|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
COPING WITH THE COMPETING DEMANDS OF SCHOOL AND SPORT:
A C OGNITIVE-MO TIVATIONAL -RELATIONAL
EXAMINATION OF CHRONIC STRESS
DANIEL EDWARD TUCCITTO
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
UNIVERSITY OF FLORIDA
Daniel Edward Tuccitto
To my parents, who cultivated my intellectual curiosity and provided the unconditional support
necessary for me to achieve this milestone
I thank the chair and members of my supervisory committee for their mentoring, patience,
and time commitment, the UF Director of Sport Clubs for assisting me with participant
recruitment, and the club sport athletes who completed my measurement protocol, all of whom
made the completion of this proj ect possible.
TABLE OF CONTENTS
ACKNOWLEDGMENT S .............. ...............4.....
LI ST OF T ABLE S ................. ...............8................
LI ST OF FIGURE S .............. ...............9.....
AB S TRAC T ............._. .......... ..............._ 10...
INTRODUCTION .............. ...............12....
Need for the Study .............. .. ............ ... ...............13....
Problem Statement, Study Purposes, and Hypotheses ................. .............................19
LITERATURE REVIEW .............. ...............22....
Stress............... ... ... .............2
P sy chol ogical Stress .............. ...............23....
Acute vs. Chronic Stress............... ...............26.
Apprai sal ................ .... ....... ...............28.......
Types of Appraisal ................ ........ ......... ........ ......... ................29
Primary appraisal ............... ...............2
Second ary appraisal ............... ...............3
Environmental variabl es............... ...............3
Person variables................. ..............3
Achievement goal orientations ................. ...............3.. 4......... ....
Summary of Apprai sal ................. ...............36................
Coping............... .. ...............37
Ego Approach ................. ...............38.................
Trait Approach............... ...............3
Process Approach .............. ...............40....
Functions of coping ................. ...............40................
Coping effectiveness .............. ...............41....
Process methodology................ ............. .. .......4
Summary of the C ognitive-Motivational -Relational Theory of Emotion ................... ....43
Coping Measurement and Research in General and Sport Psychology ................ ...............44
Coping Measurement in General Psychology .............. ...............44....
The Ways of Coping checklist .............. ...............44....
The COPE inventory .............. ..... ...............46..
Coping Measurement in Sport Psychology .............. ...............50....
Modifications of the Ways of Coping checklist ................. ......... ................50
Modifications of the COPE inventory ................. ...............54........... ...
Summary of Coping Measurement............... ..............5
Coping as a style .............. .. .. ...............58.
Coping as a personality disposition............... ..............6
Conditional coping .............. ... .... ......... .......... .............6
Summary of research on the structural aspects of coping ................. ................ ..65
Research on the Process Aspects of Coping ................. ...............66........... ..
Dimensionality of coping ................. ...............66................
Dynamics of coping .............. ...............68....
Coping effectiveness .............. .. .. ...............70
Summary of research on the process aspects of coping ................. ............... .....78
Study Rationale............... ........ ................7
Achievement Goal Orientations and Primary Appraisal ................. .................8
Achievement Goal Orientations and Secondary Appraisal ................ ............. .......81
Achievement Goal Orientations, Coping, and Affect .............. ...............82....
Primary Appraisal, Secondary Appraisal, and Coping............... ...............83.
Primary Appraisal and Affect............... .... .......... ........8
Coping, Perceived Coping Effectiveness, and Affect .............. ...............85....
Hypotheses .............. ...............86....
M ETHOD .............. ...............90....
Participants and Procedure .............. ...............90....
Study M easures................. ..............9
Demographics Questionnaire ............... ..................... ...............9
The Task and Ego Orientation for Sport Questionnaire (TEOSQ) ..........................92
Appraisal Measures ................ ...............93.................
The COPE Inventory .............. .... ...............94..
Perceived Coping Effectiveness Measure .............. ................ ............9
The Positive Affect and Negative Affect Schedule (PANAS) ................. ................ ..97
Data Analyses ................. ...............98.......... .....
Preliminary Analyses................ ...... ....... .......9
Psychometric Analyses (The Measurement Model)............... ...............99.
SEM Analyses (The Structural Model) .............. ...............102....
RE SULT S .............. ...............104....
Preliminary Analy ses ................. ...............104................
Descriptive Statistics .............. ...............104....
Demographic Effects ................. ...............104................
Demographic Correlations............... .............10
Correlations Between Study Variables............... ...............10
Formulation of the Proposed Structural Model ................ ..............................106
Psychometric Analyses ........._.___..... ..._. ...............107....
Overall Construct Validity .............. ...............107....
Convergent Validity .............. ...............107....
Discriminant Validity ............... ...............108...
Internal Consistency Reliability .............. ...............108....
External Consistency Reliability .............. ...............109....
SEM Analyses ................. ...............110......... ......
DI SCU SSION ................. ................. 1......... 18....
Empirical Issues ................. ...............119................
Descriptive Statistics ................. ...............119......... ......
Demographic Variables ................... ...............120......... ......
Average Variance Extracted (AVE) ....... ......._ .............._ ...........12
Substantive Issues ..............._ ...... ... ......... ............12
The C-M-R Theory of Emotion ..............._ ...............122......_....
Coping Effectiveness............... .............12
Coping Measurement. ............... ...............125....._.......
Limitations ..............._ ...............126....._.......
Sampling ..............._ ...............126....._.......
Procedure ..............._ ...............126....._.......
Statistical Analyses............... ...............12
Applications ..............._ ...............129....._.......
DEMOGRAPHICS QUESTIONNAIRE............... .............13
TASK AND EGO ORIENTATION SCALE FOR SPORT QUESTIONNAIRE (TEOSQ) ......133
APPRAISAL MEASURES .............. ...............135....
COPINTG ORIENTATION FOR PROBLEM EXPERIENCES (COPE) .............. ..................137
POSITIVE AFFECT NEGATIVE AFFECT SCHEDULE (PANAS) ................ ................ ..140
LIST OF REFERENCES ................. ...............141................
BIOGRAPHICAL SKETCH ................. ...............154......... ......
LIST OF TABLES
Table 3-1. Demographic characteristics of participants (N = 223). ................ ......................103
Table 4-1. Means, standard deviations, and skewness values for the study variables. ................111
Table 4-2. Intercorrelations among aspects of achievement motivation, appraisal, second-
order coping, and affect ........... ..... ._ ...............112...
Table 4-3. Factor loadings (h), uniqueness (E), and average variance explained (AVE) values
for the measurement model ................. ...............113...............
Table 4-4. Estimated intercorrelations and coefficient alphas for first-order latent variables
in the measurement model ........... ..... .._ ...............115..
Table 4-5. Path coefficients (P), standard errors (SE), and t-values for the structural model .....116
LIST OF FIGURES
Figure 2-1. Predicted relationships between the maj or constructs of the cognitive-
motivational-relational theory of emotion. ............. ...............88.....
Figure 2-2. Proposed path model of the relationships between goal orientations, primary
appraisal, secondary appraisal, coping, perceived coping effectiveness, and affect.
Dotted lines represent negative relationships. ............. ...............89.....
Figure 4-1. Final model tested in the psychometric and structural equation modeling
analyses. dotted lines represent negative relationships ................. ........................117
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
COPING WITH THE COMPETING DEMANDS OF SCHOOL AND SPORT:
A C OGNITIVE-MO TIVATIONAL -RELATIONAL
EXAMINATION OF CHRONIC STRESS
Daniel Edward Tuccitto
Chair: Peter R. Giacobbi, Jr.
Major Department: Applied Physiology and Kinesiology
The purposes of my study were to examine the viability of the cognitive-motivational-
relational theory of emotion as a research tool in sport psychology, and to test a theoretical
model of the explanatory relationships between club sport athletes' achievement goal
orientations, cognitive appraisals, coping efforts, and affective experiences. A sample of 223
club sport athletes at a large university in the Southeastern United States described a situation in
the preceding 7 days in which they had difficulty balancing their academic and athletic time
demands. In addition, they completed the Task and Ego Orientation in Sport Questionnaire
(TEOSQ), the threat and challenge subscales of the Stress Appraisal Measure (SAM), the
personal control sub scale of the Causal Dimension Scale II (CDS-II), sub scales of the Coping
Orientation for Problem Experiences (COPE) inventory reflecting task-focused and avoidance-
focused coping, items attached to the COPE reflecting the perceived effectiveness of task-
focused and avoidance-focused coping, and the Positive Affect Negative Affect Schedule
(PANAS). I used structural equation modeling techniques to assess the psychometric properties
of the measurement protocol, to measure the extent to which my theoretical model accurately
reflected the data, and to estimate the explanatory relationships predicted by the model. Results
indicated first that the measurement protocol I employed in my study exhibited acceptable
factorial, convergent, and discriminant validity, along with acceptable internal and external
consistency reliability. Second, my theoretical model fit the data well, thereby providing further
justification for use of cognitive-motivational theory as a guiding framework in sport psychology
research. Third, with respect to the specific explanatory relationships estimated by my model, I
found that task orientation significantly predicted challenge appraisal, which in turn predicted
task-focused coping, with perceived task-focused coping effectiveness acting as a mediator of
the relationship between task-focused coping and positive affect. While my study filled several
gaps in the extant literature, it nevertheless can be improved upon in future studies by obtaining a
random sample, employing procedures more consonant with the ipsative-normative methodology
advocated in the cognitive-motivati onal-relational theory of emotion, and conducting a more
sophisticated test of moderation within a structural equation modeling framework. However,
despite these limitations, my study has clear applications with respect to stress interventions and
the development of tailored services to help club sport athletes deal with balancing their
competing academic and athletic time demands.
Up to 40% of college students encounter serious difficulties and fail to complete their
degrees (Pantages & Creedon, 1978; Zitzow, 1984). In addition, 10% of college students are
diagnosed with depression annually (National Mental Health Association, 2004). Researchers in
psychology have identified several sources of stress that may contribute to these disconcerting
statistics regarding the psychological well-being of college students (Ross, Neibling, & Heckert,
1999; Tyrrell, 1992). They include changes in eating or sleeping habits, lack of vacations/breaks,
increased workload, new responsibilities, falling behind with coursework, finding the motivation
to study, time pressures, Einancial worries, and concerns about academic ability (Ross et al.;
Tyrrell). In addition to academic sources of stress, college students who also participate in
competitive athletics (e.g., recreation, club, and intercollegiate sports) are confronted with
additional stressors unique to their sport involvement. For example, student athletes have
indicated that their maj or sources of stress include being mentally and physically overwhelmed,
training intensity, high performance expectations, and interpersonal relationships in their sport
(Giacobbi et al., 2004; Tracey & Corlett, 1995). It seems clear then that athletics and academics
each present unique sources of psychological stress for student-athletes. Indeed, researchers have
shown considerable interest in how student-athletes deal separately with athletic (Giacobbi et al.,
2004; Giacobbi & Weinberg, 2000; Kimball & Freysinger, 2003; Ntoumanis & Biddle, 1998;
Ntoumanis, Biddle, & Haddock, 1999; Sellers, 1995; Williams & Krane, 1992) and academic
stressors. However, an understanding of the ways that individuals cope with specific athletic and
academic stressors provides only a small piece of the puzzle. What is also needed is an insight
into the ways that student-athletes cope with these stressors in combination (e.g., managing the
dual time demands of athletics and academics). I attempted to address this issue by examining
the coping thoughts and behaviors of club sport athletes.
Although my study was more theory- than population-driven, I based my decision to
sample club sport athletes on two grounds. First, sport coping researchers have focused almost
exclusively on student-athletes at the NCAA level. While the experiences of NCAA athletes are
important, there exists a broader population of college students who are both highly skilled and
highly involved in their sports. Understanding how these individuals cope with the stress of
managing athletic and academic time demands may enhance the ability of practitioners to
provide the greatest good to the greatest number. Indeed, at the University of Florida alone there
are 36 sport clubs while only 18 NCAA teams. Of these 36 sport clubs there are approximately
1,700 students who participate in a variety of team (i.e., soccer, flag football, basketball) and
individual (i.e., golf, tennis) sports.
Second, although the athletic demands placed on club sport athletes may not be as great as
those placed on their NCAA counterparts, it is likely that club sport athletes do in fact experience
stress related to balancing their athletic and academic roles. For example, club sport athletes
compete regionally and nationally, and practice regularly, all while having to keep up with their
studies. Therefore, I sampled club sport athletes in my study because they represent a relatively
untapped research population and they are likely to experience and cope with a combination of
academic and athletic sources of stress.
