Study 4 Study 5 MACL
Energetic Arousal .76 .79 .82 Tense Arousal .88 .88 .89 Hedonic Tone .95 .95 .92
RSES .91 .93 .87 Cyberball Appraisal Scale
Appraisal (Excluded) .94 .91 n/a Appraisal (Control) .90 .91 n/a
CITS-S Task-Focused Coping .83 .74 .75
Emotion-Focused Coping .90 .83 .92 Avoidance Coping .54 .45 .45
222
Appendix H: Anagram Appraisal Scale
This questionnaire aims to assess your experiences during the Anagram task. Please answer each question by marking the appropriate point on the line. 1. How much influence do you think you had in the anagram task? Not at All _________________________________________________ Extremely 2. How stressed did you feel during the anagram task? Not at All _________________________________________________ Extremely 3. To what extent did the anagram task make you feel powerless? Not at All _________________________________________________ Extremely 4. How anxious did you feel during the anagram task? Not at All _________________________________________________ Extremely 5. Would you agree that the anagram task is manageable? Not at All _________________________________________________ Extremely 6. How tense did you feel during the anagram task? Not at All _________________________________________________ Extremely Subscales: Stress: Items 2, 4, and 6. Control: Items 1, 3, and 5.
223
Appendix I: Publications
The appendices could not be included in the digital version of this thesis for copyright reasons. Please refer to the physical copy of the thesis, held in the University Library.