Need for the Study
Over the past decade, several studies have examined the coping behaviors of student-
athletes (Anshel, Williams, & Williams, 2000; Giacobbi & Weinberg, 2000; Ntoumanis &
Biddle, 1998). However, the majority of these studies have focused on the ways student-athletes
cope with competitive stress. For example, Ntoumanis and Biddle studied the coping behaviors
of intercollegiate athletes in response to a stressful competitive situation they had encountered in
the past. Similarly, Anshel and his colleagues required student-athletes to indicate their coping
responses to a predetermined list of common competitive stressors. In contrast, there has been a
dearth of research in the extant literature regarding the manner in which university athletes cope
with stressors outside of competition (See Giacobbi et al., 2004; Petrie & Stoever, 1997; Tracey
& Corlett, 1995, for exceptions). One source of non-competitive stress for university athletes that
has been identified in the literature was alluded to above, i.e., managing athletic and academic
time demands (Giacobbi et al., 2004). In my study, I examined the coping thoughts and
behaviors that club sport athletes employ to deal with this non-competitive stressor.
Additionally, the overall study of coping in sport, regardless of the population being
studied, has been limited exclusively to examination of responses to acute stress (Anshel &
Anderson, 2002; Bouffard & Crocker, 1992; Crocker & Graham, 1995; Crocker & Isaak, 1997;
Dugdale, Eklund, & Gordon, 2002; Eubank & Collins, 2000; Gaudreau, Blondin, & Lapierre,
2002; Gaudreau, Lapierre, & Blondin, 2001; Gould, Finch, & Jackson, 1993; Hammermeister &
Burton, 2001; Haney & Long, 1995; Krohne & Hindel, 1988). Anshel (1996) offered that acute
stress results from the interpretation of a particular event as stressful, while chronic stress results
from the continued interpretation of the same event as stressful over a prolonged period of time.
The distinction between acute and chronic stress has important implications for coping
researchers, as Gottlieb (1997) suggested that acute and chronic stressors may require differential
coping strategies. Therefore, there is clearly a need for research on the coping efforts of athletes
in response to chronic stress, and my study addressed this issue.
The most widely accepted conceptual framework for examining psychological stress is
Lazarus' s cognitive-motivational-relational (C-M-R) theory of emotion (1991b, 1999). In his
theory, Lazarus (1991b, 1999) posited that when individuals encounter a potentially stressful
situation, they engage in a process of cognitive evaluation called appraisal. Primary appraising is
an evaluation of whether or not the situation is relevant to and congruent with one's goals.
Lazarus (1991b, 1999) proposed that individuals will not experience psychological stress in a
situation that is not perceived as goal relevant. Goal-relevant situations are then subj ect to
secondary appraising, which involves an evaluation of one' s ability to deal with the stressor, as
well as expectations about the situation's resolution (Lazarus, 1991b, 1999).
The strategies an individual actually employs to deal with a stressful situation are referred
to as coping (Lazarus & Folkman, 1984). Lazarus and Folkman identified two functional
categories of coping behaviors. Problem-focused coping refers to cognitive and behavioral
efforts to alter the person-environment relationship that is the source of stress (Lazarus &
Folkman). In contrast, emotion-focused coping concerns cognitive and behavioral efforts to
regulate stressful emotions without attempting to alter the source of stress (Lazarus & Folkman).
According to C-M-R theory (Lazarus, 1991b, 1999), appraisal and coping are predicted to
mediate the relationship between the stressful situation and the emotions felt in response to that
situation. Furthermore, Lazarus and Folkman emphasized the dynamic, process-oriented nature
of relationships between stress, appraisal, coping, and emotion in that they are constantly
changing as the situation unfolds.
Research on coping in the general psychology literature has yielded several consistent
findings that support the predictions of C-M-R theory (Lazarus, 1991b, 1999). First, individuals
employ a number of different coping strategies in combination when dealing with a stressful
situation (Folkman & Lazarus, 1980, 1985, 1988). Second, coping is more variable than
consistent (Folkman & Lazarus, 1980, 1985). Finally, research on coping in the general
psychology literature has revealed that individual difference variables (e.g., neuroticism,
extroversion, optimism, type-A personality) are influential in the relationships between appraisal,
coping and emotion (Carver, Scheier & Weintraub, 1989; David & Suls, 1999; Endler, Kantor &
Parker, 1994). Coping research in sport has mirrored the general psychology literature in
supporting C-M-R theory (Lazarus, 1991b, 1999). Therefore, based on its wide acceptance, its
relevance to the constructs to be examined, and its considerable empirical support in both general
and sport psychology, I adopted the C-M-R theory of emotion (Lazarus, 1991b, 1999) as the
theoretical framework for my study.
It is important to note that Lazarus' (1991b, 1999) theoretical predictions have evolved
over time. For instance, Lazarus (1999) recently identified important antecedents of appraisal
that operate as moderating variables in the stress, emotion, and coping process. These variables,
which Lazarus (1991b, 1999) predicts will interact to influence appraisal, are categorized as
environmental variables and person variables. Environmental variables include demands,
constraints, opportunities, and culture. Person variables include beliefs about one's self and the
world, personal resources, and goals and goal hierarchies.
One type of goal hierarchy identified by researchers in sport psychology is achievement
goal orientation, the extent to which an athlete's achievement goals are self- or norm-referenced
(Duda, 1989). Specifically, Duda distinguished between task and ego orientations in sport.
Athletes adopting a task orientation perceive competence based on self-referenced standards
(e.g., task mastery, effort investment, skill development, etc.), and view effort as undifferentiated
from ability (i.e., ability results from increased effort rather than innate talent; Duda). In contrast,
the competence beliefs of athletes adopting an ego orientation are founded on norm-referenced
standards (e.g., defeating opponents, displaying one's superior ability to others; Duda).
Furthermore, ego oriented individuals perceive effort as differentiated from ability (i.e., ability
results from innate talent rather than increased effort; Duda).
Achievement goal theory in sport (Duda, 1989) is theoretically compatible with the C-M-R
theory of emotion (Lazarus, 1991, 1999). Specifically, both theories offer predictions with
respect to the role of motivation in appraisal and emotion. However, little, if any, research has
examined the relationships between achievement goals and the maj or constructs of C-M-R
theory. Therefore, I attempted to address this limitation of the extant literature by examining
achievement goals as antecedents of appraisal that moderate the relationships between appraisal,
coping, and emotion.
Recently, researchers in sport psychology have begun to examine the concept of coping
effectiveness (Campen & Roberts, 2001; Haney & Long, 1995; Kim & Duda, 2003; Ntoumanis
& Biddle, 1998; Pensgaard & Duda, 2003). In an extension of Folkman (1992), Gould (1996)
identified two approaches to studying coping effectiveness in sport, and recommended that
researchers utilize both simultaneously. First, Folkman (1992) advocated the goodness-of-fit
approach, and proposed that coping effectiveness is determined by the degree of fit between
situational appraisals and situational coping efforts. Specifically, Folkman predicted that in
situations appraised as controllable, problem-focused strategies are most effective. In contrast,
emotion-focused strategies were predicted most effective in situations appraised as
uncontrollable. For example, a student who experiences academic stress resulting from the
completion of his or her thesis would be best served by using problem-focused coping strategies
like increased effort, seeking advice, or other efforts directly linked to the work required. In
contrast, after submitting one's thesis, emotion-focused coping strategies would be most
effective because the situation is outside of the student' s control.
The second approach to assessing the effectiveness of a situational coping strategy has
been to determine whether it is associated with improvements in important outcome variables
(e.g. athletic performance, performance satisfaction, affect). To date, only two studies in the
sport psychology literature (Haney & Long, 1995; Kim & Duda, 2003) have applied both of the
above approaches simultaneously to the study of coping effectiveness. My study continued this
line of research by assessing perceived control, coping, and affective outcomes in sport.
Because it is difficult to infer causality when relating coping to important outcomes, some
researchers have recently focused their attention on a more subj ective measure of coping
effectiveness, perceived coping effectiveness, which is simply an individual's cognitive
evaluation of how effective his or her coping behaviors were in dealing with a particular stressful
situation (Campen & Roberts, 2001; Kim & Duda, 2003; Ntoumanis & Biddle, 1998; Pensgaard
& Duda, 2003). Preliminary evidence has indicated that an athlete's perceived effectiveness of
his or her coping efforts may be a better predictor of sport-related outcomes than the coping
efforts themselves (Ntoumanis & Biddle; Pensgaard & Duda). In my study, I further examined
the relationships between perceived coping effectiveness and important outcomes in sport.
To summarize, studies in both the general and sport psychology literature that have
adopted the C-M-R theory of emotion (Lazarus, 1991b, 1999) as a research framework have
found that (a) coping is complex in that individuals use numerous strategies, often in
combination when dealing with stress, (b) coping is characterized by both between- and within-
situation variability, and (c) coping is related to important individual difference variables like
personality. Furthermore, Lazarus's (1991b, 1999) theoretical predictions concerning the
relationships between appraisal, coping, and emotion have been demonstrated empirically by
researchers in both general (Carver et al., 1989; Folkman & Lazarus, 1980, 1985, 1988; Iwasaki,
2003; Park, Folkman, & Bostrom, 2001; Zeidner, 1995) and sport psychology (Crocker &
Graham, 1995; Gaudreau et al., 2002; Gaudreau et al., 2001; Giacobbi et al., 2004; Gould et al.,
1993; Kim & Duda, 2003; Pensgaard & Duda, 2003).
Problem Statement, Study Purposes, and Hypotheses
The rationale for my study was founded on both theory and application. To begin, because
many college students experience psychological distress, and because many of these same
individuals participate in club sports, the study of how college students cope with chronic
sources of stress related to academic and sport challenges is important in order to gain a more
complete understanding of adaptational processes. In addition, because not all individuals
succumb to the deleterious effects of stress, the study of coping also has applied implications
(Folkman, 1992). Thus, from an applied perspective, there is a need to study the coping reactions
of club sport athletes who are dealing with the stress of managing athletic and academic time
demands. As Hill (2001) pointed out, "theory defines the underlying causes of the athlete's
problems, guides the process for solving these problems, and suggests which types of
intervention techniques should be used" (p. xv). Therefore, in order to guide and improve
interventions, the theory that underlies these interventions must be validated through applied
The C-M-R theory of emotion (Lazarus, 1991b, 1999), which informs several cognitive-
behavioral interventions (e.g., The Coping with Stress model; Zeitlin, Williamson, & Rosenblatt,
1987), has received considerable support in the extant sport psychology literature. Nevertheless,
there are several relatively unexplored aspects of the theory that, if confirmed, might augment
existing coping interventions. Specifically, few studies have examined the ways that club sport
athletes cope with chronic, non-competitive stress. In addition, while personality traits such as
neuroticism, extraversion, and trait anxiety have been shown to relate to coping, appraisal, and
emotion, other person variables (e.g., achievement goal orientations) have received little
attention in the extant literature. Furthermore, while recent research has found that perceived
coping effectiveness is related to affect and athletic performance, this area of research is
relatively new and requires further exploration. Finally, few, if any, studies have simultaneously
examined all four maj or constructs of C-M-R theory (i.e., antecedents of appraisal, appraisal,
coping, and emotion).
Based on these limitations of the extant literature, my study had two related objectives.
First, the viability of C-M-R theory (Lazarus, 1991b, 1999) was considered by testing a
theoretically- and empirically-based model of the relationships between goal orientations,
primary appraisals, secondary appraisal, coping, and affect. Specifically, the relationships
depicted in this proposed model are based on the theoretical predictions of C-M-R theory
(Lazarus, 1991b, 1999) and achievement goal theory in sport (Duda, 1989; Roberts 1992), along
with the findings of research showing empirical links between these two theories (Ntoumanis,
Biddle, & Haddock, 1999). The second, related purpose of my study was to examine the ways
that club sport athletes appraise, cope with, and emotionally respond to chronic, non-competitive
stress (i.e., managing athletic and academic time demands).
In the proposed model, I hypothesized that athletes who score highly on a measure of task
orientation appraise the chronic stress of managing athletics and academic time demands as
challenging and within their control, and therefore employ problem-focused coping efforts. In
contrast, ego-oriented athletes appraise the stressor as threatening and outside their control, and
therefore employ emotion-focused coping to deal with the threat. Finally, both task- and ego-
oriented athletes experience increased positive affect if they perceive their coping efforts to be
effective, and experience increased negative affect if they perceive their coping efforts to be
Levels of Analysis
Stress is a seemingly ubiquitous feature of the human experience and, therefore, an
understanding of the causes, mechanisms, and effects of stress has implications for nearly all
individuals. However, the study of human stress must begin with a concrete definition of the
stress construct. Unfortunately, because it is studied by researchers in scientific disciplines as
varied as medicine, psychology, sociology, and sport sciences, a consensus operational definition
of stress has proven elusive. Lazarus (1999) offered that the major reason for ambiguity
regarding stress definitions is that the different scientific disciplines mentioned above have
studied stress at three different levels of scientific analysis: (a) sociocultural; (b) physiological;
and (c) psychological.
First, sociologists and cultural anthropologists have studied stress at its broadest level and
have been concerned with the ways that the structure of societies and cultures influence the stress
processes of the individuals within them (Lazarus, 1999). In sociology, scholars have ascribed
the term social strain to the social conditions (e.g., war, racism) and crises (e.g., economic
depression, social anarchy) that trigger stress reactions in individuals (Lazarus, 1999). For
example, Hollingshead and Redlich (1958) observed that social deviance and mental illness
increased under conditions of social strain. Cultural anthropologists also have been interested in
the social aspects of stress, and have studied how particular values of a culture define what is
considered stressful by its members, and how these values differ between cultures (Lazarus,
At the opposite extreme, physiologists have studied stress at its most basic level and have
been concerned with ways that physical stimuli initiate biological, neural, and hormonal
adaptations in the human body (Lazarus, 1999). Selye's (1956/1976) general adaptation
syndrome (GAS) is a prototypical example of stress studied at the physiological level of analysis.
Selye proposed that when confronted with a harmful stimulus, the human body proceeds through
a series of general adaptational stages, utilizing a number of pathways in the nervous system
designed to maintain internal homeostasis.
Finally, psychologists have studied stress between the broadest (i.e., sociocultural) and
most basic (i.e., physiological) levels and have been concerned with the human mind and
behavior (Lazarus, 1999). Because human cognition and behavior influence society, as well as
bodily function, the psychological level of analysis complements, rather than diverges from, the
sociocultural and physiological levels of analysis. Examples of this complementary relationship
are (a) the study by Hollingshead and Redlich (1958) discussed above which dealt with the
effects of social strain on mental illness, and (b) Selye' s (1956/1976) proposal that the GAS,
though physiological in nature, can be initiated in response to psychological stress.
To summarize then, part of the confusion regarding the definition of stress is due to the
different levels of analysis, ranging from basic (physiological) to general socioculturall), through
which researchers have sought to examine this construct. Because my study was concerned with
psychological stress, what follows is a conceptual review of how psychologists have defined the
One approach to defining psychological stress has been the stimulus-response approach.
Traditionally, stimulus and response definitions of stress have been considered separately. The
stimulus-based approach has defined psychological stress as an intrinsic property of particular
situations. For example, Stone and Neale (1984) adopted a stimulus-based definition of stress in
their development of the Daily Life Events Checklist (DLE), which requires individuals to
indicate the desirability or undesirability of situations (e.g., death of a relative, change in job
status) they encountered on a given day. Underlying this measurement approach is the
presumption that daily events rated as undesirable are stressful, while events rated as desirable
are not stressful. For example, if an individual indicates that receiving a job promotion is
desirable, it is presumed that he or she did not experience stress. However, an obvious limitation
of this approach is that many desirable situations are nevertheless stressful. For example,
receiving a job promotion may be desirable, but it may also be stressful in that the individual
might realize that he or she is now personally accountable for the failures of his or her
subordinates. Therefore, as Lazarus (1999) argued, certain situations are indeed stressful, but it is
the person' s perception of the situation, not the situation itself, that generates stress.
In contrast to the stimulus-based approach is the response-based approach, which has
defined stress as the emotional and physiological reaction to stressful stimuli (Lazarus, 1999).
Researchers adapting Selye's GAS (1956/1976) to examine the body's physiological response to
psychological stressors (e.g., ego threat) have adopted this response-based definition of
psychological stress (Mason et al., 1976; Symington; Currie, Curran, & Davidson, 1955).
Specifically, these researchers have viewed pituitary release of adrenocorticotrophic hormone
(ACTH) as the defining feature of psychological stress. For example, Mason and colleagues
(1976) found that individuals in an ego-threatening (i.e., stressful) situation exhibited elevated
levels of ACTH.
Recently, Lazarus (1999) argued that stimulus-based and response-based definitions of
stress are two sides of the same coin, and therefore, should be discussed in terms of a combined
stimulus-response approach. Specifically, he asserted that a stimulus cannot be defined as
stressful without the observation of a stress response. For example, a situation listed on the DLE
(Stone & Neale, 1984) cannot be considered stressful until an individual responds that the
particular event was undesirable. Similarly, Lazarus (1999) argued the stress response cannot be
defined without reference to the stimulus that elicited it. For instance, in the study by Mason et
al. (1976) cited above, these researchers demonstrated that pituitary secretion of ACTH (i.e., the
physiological response) only emerged in response to particular situations (i.e., stimuli) that were
perceived as threatening to one's personal well-being. Therefore, to define stress based on a
stimulus or a response alone ignores the necessary relationship between the two and limits the
predictive utility of each (Lazarus, 1999).
Even when combining the approaches, stimulus-response definitions of stress remain
limited in that they ignore the individual difference variables that influence both the stimulus
(i.e., what situations an individual considers stressful) and the response (i.e., how an individual is
going to respond to a stressful situation; Lazarus, 1999). As Lazarus (1999) stated, "In effect, it
takes both the stressful stimulus condition and a vulnerable person to generate a stress reaction.
Putting the person into the equation is the only way to solve the dilemma" (p. 53). Stated
differently, Lazarus (1999) argued that stress is defined, not by a stimulus and a response, but by
an individual's cognitive processes, which intervene between the stressor and its related
In order to account for these intervening processes, Lazarus (1991b, 1999) developed the
C-M-R theory of emotion, which has emerged as the most widely accepted conceptual
framework for examining psychological stress and coping (David & Suls, 1999; Gould, 1996).
The basic tenet of C-M-R theory is that the relational meaning generated by a person's
constantly changing, mutually reciprocal transaction with the environment is the primary
predictor of stress, coping, and emotion in such an encounter. For instance, the way a player
reacts to a poor call by an official depends on the meaning of the situation to the player rather
than the simple interaction of variables related to the official's competence (i.e., the
environment) and the player' s stress reactivity (i.e., the person). Therefore, Lazarus and Folkman
(1984) defined psychological stress as, "a particular relationship between the person and the
environment that is appraised by the person as taxing or exceeding his or her resources and
endangering his or her well-being" (p. 19).
Since its introduction over 20 years ago, the C-M-R theory of emotion (Lazarus, 1991b,
1999) has stimulated a wealth of research in both the general and sport psychological domains
(Bouffard & Crocker, 1992; Crocker & Graham, 1995; David & Suls, 1999; Dugdale et al.,
2002; Folkman & Lazarus, 1985, 1988; Gaudreau et al., 2002; Gaudreau et al., 2001; Giacobbi &
Weinberg, 2000; Gould et al., 1993; Hammermeister & Burton, 2001; Haney & Long, 1995;
Ntoumanis & Biddle, 1998; Pensgaard & Duda, 2003). In general, this research supported the
basic constructs of C-M-R theory (Lazarus, 1991b, 1999) and confirmed that a process-oriented,
cognitive-mediational approach is the most viable alternative for research on psychological
stress, coping, and emotion. Given this empirical support and its relevance to the constructs to be
examined, I adopted the definitional approach advocated by the C-M-R theory of emotion
(Lazarus, 1991b, 1999).
Acute vs. Chronic Stress
Lazarus and Folkman (1984) distinguished between acute and chronic stress in their
preliminary work on C-M-R theory. Specifically, they viewed acute stress to result from the
interpretation of a particular event as stressful. In contrast, chronic stress results from the
continued interpretation of the same event as stressful over a prolonged period of time. Recently,
Lazarus (1999) further argued that chronic stress arises from "harmful or threatening, but stable,
conditions of life, and from the stressful roles people continually fulfill at work and in the
family" (p. 144). For example, the threat of being unable to successfully manage athletic and
academic time demands would be considered a chronic stressor by an individual continually
fulfilling the role of student-athlete. Indeed, my study focused on this particular source of
The distinction between acute and chronic stress has important implications for coping
research and intervention. First, Gottlieb (1997) argued that different strategies are required to
effectively cope with chronic stress and acute stress. Therefore, in order to get a complete picture
of how individuals cope with stress, researchers must examine chronic stressors (e.g., daily
hassles) in addition to acute stressors (e.g., major life events, Lazarus, 1999). Second, as Aldwin
and Brustrom (1997) pointed out, chronic stressors are often dealt with and managed rather than
fully resolved (e.g., terminal illness). Therefore, researchers must measure the effectiveness of
coping efforts employed in response to a chronic stressor differently than for efforts employed in
response to an acute stressor. Despite these implications, the study of acute stress has received
the vast maj ority of attention in extant research on stress, coping, and emotion in both general
and sport psychology (Anshel & Anderson, 2002; Bouffard & Crocker, 1992; Carver et al., 1989;
Crocker & Graham, 1995; Crocker & Isaak, 1997; David & Suls, 1999; Dugdale et al., 2002;
Eubank & Collins, 2000; Folkman & Lazarus, 1985; Endler et al., 1994; Gaudreau et al., 2002;
Gaudreau et al., 2001; Gould et al., 1993; Hammermeister & Burton, 2001; Haney & Long,
1995; Krohne & Hindel, 1988). Therefore, I attempted to fill this gap in the literature by
examining how individuals who occupy both athletic and academic roles (i.e., university club
sport athletes) appraise, cope with, and experience affective responses to the chronic stress of
managing athletic and academic time demands.
Regardless of whether a stressor is acute or chronic, two related processes are contained
within Lazarus and Folkman's (1984) definition of stress: (a) appraisal; and (b) coping.
Therefore, what follows in the next two sections is a discussion of these two constructs. First, the
antecedents and types of appraisals proposed by Lazarus (1999) in the C-M-R theory of emotion
will be described in detail. Subsequently, the major theoretical approaches to the study of coping
will be outlined, with emphasis on the process approach advocated by Lazarus and Folkman
(1984). Finally, the implications of adopting this process approach for my study will be
identified and discussed.
According to Lazarus (1991Ib), appraisal is defined as "an evaluation of the personal
significance of what is happening in an encounter with the environment" (p. 820). With respect
to the causal role of cognitive appraisal in emotion, Lazarus (1991a) adopted the cognitive-
mediational view that appraisal is both necessary and sufficient for an emotion to result. Stated
differently, appraisals are not simply capable (sufficient) of producing emotions. Rather, without
appraisals emotions would not occur (Lazarus, 1991a). In addition, Lazarus and Smith (1988)
emphasized the conceptual distinction between appraisal and knowledge. In contrast to
knowledge, which is a cold cognition about the facts of a situation (i.e., who, what, where, when,
why, and how), appraisals are warm or hot cognitions regarding the personal meaning of a
situation in terms of an individuals goals, beliefs, attitudes, and well-being (Lazarus & Smith;
Lazarus, 1999). In the C-M-R theory of emotion, Lazarus (1991b, 1999) specified both the
specific types of appraisals and their antecedent conditions, each of which will be discussed
Types of Appraisal
Lazarus (1991b, 1999) proposed that individuals engage in two transactional processes
when in a potentially stressful encounter: primary and secondary appraising. The products of
primary and secondary appraising represent the two types of appraisals identified by Lazarus
(1991b, 1999) in the C-M-R theory of emotion: primary and secondary appraisal. Primary
appraisal concerns "whether and how the encounter is relevant to the person's well being"
(Lazarus & Smith, 1988, p.284). Secondary appraisal, on the other hand, is a judgment of the
resources available to deal with a stressful encounter and its potential outcomes (Lazarus, 1991b,
1999). In addition, Lazarus (1991b, 1999) identified subcomponents of both primary and
secondary appraisal. The next two sections of this paper will offer a theoretical discussion of
each appraisal subcomponent.
Lazarus (1991Ib, 1999) distinguished between three subcomponents of primary appraisal:
(a) goal relevance, (b) goal congruence, and (c) type of ego involvement. Goal relevance refers
to a determination of what, if anything, is at stake in an encounter (Lazarus, 1991b). For
example, in preparation for a midterm exam, a college student may appraise his or her
performance on the exam as being relevant for future goals respecting graduation, and thus,
earnings potential. Goal congruence concerns "the extent to which a transaction is consistent or
inconsistent with what the person wants" (e.g. graduation, earnings; Lazarus & Smith, 1988, p.
289). Finally, type of ego involvement involves the kind of goal at stake (Lazarus, 1991b, 1999).
Lazarus (1999) identified six types of ego involvement: (a) social and self-esteem, (b) moral
values, (c) ego ideals, (d) meanings and ideas, (e) other persons and their well-being, and (f) life
goals. With respect to the role of each type of primary appraisal in emotion, Lazarus (1991b)
proposed that goal relevance determines whether or not an emotion will occur, goal congruence
establishes the valence of the experienced emotion (i.e., whether it is positive or negative), and
type of ego involvement distinguishes between the specific harm, threat, and challenge emotions
Most relevant to the purposes of my study are goal congruence appraisals because, as
Lazarus (1991b) argued, they determine the relational meaning of a stressful encounter. Lazarus
and colleagues (Lazarus, 1991b, 1999; Folkman & Lazarus, 1980) proposed three distinct
relational meanings indicative of psychological stress: (a) harm, a loss that has already occurred;
(b) threat, an anticipated loss; and (c) challenge, a difficult-to-attain, yet anticipated gain.
Furthermore, Lazarus (1991b, 1999) proposed that these three relational meanings result in three
maj or types of stress emotions: harm emotions, threat emotions, and challenge emotions. Harm
emotions are outcome-based and include anger, guilt, shame, and sadness (Lazarus, 1991b,
1999). Threat emotions are anticipatory and include anxiety and fright (Lazarus, 1991b, 1999).
Finally, challenge emotions are anticipatory, reflective of movement toward goal attainment, and
include happiness, pride, gratitude, and love (Lazarus, 1991b, 1999).
According to the C-M-R theory of emotion (Lazarus, 1991b, 1999), the secondary
appraisal process comprises three judgments: (a) future expectations, (b) blame or credit, and (c)
coping potential. Future expectations reflect an individual's judgment as to whether or not the
stressful encounter will be resolved favorably or unfavorably for his or her well-being (Lazarus,
1991Ib, 1999). Blame or credit is a determination of whom or what is responsible for the
particular harm, threat, or challenge appraised in a stressful situation (Lazarus 1991b, 1999).
With regard to emotion, Lazarus (1991b, 1999) argued that appraising blame or credit to oneself
results in different emotions than directing blame towards others (e.g., anger vs. guilt, shame,
Finally, coping potential concerns whether, and in what ways, an individual can influence
the person-environment relationship (Lazarus, 1991b, 1999). Specifically, in a given stressful
encounter, one must judge whether there is any potential to reduce or eliminate a harm or threat,
or attain a challenging goal. According to Lazarus (1991b, 1999), secondary appraisal of coping
potential has important implications for the ways that individuals cope with stress. Therefore, the
maj or predictions of C-M-R theory (Lazarus, 1991Ib, 1999) regarding relationships between
coping potential, actual coping efforts, and emotion will be discussed in greater detail in the
coping section of this paper. Attention will now turn toward the maj or cognitive-motivational-
relational constructs that influence how an individual appraises a stressful encounter.
Antecedents of Appraisal
Lazarus (1999) identified two broad categories of appraisal antecedents: (a) environmental
variables, and (b) person variables. Lazarus viewed these aspects of the person and the
environment as primary moderators, through the appraisal process, of the relationships between
stress, coping, and emotion. Stated differently, Lazarus argued that an individual's appraisal of
an event exerts a direct influence on stress, coping, and emotion, while the influence of the
person or environment is indirect. The maj or types of environmental and person variables will be
Lazarus (1999) identified the following four categories of environmental variables that
interact with person variables as antecedents of appraisal: (a) demands, (b) constraints, (c)
opportunities, and (d) culture. Opportunities and culture will not be discussed in detail here for
two reasons. First, they were not fully detailed by Lazarus (1999) in his presentation of the C-M-
R theory of emotion. Second, they are not germane to the purposes of my study. Therefore, this
section will focus on demands and constraints as antecedents of appraisal (See Lazarus, 1999, for
a discussion of opportunities and culture as antecedents of appraisal).
Demands are defined as "the implicit or explicit pressures from the social environment to
act in certain ways and manifest socially correct attitudes" (Lazarus, 1999, p. 61). Stated
differently, demands influence the appraisal process through the pressure to conform to social
conventions. For example, students who participate in organized university athletics (e.g., club
and intercollegiate sports) are pressured by their professors and academic advisors to behave in
ways that will help them excel in academics (e.g., studying for exams, going to class, keeping up
with required readings), while at the same time being pressured by their coaches and teammates
to behave in ways that will develop excellence in athletics (e.g., attending all practices and
meetings, participating in team workouts). When faced with a stressful situation in which he or
she must choose between an academic or athletic activity (e.g., having to postpone studying for
an exam in order to attend a mandatory team workout), the ways that a student-athlete appraises
and copes with these competing environmental demands is influenced by his or her perception of
which is more salient. Indeed, in my study, I examined the ways that student-athletes appraise
and cope with the competing demands presented in this example.
In contrast to demands, constraints are socially unacceptable ways of behaving that affect
appraisal by limiting one's coping potential and affecting one's social and self-esteem (Lazarus,
1999). For example, in the stressful situation described above, a student-athlete might decide that
his or her academic demands are more salient, and therefore cope by focusing less on his or her
athletic responsibilities. However, such a coping strategy would be constrained by the practice
and workout attendance rules imposed by his or her coach, thereby limiting his or her coping
Person variables include (a) beliefs about self and world, (b) personal resources, and (c)
goals and goal hierarchies (Lazarus, 1999). First, one's beliefs about self and the world influence
appraisals through their effect on situational expectations. For example, an athlete playing for a
vociferous (vs. calm) coach might be more likely to appraise poor individual performance as
threatening because he or she expects to be berated by the coach.
According to Lazarus (1999), personal resources (e.g., intelligence, money, attractiveness,
social skills, social support network) are aspects of the individual that influence appraisal
through their effects on one's chances for adaptational success. For example, at the beginning of
his or her first year, a student-athlete might appraise the intense intercollegiate workout regimen
as highly stressful. However, at the end of the year, he or she might appraise this same regimen
as less stressful because of the strong social support network (i.e., personal resource) he or she
cultivated throughout the course of the year. Giacobbi and colleagues (2004), in their
longitudinal study of freshmen female athletes, found support for Lazarus predictions' regarding
Finally, Lazarus (1999) recognized that goals and goal hierarchies are antecedents of
appraisal because individuals are more likely to appraise a situation as meaningful if it involves
more important, as opposed to less important, goals. In sport, researchers have identified several
individual difference (person) variables reflecting goals and goal hierarchies, including the extent
to which an individual priorities athletic goals (athletic identity; Brewer, Van Raalte, & Linder,
1993) and the extent to which an athlete' s achievement goals are self- or norm-referenced (task
and ego orientations; Duda, 1989). I examined the role of this latter person variable, achievement
goal orientation, as an antecedent of appraisal, coping, and emotion. Therefore, a brief theoretical
discussion of achievement goal orientations is appropriate.
Achievement goal orientations
Nicholls (1984, 1989) in the educational psychology literature and Duda (1989) in the
sport psychology literature proposed that an individual's dispositional motivation in an
achievement context (e.g., school, athletics) depends on his or her criterion for competence-
related beliefs. Furthermore, Duda and Nicholls (1992) argued that these beliefs "cut
across...domains," and therefore are "theories in the larger sense of world views" (p. 297).
Specifically, Nicholls (1984, 1989) and Duda (1989) distinguished between two
achievement motives: task orientation and ego orientation. Individuals adopting a task
orientation perceive personal competence based on self-referenced standards (e.g., task mastery,
effort investment, skill development, etc.), and view efforts as undifferentiated from ability (i.e.,
ability results from increased effort rather than innate talent; Duda, 1989; Nicholls, 1984, 1989).
In contrast, the competence beliefs of individuals adopting an ego orientation are founded on
norm-referenced standards (e.g., defeating opponents, displaying one's superior ability to others;
Duda, 1989; Nicholls, 1984, 1989). Furthermore, individuals with a strong ego orientation
perceive effort as differentiated from ability (i.e., ability results from innate talent rather than
increased effort; Duda, 1989; Nicholls, 1984, 1989). Achievement goal theory has important
implications for the stress process. However, little, if any, research has examined the impact of
achievement goals within the C-M-R theory of emotion (Lazarus, 1991b, 1999). Therefore, I
attempted to address this limitation of the extant literature by examining achievement goals as
antecedents of appraisal that moderate the relationships between appraisal, coping, and emotion.
For the purposes of my study, achievement goal theory (Duda, 1989; Nicholls, 1984, 1989)
and the C-M-R theory of emotion (Lazarus, 1991b, 1999) are linked theoretically in at least three
ways. First, achievement goal orientations can be considered person variables in the appraisal
process. As stated previously, Lazarus (1999) emphasized the motivational aspects of stress by
(a) identifying goal hierarchies as person variables that antecede appraisal, and (b) proposing that
the perception of psychological stress in an encounter is dependent on a primary appraisal of
goal relevance. Indeed, a central tenet of C-M-R theory is that an individual's motives in an
encounter are influential in determining how he or she will appraise the encounter (Lazarus,
1999). Compatible with this view is that of Duda and Hall (2001), who posited that
"achievement goals are presumed to be the organizing principle influencing how we interpret,
feel about, and react to our achievement-related endeavors" (p. 417).
Second, both achievement goal theory (Duda, 1989; Nicholls, 1984, 1989) and the C-M-R
theory of emotion (Lazarus, 1991b, 1999) propose that an individual's perception of control
influence his or her cognitive and behavioral strategies in an adaptational encounter. Specifically,
achievement goal theory posits that, in the face of adversity, individuals adopting a task
orientation perceive successful (or unsuccessful) outcomes to be within their control (i.e., due to
effort), and are therefore more likely to be challenged and employ adaptive strategies (e.g.,
problem-solving, planning, and increased effort; Roberts, 1992). In contrast, individuals with a
strong ego orientation perceive success (or failure) as uncontrollable (i.e., due to innate talent),
and are therefore more likely to be threatened by adversity and employ maladaptive strategies
(e.g., task-irrelevant thoughts; Roberts, 1992). In general, research in both general and sport
psychology has supported the prediction that perceptions of control are related to adaptational
thoughts and behaviors (Folkman & Lazarus, 1980; Folkman & Lazarus, 1985; Hammermeister
& Burton, 2001; Kim & Duda, 2003; Peacock & Wong, 1996; Reese, Kliewer, & Suarez, 1997).
Specific studies examining the relationships between achievement goal orientations, appraisal,
coping, and perceptions of control will be presented in subsequent sections of this paper.
Finally, both achievement goal orientations (Duda, 1989; Nicholls, 1984, 1989) and
cognitive-motivational -relational constructs (Lazarus, 1991b, 1999) are proposed to influence
emotion. For example, Roberts (1986) argued that individuals scoring highly on measures of task
orientation would be less likely than their ego-oriented counterparts to experience competitive
anxiety because they are more likely to appraise a competition as challenging (as opposed to
threatening). Such an argument is consistent with the cognitive-motivational -rel ati onal
categorization of anxiety as a threat emotion. Unfortunately, researchers only recently have
begun to examine the relationships between goal orientations, relational meanings (i.e., harm,
threat, and challenge), and affective experiences. Emerging evidence supporting these
relationships will be presented in the subsequent sections of this paper.
To summarize, achievement goal theory (Duda, 1989; Nicholls, 1984, 1989) and the C-M-
R theory of emotion (Lazarus, 1991b, 1999) share in common a theoretical framework that
attempts to explain an individual's thoughts, feelings, and behaviors in personally meaningful
situations. Specifically, it appears theoretically justified to classifying achievement goal
orientations as person variables that antecede appraisal and thereby influence the relationships
between appraisal, coping, and emotion. However, in the extant sport psychology literature,
researchers have shown little interest in examining achievement goal orientations within a
cognitive-motivational -rel ati onal approach to stress. My study specific cally addressed thi s
Summary of Appraisal
According to the C-M-R theory of emotion (Lazarus, 1991b, 1999), stress emotions result
from an individual's cognitive appraisal that a stressful situation is meaningful for his or her
well-being. Lazarus (1991b, 1999) identified two types of cognitive appraisal: primary appraisals
and secondary appraisals. Primary appraisals concern whether an encounter has implications for
one's well-being and include (a) goal relevance, (b) goal congruence, and (c) type of ego-
involvement. Secondary appraisals involve an evaluation of what can be done about the stressor
and include (a) blame or credit, (b) coping potential, and (c) future expectations. Both the
primary and secondary appraisal processes are influenced by person and environmental
variables, which serve as antecedents of appraisal. In my study, person variables were
represented by achievement goal orientations (Duda, 1989; Nicholls, 1984, 1989), which are
analogous to the concept of goals and goal hierarchies in the C-M-R theory of emotion (Lazarus,
Lazarus and Folkman (1984; Folkman & Lazarus, 1988) asserted that coping and appraisal
are joined as mediators of stress emotions. From the above discussion of appraisal, one can glean
that the coping process greatly influences, and is influenced by, the way an individual appraises
the adaptational significance of an event. However, prior to the 1970s, the term coping was used
infrequently to describe the ways that people manage their stressful life conditions (Lazarus,
1999). Instead, theorists and researchers viewed stress management efforts as reflective of one's
personality, and therefore referred to these efforts in terms consonant with specific personality
theories. Based on this history, I will now endeavor to briefly outline the traditional approaches
that researchers have adopted to study coping. First, coping as an ego defense mechanism within
psychoanalytic theory will be discussed. Next, the trait approach, which focuses on the structural
aspects of coping, will be presented. Finally, because I adopted the C-M-R theory of emotion
(Lazarus, 1991Ib, 1999) as a guiding framework for my study, the central tenets of the process
approach advocated by these authors will be described in detail.
In its earliest conceptualizations, coping was studied by researchers from an ego-based
perspective that emerged from the psychoanalytic concept of ego defense (Lazarus, 1999).
Specifically, these researchers defined coping as the employment of ego defense mechanisms to
reduce the tension resulting from a stressful encounter (Folkman & Lazarus, 1980). Typically,
ego defenses were stratified in terms of their adaptational utility (Folkman & Lazarus, 1980). For
example, some ego psychologists (Haan, 1969; Menninger, 1954) argued that defense
mechanisms are not based in reality, and therefore represent the most maladaptive form of
coping (Lazarus, 1999). In contrast, Vaillant (1977) considered certain ego defenses (e.g.,
altruism) to be more mature than others (e.g., reaction formation).
The ego approach is limited in at least two ways. First, its emphasis on tension reduction
underestimates other possible functions of coping (e.g., problem-solving; Folkman & Lazarus,
1980). Second, and most importantly, by focusing solely on the defensive stress response, the
ego approach fails to consider what it is about the person or the environment that influences such
a response (Folkman & Lazarus, 1980). For instance, from an ego perspective, reaction
formation is defined as "behavior in a fashion diametrically opposed to an unaccepted instinctual
impulse" (Vaillant, 1977, p. 385). Therefore, if an athlete restrains from retaliating toward a
verbally abusive coach, he or she is described as having employed reaction formation as an ego
defense against the stress of the abusive encounter. However, this analysis is inadequate because
it offers little information about the athlete' s appraisals of the encounter (e.g., interpreting the
coach's comments as a threat or challenge), which, as already discussed, are determined by the
combined influence of person and environmental variables. Furthermore, the athlete' s actual
motives for restraining (e.g., remaining on the team roster), and his or her emotions (e.g., anger,
shame, gratitude) are also ignored. Based on these limitations, researchers began to study coping
from a trait perspective, which considers the stable aspects of the individual as maj or predictors
of situational coping efforts (Folkman & Lazarus, 1980).
Traditionally, researchers have examined trait coping in three distinct ways varying in
theoretical sophistication (Lazarus, 1999). The least sophisticated approach has been to examine
coping style, which can be defined as an individual's tendency to employ a certain type or
category of coping strategy across time and situations (Lazarus, 1999). Typically, coping styles
have been expressed in terms of dichotomous dimensions derived from the ego defense approach
discussed above (e.g., approach-avoidance, repression-sensitization; Lazarus, 1999). A more
sophisticated approach has involved the identification of specific personality traits (e.g.,
neuroticism, trait anxiety) that might influence an individual's coping style (Lazarus, 1999). For
example, an individual who is high trait anxious might consistently use an avoidance coping
style (i.e., denial, fantasy) across time and situations. Finally, the most sophisticated approach
has been the conditional trait approach, which argues that personality traits influence coping
styles only in situations rendered functionally equivalent by the specific personality trait (Wright
& Mischel, 1987). From this perspective, a basketball player who is high trait anxious might
consistently employ an avoidant coping style (e.g., pass the ball to a teammate instead of
shooting late in a game) during competitions, but not during practice.
Regardless of the specific way researchers have examined trait coping, the overall trait
perspective is limited in that it reduces coping to a contrast between competing styles,
personalities, or conditions (Lazarus, 1999). Specifically, in the trait approach, if an individual is
described as exhibiting a particular coping style across situations, several aspects of the coping
process are ignored, including (a) the multitude of various coping thoughts and behaviors that the
individual actually employed both across and within situations, (b) the aspects of the situation
that influenced those coping thoughts and behaviors, and (c) the shifts in coping thoughts and
behaviors that occurred within the specific situation (Folkman & Lazarus, 1980; Lazarus, 1999).
Based on the limitations of the ego and trait approaches to coping, Lazarus and Folkman
(1984) advocated a process approach that defines coping multidimensionally as the "constantly
changing cognitive and behavioral efforts to manage specific external and/or internal demands
that are appraised as taxing or exceeding the resources of the person" (p. 141). Three important
issues underlie this process definition of coping: (a) the functions of coping, (b) the measurement
of coping effectiveness, and (c) the methods for studying coping from a process approach
(Lazarus, 1999). Each of these theoretical challenges will now be detailed along with how
Lazarus and colleagues' work is relevant to my study.
Functions of coping
Lazarus and Folkman (1984) proposed that coping is multidimensional in that an
individual's coping efforts can ameliorate stress by either (a) altering the nature of the situation
so that it is less stressful or (b) reducing emotional distress. Lazarus and Folkman termed the
cognitive and behavioral efforts aimed at transforming the nature of a stressful situation as
problem-focused coping. For example, a football wide receiver who is ashamed after dropping a
potential touchdown pass during a game might cope with the shame in a problem-focused
manner by catching practice passes from a teammate on the sideline or by talking to the
quarterback about ways to repair the timing of a particular play.
In contrast to problem-focused coping, emotion-focused coping refers to an individual's
efforts aimed at changing the relational meaning (i.e., threat, harm, and challenge) of the
stressful situation, rather than the situation itself. Emotion-focused coping efforts, although less
active than problem-focused coping, are often more cognitive in nature as an individual may
internally restructure the meaning of the event, and therefore his or her emotional reaction
(Lazarus, 1991b). Referring to the previous example, the football wide receiver might cope with
shame in an emotion-focused way by venting to a teammate, psychologically distancing himself
from blame (i.e., engaging in denial), or using self-talk to reinterpret or reframe the reasons)
why he dropped the pass.
With regard to the functions of coping, Lazarus (1999) argued that problem-focused and
emotion-focused coping strategies are neither inherently adaptive nor inherently maladaptive. In
other words, Lazarus took the position that individuals cope with events regardless of whether
their efforts result in a specified outcome. This begs the question: How then can researchers
assess the adaptational utility of problem- or emotion-focused coping in a given stressful
encounter? One option has been the outcome approach, which assesses only the important
outcomes related to a stressful encounter (e.g., sport performance, test grades). However,
Lazarus (1999) and Folkman (1992) argued against this approach because adaptational outcomes
often do not reflect the effectiveness of an individual's coping efforts. For example, some
stressors (e.g., one's partner suffers from HIV/AIDS) are such that there is little the individual
can do to improve the outcome (e.g., death of partner). Therefore, in addition to outcomes,
researchers should focus on cognitive appraisal as an important predictor of coping effectiveness
(Folkman; Lazarus, 1999). Following this recommendation, Folkman proposed an approach to
the measurement of coping effectiveness that emphasizes the quality of fit between an
individual's coping efforts and his or her situational appraisals. Specifically, in Folkman's
goodness-of-fit approach, coping effectiveness depends on two fits between the environment and
an individual's situational appraisal. First, an individual must realistically appraise the personal
significance of the stressor (Folkman). Folkman argued that over- or underestimation of the
personal significance of a stressful encounter leads to ineffective coping because efforts are less
likely to be sufficient in meeting situational demands. For example, a football player who, during
a game, blames himself for a teammate's injury (i.e., overestimates personal significance) is
unlikely to cope effectively with his guilt because the situation demands continued focus on
game play rather than rumination and negative self-talk.
Second, there must be a fit between an individual's coping efforts and his or her secondary
appraisal of control over the stressor (Folkman, 1992). As suggested previously with respect to
the functions of coping, Folkman contended that problem-focused coping strategies are most
effective during situations appraised as controllable because the situation is amenable to such
efforts. Similarly, emotion-focused coping strategies are most effective during situations
appraised as uncontrollable because the situation is not amenable to change; and therefore,
coping efforts should instead be directed towards regulating one's emotions (Folkman). For
example, according to the goodness-of-fit approach (Folkman), increasing one's focus on the
source of stress (i.e., a problem-focused coping strategy) would be more effective for coping
with poor end-of-game performance (i.e., a controllable stressor) than for dealing with a hostile
crowd (i.e., an uncontrollable stressor). In contrast, ignoring the stressor (i.e., an emotion-
focused coping strategy) would be more effective for coping with a hostile crowd (i.e., an
uncontrollable stressor), but less effective for coping with poor end-of-game performance (i.e., a
Folkman and Lazarus (1985) recommended that process-oriented studies of coping should
satisfy three methodological criteria. First, coping should be studied in the context of an ongoing
stressful situation (Folkman & Lazarus, 1985). While necessary to some extent given the
difficulty of studying cognitive processes, retrospective reports of past stressful encounters might
reflect an individual's intuitive theories and abstract knowledge about what happened, rather
than the actual situational appraisal and coping that took place (Lazarus & Smith, 1988). This
concern was supported by Smith, Leffingwell, and Ptacek (1999), who found that daily and 7-
day retrospective accounts of coping shared only 25% of common variance. Second, Folkman
and Lazarus (1985) recommended that researchers should examine how an individual actually
copes with a real stressful situation, rather than how they would hypothetically cope.
Methodologies that employ imagery or role-playing exercises are inadequate in that they limit
the relational meaning of the person-environment relationship, which is critical for appraisal and
coping processes (Lazarus & Smith). Finally, Folkman and Lazarus (1985) recommended that
researchers should collect repeated assessments of coping, thereby facilitating examination of
changes in coping in relation to the stages of the stressful encounter. Therefore, based on these
three criteria, Lazarus (2000) advocated an ipsative-normative longitudinal design whereby a
sample of individuals is studied in different situations at different times over an extended period.
This methodology simultaneously allows for analysis of both within-participant (ipsative) and
between-participant (normative) differences in coping (Lazarus, 2000).
Summary of the Cognitive-Motivational-Relational Theory of Emotion
According to the C-M-R theory of emotion (Lazarus, 1991b, 1999), four general constructs
influence the stress process: (a) antecedents of appraisal, (b) appraisal, (c) coping, and (d)
emotion. A model of the theoretical relationships between these variables is depicted in Figure 2-
1. On the far left of Figure 2-1 are person and environmental variables, which antecede appraisal
and therefore influence the mediating processes located in the middle of the model (i.e., primary
appraisal, secondary appraisal, and coping). When confronted with a potentially stressful
situation, an individual generates a primary appraisal concerning the relational meaning of the
situation (i.e., whether it is harmful, threatening, and/or challenging). Based on this relational
meaning, specific stress emotions (e.g., anger, anxiety, sadness) are experienced. To reduce this
emotional distress, the individual evaluates his or her coping potential (i.e., secondary appraisal),
and engages in efforts to either change the nature of the situation (i.e., problem-focused coping)
or directly reduce the emotional distress (i.e., emotion-focused coping). Depending on both the
outcomes of these coping efforts and the degree to which they fit the adaptational demands of the
situation, specific stress emotions will either disappear or remain (far right of Figure 2-1).
Coping Measurement and Research in General and Sport Psychology
Over the past 20 years, the C-M-R theory of emotion (Lazarus, 1991b, 1999) has gained
widespread acceptance in both general and sport psychology as a framework for examining
stress, appraisal, coping, and emotion. Therefore, with the relevant constructs and theoretical
perspectives defined and discussed, focus will now shift toward the research in general and sport
psychology specifically relevant to the predictions of C-M-R theory. First, the maj or
measurement instruments employed by coping researchers will be identified and described. Next,
empirical evidence for both the structural and process aspects of coping will be presented.
Finally, specific sport coping research related to the purposes of my study will be discussed.
Coping Measurement in General Psychology
The two instruments most commonly used by coping researchers in general psychology to
assess coping thoughts and behaviors are the Ways of Coping checklist (Folkman & Lazarus,
1980) and the COPE inventory (Carver et al., 1989). The development, content, psychometric
properties, strengths, and weaknesses of both of these instruments will be discussed in this
The Ways of Coping checklist
Folkman and Lazarus (1980) developed the Ways of Coping checklist to assess the coping
strategies that an individual might use in a specific stressful situation. The original 68 items,
derived from both theory (i.e., C-M-R theory; Lazarus, 1991b, 1999) and extant research, were
classified by 10 raters according to the two proposed functions of coping: problem-focused
coping and emotion-focused coping. For example, the item, "Made a plan of action and followed
it," was categorized as a form of problem-focused coping, while the item, "Accepted sympathy
and understanding from someone," was hypothesized to represent a form of emotion-focused
coping. The instrument instructed individuals to think of a specific stressful situation and
respond yes or no to each item, indicating whether they did or did not employ the given coping
strategy in response to that situation.
Folkman and Lazarus (1980) reported conflicting findings regarding the reliability and
validity of the original Ways of Coping checklist. First, the inter-rater reliability of item
classification (r = .91), and Cronbach's alpha for the two sub scales (problem-focused coping =
.80, emotion-focused coping = .81) were adequate. Second, principal components factor analysis
with varimax rotation supported the existence of the two coping factors. However, while the
amount of shared variance between the problem- and emotion-focused subscales was low enough
(r2 = .19) to support their relative independence, the two sub scales were nevertheless
intercorrelated (r = .44). Furthermore, while factor analysis supported the problem-focused and
emotion-focused subscales, it also determined that only 49 of the 68 items actually loaded onto
these two factors. Despite these conflicting findings, Folkman and Lazarus (1980) approached
the item revision procedure conservatively, deleting only four items and reclassifying one item as
emotion-focused rather than problem-focused.
Carver and colleagues (1989) raised three additional concerns with the Ways of Coping
checklist (Folkman & Lazarus, 1980). First, they argued that the classification of coping
strategies as either problem- or emotion-focused is an oversimplification of the diversity and
complexity of the thoughts and behaviors that individuals use to cope with stress. Indeed,
subsequent factor analyses performed on the Ways of Coping checklist have typically revealed
solutions comprised of more than two factors (Folkman & Lazarus, 1985; Folkman, Lazarus,
Dunkel-Schetter, DeLongis, and Gruen, 1986). Second, some of the items on the Ways of
Coping checklist (Folkman & Lazarus, 1980) are ambiguously worded (Carver et al.). For
example, the item, "Took a big chance or did something risky," describes an act without
indicating why the act is being done. Both the act itself and the rationale for its use have
important adaptational implications. For instance, to cope with a stressful coach-player
relationship, two teammates might engage in risk-taking behaviors for entirely different reasons.
One athlete might engage in binge drinking with teammates because he feels the need to distance
himself from the emotions of the situation, while another might engage in binge drinking with
the coach because he feels that this bonding experience will directly improve the relationship. In
addition, other items on the Ways of Coping checklist (Folkman & Lazarus, 1980) are
ambiguous in that they combine two conceptually distinct qualities (Carver et al.). For example,
in the item, "I did something which I didn't think would work, but at least I was doing
something," it is unclear whether it is more important that the individual is taking action or that
the individual believes his actions will be unsuccessful (Carver et al.).
The COPE inventory
Based on their dissatisfaction with the Ways of Coping checklist, Carver et al. (1989)
developed the COPE as a means of gaining a clearer understanding of the complex thoughts and
behaviors individuals employ to cope with stress. The COPE (Carver et al.) comprises 13
conceptually distinct four-item subscales that were informed by the C-M-R theory of emotion
(Lazarus, 1991b, 1999), Carver and Scheier' s (1985) model of behavioral self-regulation, and
previous empirical research findings. The 13 COPE (Carver et al.) subscales are as follows: (a)
active coping, i.e., "taking active steps to try to remove or circumvent the stressor or ameliorate
its effects" (p. 268); (b) planning, i.e., "thinking about how to cope with the stressor" (p. 268);
(c) suppression of competing activities, i.e., "putting other proj ects aside, trying to avoid
becoming distracted by other events, even letting other things slide, if necessary, in order to deal
with the stressor" (p. 269); (d) restraint coping, i.e., "waiting until an appropriate opportunity to
act presents itself, holding one-self back, and not acting prematurely" (p. 269); (e) seeking social
support for instrumental reasons, i.e., "seeking advice, assistance, or information" (p. 269); (f)
seeking social support for emotional reasons, i.e., "getting moral support, sympathy, or
understanding" (p. 269); (g) focusing on and venting of emotions, i.e., "the tendency to focus on
whatever distress or upset one is experiencing and to ventilate those feelings" (p. 269); (h)
behavioral disengagement, i.e., "reducing one's effort to deal with the stressor, even giving up
the attempt to attain goals with which the stressor is interfering" (p. 269); (i) mental
disengagement, i.e., "activities that serve to distract the person from thinking about the
behavioral dimension or goal with which the stressor is interfering" (p. 269); (j) positive
interpretation and growth, i.e., "coping aimed at managing distress emotions rather than at
dealing with the stressor per se" (p. 269); (k) denial, i.e. "refusal to believe that the stressor exists
or trying to act as though the stressor is not real" (p. 270); (1) acceptance, accepting the reality of
a stressful situation; and (m) turning to religion, i.e., "the tendency to turn to religion in times of
stress" (p. 270). Two additional subscales, humor and alcohol/drug use, were not in the published
version of the COPE (Carver et al.), but were developed and included later. Humor refers to
making j okes about the stressor, while alcohol/drug use represents turning to the use of alcohol
or other drugs as a way of disengaging from the stressor.
The COPE (Carver et al., 1989) can be administered in either a dispositional or situational
form. The dispositional version of the COPE (Carver et al.) instructs respondents to indicate the
extent to which they usually employ a given coping strategy when they experience stress. Items
are phrased in the present tense (e.g., "I make a plan of action"), and responses are made on a 4-
point Likert scale (1 = I usually don't do this at all; 4 = I usually do this a lot). The situational
version of the COPE (Carver et al.) requests respondents to think of their most stressful situation
over the past two months and indicate the degree to which they employed a given coping strategy
to deal with that specific situation. In contrast to the dispositional version, items are phrased in
the past tense to assess state coping in response to specific stressful encounters (e.g., "I made a
plan of action"). Using this format, respondents are instructed to answer on a 4-point Likert scale
ranging from 1 (I didn't do this at all) to 4 (I did this a lot).
Carver and colleagues (1989) assessed the factor structure of the dispositional COPE with
a sample of 978 college undergraduates. Principal components factor analysis (with oblique
rotation) yielded 12 factors with eigenvalues greater than 1.0. However, one of these factors had
no item loadings greater than .30, and was therefore discarded. The structure of the 11 remaining
factors deviated from expectations in only three ways. First, with regard to a priori assignment of
items to scales, all items loaded on their hypothesized scales except for the active coping and
planning items, which loaded onto a single factor. Similarly, items from the seeking social
support for instrumental reasons and seeking social support for emotional reasons loaded onto a
single factor. Second, two items each from the mental disengagement and positive
reinterpretation and growth factors loaded below the standard .30 level. Finally, one item related
to alcohol and drug abuse, hypothesized as an aspect of mental disengagement, failed to load on
One important issue to note about the factor analytic results regards the correlations among
the subscales of the dispositional COPE. Specifically, Carver et al. (1989) employed an oblique
rotation in their factor analysis to allow for correlations among factors. Despite this
methodology, results revealed that, with few exceptions, the COPE (Carver et al.) subscales were
not strongly correlated. For example, the diametrically opposed subscales of acceptance and
denial were found to inversely correlate only moderately (r = -.21). The fact that these
intercorrelations were low to moderate implies that the subscales of the dispositional COPE are
relatively independent and can therefore be studied separately.
In addition to its factor structure, Carver and colleagues (1989) assessed the reliability of
the dispositional COPE using estimates of internal consistency and test-retest reliability.
Calculation of Cronbach' s alpha reliability coefficients for each subscale revealed that only the
mental disengagement subscale (a = .45) did not exhibit acceptable internal consistency
reliability. The test-retest reliability of the dispositional COPE (Carver et al.) subscales was
supported by data collected from two additional samples of college undergraduates, with
correlations ranging from .46 to .86.
Carver and colleagues (1989) also administered the situational version of the COPE to 156
participants three weeks after they completed the dispositional version. Exploratory factor
analysis of the responses revealed that the factor structure of the situational COPE (Carver et al.)
differed from the dispositional version in only two ways. First, unlike its counterpart on the
dispositional version, all items on the situational mental disengagement factor exhibited
acceptable loadings. Second, positive reinterpretation and growth on the situational version split
into two separate factors (although Cronbach's alpha for the four items together was adequate; a
=.74). In addition to vast similarities in factor structure, the dispositional and situational versions
of the COPE (Carver et al.) also paralleled in terms of the correlations among subscales. Finally,
reliability analysis of the situational COPE (Carver et al.) revealed that subscale alpha
coefficients tended to by higher than on the dispositional version.
Coping Measurement in Sport Psychology
The maj or instruments employed by researchers to measure coping in sport have been
adapted from those used in general psychology. Specifically, both the Ways of Coping checklist
(Folkman & Lazarus, 1980) and the COPE inventory (Carver et al., 1989) have been modified to
increase their applicability to the sport domain. In this section, each of these sport-modified
instruments will be discussed.
Modifications of the Ways of Coping checklist
The Ways of Coping checklist was modified by Madden and colleagues (Madden, Kirkby,
& McDonald, 1989; Madden, Summers, & Brown, 1990) for use in their studies on the coping
efforts of middle distance runners and basketball players. Their 54-item Ways of Coping
Checklist for Sport (WOCS) comprises eight subscales: (a) problem-focused coping, (b) seeking
social support, (c) general emotionality, (d) increased effort and resolve, (e) detachment, (f)
denial, (g) wishful thinking, and (h) emphasizing the positive. The subscales differ in number of
items, ranging from three (emphasizing the positive) to seven (problem-focused coping).
Although Madden and colleagues (Madden et al., 1989; Madden et al., 1990) provided
acceptable evidence for the internal consistency of the composite measure (a = .91), the
reliability of individual subscales was not reported. The lack of internal consistency estimates for
the eight subscales of the WOCS (Madden et al., 1989, 1990) has important implications for its
utility in research settings. First, because the number of items influences internal consistency
estimates, it remains possible that the individual subscales-especially those with only three
items--exhibit far lower, perhaps even unacceptable levels of reliability than the composite
measure (Crocker, Kowalski, & Graham, 1998). Second, Madden et al. (1989, 1990) proposed
that the eight subscales represent different coping strategies. However, without adequate
evidence for the reliability of each subscale, researchers are prevented from studying each of the
proposed coping strategies separately (Crocker et al., 1998). In addition to these reliability
concerns, there is little evidence for the validity of the WOCS (Madden et al., 1989, 1990) aside
from the actual studies for which it was developed (Crocker et al., 1998).
Crocker (1992) subsequently modified the original Ways of Coping checklist (Folkman &
Lazarus, 1980) to measure how athletes cope with a recent stressful athletic situation.
Modifications included (a) adding sport-specific items to the original Ways of Coping checklist,
(b) deleting irrelevant items, and (c) rewording existing items to increase sport relevance.
Crocker administered the modified instrument to 237 competitive athletes and instructed them to
indicate on a 4-point Likert-type scale the extent to which they used each strategy to cope with a
recent stressful situation (1 = not used; 4 = used very much). Principal axis factor analysis of this
data (with varimax rotation) resulted in a final 3 8-item version of the instrument which consisted
of the following eight subscales: (a) active coping, (b) problem focused, (c) social support, (d)
positive reappraisal, (e) wishful thinking, (f) self-control, (g) detachment, and (h) self-blame.
The subscales differed in number of items, ranging from two (self-blame) to seven (active
Reliability and validity evidence for Crocker' s (1992) Modified Ways of Coping checklist
provided only partial support for its use in sport psychology research (Crocker et al., 1998). First,
half of the subscales (positive reappraisal, self-control, detachment, and self-blame) did not
exhibit acceptable levels of internal consistency (as = .68, .60, .58, and .68 respectively). Second,
while it appears to be content valid, the modified Ways of Coping checklist (Crocker) is limited
by its lack of factorial validity (Crocker et al., 1998). Specifically, several items on the modified
Ways of Coping checklist loaded on factors inconsistent with both the original Ways of Coping
checklist (Folkman & Lazarus, 1980) and the WOCS (Madden et al., 1989, 1990).
Haney and Long (1995) subsequently modified the revised two-factor version of the Ways
of Coping checklist (Folkman, Lazarus, Dunkel-Schetter, et al., 1986) to measure higher-order
dimensions of coping with competitive stress. Specifically, these authors employed the
engagement and disengagement coping dimensions previously identified by Tobin, Holroyd,
Reynolds, and Wigal (1989). Engagement coping involves active efforts to manage the situation,
while disengagement coping refers to distancing oneself from the situation.
As part of the scale revision process, Haney and Long (1995) conducted a pilot study to
both (a) test the modified measure's higher-order factorial validity and (b) reduce the number of
items for ease of use. They administered an initial 50-item version of their measure, consisting of
the 46-item revised Ways of Coping Checklist (Folkman, Lazarus, Dunkel-Schetter, et al., 1986)
and four items from the suppression of competing activities sub scale of the COPE (Carver et al.,
1989), to a sample of 106 athletes. For each item, participants were instructed to indicate on a 4-
point Likert scale (0 = does not apply or not used; 3 = used a great deal) the extent to which they
used each coping strategy during the physical activity they had just performed. Confirmatory
factor analysis (CFA) revealed that the hypothesized factor structure of the initial measure poorly
fit the data. Subsequent CFAs were performed on progressively refined versions of the measure
until further modification did not result in significant model improvement. The final 18-item
version of the engagement-di engagement checklist (Haney & Long) comprised 1 1 engagement
items and seven disengagement items. The goodness-of-fit index (GFI) for the two-factor model
was .84 and the root mean square residual (RMSR) was .08. In addition, internal consistency
reliability was adequate for both the engagement (a = .82) and disengagement (a = .75)
Crocker and colleagues (1998) identified at least two limitations of the engagement-
disengagement checklist (Haney & Long, 1995). First, the GFI value of .84 was lower than the
acceptable level of .90. One possible explanation is that the items of the engagement-
disengagement checklist (Haney & Long) were taken from measures designed to assess coping at
a more specific level (i.e., the Ways of Coping checklist and COPE), and therefore may be
inadequate for measuring high-order coping dimensions. Second, inspection of the means and
standard deviations revealed that, for the disengagement subscale, the vast maj ority of athletes
reported only minimal use of disengagement coping. Finally, it should be noted that the small
sample size in the CFAs performed by Haney and Long may have impacted their observed fit
indices and subsequent conclusions (See Hu & Bentler, 1995, for a discussion of sample size
recommendations when performing a CFA).
Furthermore, Stone, Greenberg, Kennedy-Moore, and Newman (1991) offered three
limitations of the original Ways of Coping checklist (Folkman & Lazarus, 1980) that are equally
applicable to the three sport-modified checklists (Modified Ways of Coping checklist, Crocker,
1992; engagement-disengagement checklist, Haney & Long, 1995; WOCS, Madden et al., 1989,
1990). First, some items measure coping strategies that are not equally relevant in different sport
situations, thereby limiting the ability to compare subscale scores across situations. For example,
the item, "I avoided being with people in general," on Haney and Long's engagement-
disengagement checklist may not be relevant for team sport contexts where social withdrawal is
difficult (Crocker et al., 1998). Second, the three sport-modified measures do not identify a well-
defined period of time for which coping is assessed (Stone et al.). For example, in Crocker' s
Modified Ways of Coping checklist, participants are instructed to recall coping with a recent
stressful situation without respect to the particular stage of the situation. Given that coping has
been shown to be a dynamic process whereby coping efforts change as a stressful situation
progresses, this is a particularly important problem (Crocker et al.). Finally, the three sport
modifications of the Ways of Coping checklist are limited in that the instructions indicate a
response set that might be interpreted differently by different participants or by the same
participant between items (Stone et al.). Specifically, it is unclear whether the "extent" to which
a participant uses a particular coping strategy is defined as the frequency of use, the duration of
use, or the amount of effort expended to employ the strategy (Crocker et al.).
Modifications of the COPE inventory
Based on the above criticisms of the Ways of Coping checklist (Folkman & Lazarus,
1980), and the recommendations of Gould et al. (1993), sport psychology researchers have
preferred modifying the COPE inventory (Carver et al., 1989) to measure both dispositional
(Bouffard & Crocker, 1992; Eubank & Collins, 1999; Giacobbi & Weinberg, 2000) and
situational (Crocker & Graham, 1995; Crocker & Isaak, 1997; Dugdale et al., 2002; Gaudreau et
al., 2002; Gaudreau et al., 2001; Giacobbi & Weinberg, 2000; Ntoumanis & Biddle, 1998;
Pensgaard & Duda, 2003; Hammermeister & Burton, 2001) coping in sport. Because the COPE
subscales have been found to represent different coping strategies (Carver et al.), sport
psychology researchers have modified the COPE (Carver et al.) by selecting the most relevant
sub scales and rewording the items to meet their specific study purposes.
Crocker and Graham's (1995) modified COPE (MCOPE) has been the most consistently
employed sport modification of the COPE inventory (Carver et al., 1989). Specifically, the
MCOPE is a 48-item scale comprised of 12 four-item subscales designed to measure the coping
strategies used by athletes in recent stressful performance situations. Eight of the subscales were
selected directly from Carver and colleagues' situational COPE: (a) active coping; (b) planning;
(c) suppression of competing activities; (d) seeking social support for instrumental reasons; (e)
seeking social support for emotional reasons; (f) focusing on and venting of emotions; (g)
behavioral disengagement; and (h) denial. Three other MCOPE (Crocker & Graham) subscales
were derived from research by Crocker (1992) and Madden and colleagues (1990): (a) self-
blame (e.g., "I decided I was at fault for my performance"); (b) wishful thinking; (e.g., "I
day dreamed about a better performance"); and (c) increasing effort (e.g., "I put more effort into
my play"). The final sub scale of the MCOPE (Crocker & Graham), humor (e.g., "I made j okes
about my performance"), was developed for the original COPE (Carver et al.) but was excluded
from the published version of the instrument.
In addition to the selection of subscales, Crocker and Graham (1995) modified the
situational COPE (Carver et al., 1989) in four ways. First, items referring to a problem were
revised to refer to a performance. Second, the response set was changed to a 5-point Likert scale
anchored at 1 (used not at all/very little) and 5 (used very much). Third, the instruction set was
changed to, "For each item, indicate how much you used each strategy during the stressful
performance situation." Finally, several items were reworded to be understandable at a fifth
Crocker and Graham (1995) assessed the reliability of the MCOPE by administering it to
235 athletes. Of the 12 subscales, only denial (a = .42) demonstrated unacceptably low internal
consistency. Subsequent studies have also found the denial sub scale to be psychometrically
problematic (Crocker & Isaak, 1998; Gaudreau et al. 2001; Giacobbi & Weinberg, 2000). For
example, after administering both dispositional and situational versions of the MCOPE (Crocker
& Graham) to 237 college athletes, Giacobbi and Weinberg found that denial was the only
situational subscale to exhibit inadequate internal consistency (a = .67).
In addition to persistent problems with the denial subscale, several other MCOPE (Crocker
& Graham, 1995) subscales have been found to suffer from unacceptable reliability in several
studies. For example, Crocker and Isaak (1997) and Dugdale et al. (2002) obtained coefficient
alphas of .50 and .52 respectively for the behavioral disengagement subscale. Dugdale and
colleagues also found the internal consistency of the suppression of competing activities subscale
to be unsatisfactory (a = .59). Similarly, Gaudreau et al. (2002) found that the wishful thinking
and self-blame subscales exhibited inadequate reliability (a < .60) at one or more phases of their
three-phase study. Taken together, it appears that these subscales (behavioral disengagement,
suppression of competing activities, wishful thinking, and self-blame) are problematic from a
reliability standpoint. However, given that these studies varied with respect to instruction set
(i.e., specific stressful situation vs. most stressful recent situation), type of assessment (i.e.,
prospective vs. retrospective), and type of sport (i.e., team vs. individual), strong conclusions
cannot be drawn about the reliability of the above-mentioned subscales.
Evidence for the factorial validity of the MCOPE (Crocker & Graham, 1995) was provided
by Eklund, Grove, and Heard (1998), who administered the instrument to 621 athletes coping
with the stress of a performance slump. Four hypothetical models were tested using confirmatory
factor analysis: (a) the original 12-factor model of the MCOPE, (b) an 11i-factor model
(MCOPE-11AP) that combined the active coping and planning subscales, (c) an 11-factor model
(MCOPE-11SS) that combined the two social support subscales, and (d) a 10-factor model
(MCOPE-10) that contained both the active coping/planning and instrumental/emotional social
support combinations. None of the hypothesized models were clearly supported by the calculated
fit indices. Nevertheless, Eklund and colleagues advocated use of the 10-factor model (MCOPE-
10) to measure slump-related sport coping because in each of the other hypothesized models, the
active coping and planning subscales, as well as the two social support subscales were
exceedingly intercorrelated. Essentially, these subscales were intercorrelated to the point of
functional equivalence (Eklund et al.). However, as Eklund and colleagues pointed out, given
that all fit indices failed to reach the superior .90 level (Hoyle & Panter, 1995), further research
is clearly needed to provide more robust evidence for the factorial validity of the MCOPE.
Summary of Coping Measurement
In summary, while the development of the Ways of Coping checklist (Folkman & Lazarus,
1980) was a watershed event in the study of coping in general psychology, problems with its
psychometric properties have led researchers to more frequently employ the COPE (Carver et al.,
1989) in their studies. Echoing similar concerns with sport modifications of the Ways of Coping
checklist (Madden et al., 1989, Madden et al., 1990), researchers examining how athletes cope
with sport-specific stressors have preferred the sport modified COPE (Crocker & Graham, 1995;
Eklund et al., 1998). However, rather than relying on modifications of general coping
instruments, what is needed in sport coping research is the development of sport-specific coping
measures that tap the unique coping strategies utilized by athletes in response to athletic stressors
(Crocker et al., 1998).
Nevertheless, given that the Ways of Coping checklist (Folkman & Lazarus, 1980), the
COPE inventory (Carver et al., 1989), and their sport-adapted counterparts (Modified Ways of
Coping checklist, Crocker, 1992; engagement-di engagement checklist, Haney & Long, 1995;
WOCS, Madden et al., 1989, 1990) have been the maj or instruments in coping research, studies
that have employed these measures in both general and sport psychology will be reviewed in the
next section of this paper. Specific attention will be given to research with implications for the
structural and process aspects of coping as described C-M-R theory (Lazarus, 1991b, 1999).
Research on the Structural Aspects of Coping
As mentioned in the previous section on approaches to coping, trait (or structural) coping
has been described as a cognitive-behavioral style, a personality disposition, or a conditional
trait. Therefore, what follows is a discussion of the maj or findings of general and sport
psychology research in each of these three areas.
Coping as a style
General psychology research. As mentioned previously, an individual's coping style is
inferred from his or her habitual use of certain coping strategies across time and situations.
Whether or not coping is stable (consistent) or variable situationallyy determined) has important
theoretical and practical implications. First, if coping is stable, this would imply that
educationally or clinically based interventions would be less efficacious because individuals
would have highly routine or typical ways of responding to stress. Variability across time and
situations would suggest that coping is more malleable or open to change, thus allowing for more
effective efforts aimed at teaching individuals how to cope effectively with stress. However,
systematic examinations of cross-situational coping consistency are scarce in the general
psychology literature. Two notable exceptions were studies conducted by Folkman and
colleagues (Folkman, Lazarus, Dunkel-Schetter, et al., 1986; Folkman, Lazarus, Gruen, &
DeLongis, 1986) on the effects of coping on health status. These authors found that, across five
stressful encounters over several months, participants exhibited moderate consistency in their use
of some coping strategies (e.g., positive reappraisal), while displaying considerable variability in
others (e.g., seeking social support).
Instead of examining cross-situational consistency, the dominant method in general
psychology for investigating coping styles has been to calculate correlations between
dispositional and situational coping measures. Employing this method, researchers have provided
some evidence supporting the coping style view (Carver et al., 1989; Endler et al., 1994;
Kohlmann, 1993; Krohne, Slangen, & Kleeman, 1996; Miller, Brody, & Summerton, 1988). For
example, Carver and colleagues found that 9 of the 13 subscales on the dispositional COPE were
significantly correlated with their situational counterparts. Although only the significant
correlation for turning to religion was strong (r = .76; remaining rs ranged from .22 to .50),
Carver and colleagues suggested that a coping style could be inferred because the low to
moderate coefficients resembled those found in other studies between personality and coping
variables. A subsequent study by Endler et al. on coping with examination stress also found
moderate correlations between reports of dispositional and situational coping (rs ranged from .50
Although it appears from the above studies that coping is stylistic to some degree, a strong
conclusion cannot be reached for several reasons. First, as evinced by the reported correlations,
the amount of situational coping variance explained by coping styles is at least equal to that
observed across situations. Therefore, the predictive utility of coping styles is limited. Second,
several systematic studies of cross-situational coping consistency suggest that coping exhibits a
greater degree of cross-situational variability than consistency (Cohen & Lazarus, 1973;
Folkman & Lazarus, 1985; Folkman, Lazarus, Gruen, et al., 1986; Manzi, 1986). Finally, the
explanatory utility of the coping style approach is thus far inadequate in that the standard
methodology provides more information about the coping strategies than the individuals who
employ them in a stressful encounter (Lazarus, 1999).
Sport psychology research. Several studies have assessed the cross-situational
consistency of coping with athletic stressors (Bouffard & Crocker, 1992; Crocker & Isaak, 1997;
Gaudreau et al., 2001; Giacobbi & Weinberg, 2000). Taken together, these studies suggest that
coping in sport exhibits both consistency and variability across situations, supporting a trait-state
interactional approach to coping. For example, Bouffard and Crocker administered the COPE
(Carver et al., 1989) to 30 physically disabled individuals three times over a 6-month period.
Participants were requested to indicate the coping strategies they employed in response to recent
challenging physical activity situations. Results revealed that, except for the religion subscale,
more than 50% of the total coping variance was explained by a person-by-situation interaction,
indicating participants did not consistently use the same coping strategies. In contrast, Crocker
and Isaak found that the person-by-situation interaction component of coping variance exceeded
50% for only 3 of 12 subscales, thus supporting the coping style view. It is important to note,
however, that both the Bouffard and Crocker (1992) and Crocker and Isaak (1997) studies were
limited in that there was no formal assessment of whether participants actually perceived stress
in the situations to which they were responding.
Finally, employing the trait-state correlational approach dominant in general psychology,
Giacobbi and Weinberg (2000) administered both the trait and situational versions of the
MCOPE (Crocker & Graham, 1995) to 273 college athletes. For the situational version, athletes
were asked to indicate their coping thoughts and behaviors with respect to their two most
stressful sport situations. In addition, Giacobbi and Weinberg addressed the limitation of
previous sport coping consistency research by assessing whether the athletes actually perceived
these situations as stressful. Correlational analysis revealed that responses on the dispositional
COPE were moderately correlated with responses on the two situational versions (rs ranged from
.52 to .80), indicating that sport coping appears to be more stable than situationally variable.
Although the studies cited in this section appear to support the view that coping has both
trait and state determinants, they were each limited in that the methodologies employed were
theoretical. Specifically, these studies did not offer explanations regarding what aspects of an
individual (i.e., person variables) influenced whether their coping is variable or consistent.
Researchers in both general and sport psychology have addressed this limitation by examining
personality correlates of coping, and their findings will be reviewed next.
Coping as a personality disposition
General psychology research. Researchers in general psychology have consistently found
that coping thoughts and behaviors are related to aspects of personality (Boland & Cappeliez,
1997; Carver et al., 1989; Chang, 1998; David & Suls, 1999; Elliott, Chartrand, & Harkins,
1994; Endler et al., 1994; Giacobbi et al., 2004; Gunthert, Cohen, & Armeli, 1999; Halamandaris
& Power, 1999; Houston, 1977; O'Brien & DeLongis, 1996; Penley & Tomaka, 2002; Shen, Xu,
& Cui, 2002; Terry, 1994). For example, in their assessment of the convergent and divergent
validity of the COPE, Carver and colleagues administered the dispositional COPE along with
measures of optimism, self-esteem, locus of control, hardiness, Type-A personality, and trait
anxiety. Results revealed that all of these personality traits were significantly related to at least
three COPE subscales. Specifically, they found that optimism, locus of control, self-esteem,
hardiness, and Type-A personality appear to reflect an individual's tendency toward the use of
problem-focused coping, while trait anxiety reflects the tendency to use emotion-focused coping.
David and Suls (1999) conducted a more process-oriented examination of the relationship
between personality dimensions and coping. Specifically, to assess the Big Five personality traits
(Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness),
these authors administered the NEO Personality Inventory (NEO-PI; Costa & McCrae, 1985) to
95 community-residing adult males. In addition, participants completed Stone and Neale's
(1984) measure of daily coping for eight consecutive nights. Hierarchical linear modeling
(HLM), a statistical technique that allows for the simultaneous assessment of both within-
participant (i.e., process) and between- participant (i.e., structural) variability, was used to
determine relationships between the Big Five personality traits and daily coping efforts. Results
revealed that neuroticism was positively associated with the use of catharsis and relaxation, as
well as with the employment of more coping strategies overall. Similarly, participants who
scored higher on extraversion reported increased coping overall, with specific increases in
catharsis and redefinition. Openness to experience was positively associated with direct action,
but inversely associated with distraction. Finally, conscientiousness was negatively related to
religious coping. In addition to providing support for the prediction that personality traits are
related to coping, the study by David and Suls represented a maj or development in coping
research by demonstrating empirically that HLM, with its simultaneous analysis of within- and
between-participant relationships among variables, is a useful statistical approach that
complements the process-oriented methodological framework recommended by Folkman and
Lazarus (1985; Lazarus, 2000). Specifically, HLM improved upon the traditional analysis of
daily diary data, which entailed either (a) exclusive focus on between-participant (structure-
oriented) variation through the calculation of correlations from data aggregated across days and
participants, or (b) exclusive focus on within-participant (process-oriented) variation through the
calculation of multiple regression models. As David and Suls argued, these two analytical
techniques are inadequate when a study's purpose is to simultaneously examine both the daily
changes in appraisal and coping (i.e., within-participant variation), as well as the moderating
influence of person variables (e.g., Big Five personality traits) on relationships between appraisal
and coping (i.e., between-participant variation). Indeed, David and Suls reported that when they
aggregated coping data across days and participants, only 2 of a possible 40 zero-order
correlations between the Big Five personality traits and coping were significant. Therefore, in
addition to providing support for the C-M-R theory of emotion (Lazarus, 1991b, 1999), David
and Suls advanced the extant literature by applying a new, more sophisticated, statistical
technique especially suited for testing complex theoretical relationships. In my study, I attempted
to do the same by using structural equation modeling to test a system of explanatory
relationships predicted by C-M-R theory.
Sport psychology research. Few studies in the extant sport psychology literature have
examined personality predictors of coping (See Giacobbi & Weinberg, 2000; Grove & Heard,
1997; Grove, Lavallee, & Gordon, 1997; Krohne & Hindel, 1988; and Ntoumanis et al., 1999 for
exceptions). In general, these studies have supported the view that coping is related to aspects of
personality. For example, Grove and Heard assessed dispositional optimism, trait sport
confidence, and slump-related coping in a sample of 336 team and individual sport athletes.
Hierarchical regression analyses revealed that a significant, albeit modest, amount of variance in
task- and emotion-focused coping was predicted by higher levels of optimism and trait sport
confidence (Grove & Heard).
Other studies examining the relationships between coping and aspects of personality have
shown that problem-focused coping is more likely to be employed by task-oriented athletes
(Ntoumanis et al., 1999), while aspects of emotion-focused coping are preferred by athletes who
rate highly in trait anxiety (Giacobbi & Weinberg, 2000; Krohne & Hindel, 1988), athletic
identity (Grove et al., 1997), and ego orientation (Ntoumanis et al.). In my study, I further
examined personality predictors by assessing the relationships between goal orientations and
General psychology research. The most sophisticated account of the structural aspects of
coping is the conditional trait approach, which, as stated previously, predicts personality traits to
influence coping styles only in situations rendered functionally equivalent by the specific
personality trait (Wright & Mischel, 1987). Although researchers have yet to empirically identify
specific environments that are rendered functionally equivalent by personality traits, the
influence of situational context on coping has been demonstrated (Folkman & Lazarus; 1980;
Schwartz & Stone, 1993; Stern, Norman, & Komm, 1993). Specifically, in their often-cited study
on the coping behaviors of community-residing adults, Folkman and Lazarus (1980) found that
the degree to which participants employed problem- or emotion-focused coping depended in part
on whether the stressor concerned their work, health, or family. Folkman and Lazarus's (1980)
Ending is important because it distinguished situational and contextual variations in coping.
Coping contexts represent the broad, relatively stable aspects of life, while coping situations are
the emotional encounters within them (Folkman & Lazarus, 1980). For example, within the
broad context of academics, a student might be confronted with specific stressful situations
related to Einal examinations, interpersonal relationships with teachers and classmates, high
workload, etc. The C-M-R theory of emotion (Lazarus, 1991b, 1999) posits that the specific
coping strategies employed by this student (e.g., active coping, seeking social support, or denial)
might vary with respect to each stressful situation, while the total repertoire of coping strategies
(e.g., active coping, seeking social support, and denial) remains relatively stable across stressors
within the academic context. Therefore, the context of a stressful encounter represents a
structural aspect of coping, while situations represent a process aspect of coping.
Sport psychology research. As is the case with general psychology, systematic
identification of environments rendered functionally equivalent by personality traits are virtually
nonexistent in the extant sport psychology literature. However, differences in coping were shown
to depend on the situational context in the study by Crocker and Isaak (1997) discussed earlier.
Specifically, in addition to assessing the training-related coping efforts of youth swimmers, these
authors measured the same athletes' coping in a competitive context. Crocker and Isaak found
that, in contrast to their relatively consistent use of specific coping strategies across training
stressors, youth swimmers exhibited considerable coping variability when dealing with
competitive stress. One possible explanation offered by these authors was that training sessions
are highly regimented, and therefore place a highly consistent set of demands on swimmers.
Familiarity with the adaptational demands of the training context might allow swimmers to
develop a repertoire of coping behaviors specific to the training context. While this explanation
is plausible, Crocker and Isaak did not assess personality in their study, and therefore more
theoretical explanations remain open to investigation.
Summary of research on the structural aspects of coping
With regard to the structural aspects of coping, the findings of research in general
psychology have been consonant with those of sport psychology. First, both fields have
demonstrated that coping is somewhat consistent across situations (Bouffard & Crocker, 1992;
Crocker & Isaak, 1997; Folkman, Lazarus, Dunkel-Schetter, et al., 1986; Folkman, Lazarus,
Gruen, et al., 1986). However, due to methodological limitations, and the theoretical nature of
the research, more sophisticated inquiry into the cross-situational consistency of coping is
warranted. Second, researchers in both general and sport psychology have shown coping to be
related to aspects of personality (e.g., trait anxiety, optimism, neuroticism, extraversion, and goal
orientation; Carver et al., 1989; David & Suls, 1999; Giacobbi & Weinberg, 2000; Grove &
Heard, 1997; Krohne & Hindel, 1988; Ntoumanis et al., 1999). Finally, although researchers in
both general and sport psychology have found that coping is related to the situational context
(e.g., work, family, training, and competition; Crocker & Isaak, 1997; Folkman & Lazarus,
1980), the specific personality traits that influence these contextual coping differences have not
been identified. In the next section of this paper, attention will turn toward research examining
the process aspects of coping.
Research on the Process Aspects of Coping
As alluded to throughout this review, the C-M-R theory of emotion (Lazarus, 1991b, 1999)
emphasizes the dynamic relationship between personal resources (e.g., personality, coping
resources, etc.) and situational demands as the main predictors of coping in a stressful encounter.
Three unresolved empirical issues emerge from this perspective. First, do individuals use a single
coping strategy to deal with a given stressor, or do they employ multiple coping efforts? Second,
do individuals use the same coping strategy throughout a stressful situation, or do they employ
different coping efforts at different stages of the stressor? Finally, what constitutes effective
coping in a given stressful situation? What follows is a review of the major general and sport
psychological research related to each of these issues.
Dimensionality of coping
General psychology research. Lazarus and colleagues (Folkman & Lazarus, 1980;
Lazarus, 1999) argued that unidimensional conceptualizations of coping are inadequate because
stressful encounters typically involve more than one psychological facet (e.g., goals, threats,
emotions). Initial evidence in support of this argument was provided by naturalistic observation
(Mechanic, 1962; Moos & Tsu, 1977; Murphy, 1974; Visotsky, Hamburg, Goss, & Lebovits,
1961). For example, Moos and Tsu noted that individuals coping with an incapacitating physical
illness had to simultaneously manage both the illness itself (e.g., pain, interactions with hospital
staff, surgical procedures) and the emotions that resulted from having a debilitating illness (e.g.,
sadness, despair, anger).
Researchers in general psychology have subsequently provided support for the
multidimensionality of coping (Carver et al., 1989; David & Suls, 1999; Folkman & Lazarus,
1980; Folkman & Lazarus, 1985; Folkman & Lazarus, 1988; Manzi, 1986). For example, in their
study of middle-aged adults, Folkman and Lazarus (1980) interviewed participants seven times
at 4-week intervals about stressful situations they had experienced during the previous month. In
addition to a description of the stressful events, participants indicated on the Ways of Coping
checklist (Folkman & Lazarus, 1980) the manner in which they coped with each encounter.
Participants reported a total of 1,332 stressful situations over the course of the study. Results
revealed that participants employed both problem- and emotion-focused coping strategies in 98%
of the situations. Similar results were reported by Folkman and Lazarus (1985) who found that
94% of participants employed both problem- and emotion-focused strategies across the stages of
a stressful college examination. Furthermore, in a study of the coping and emotions of young and
old adults, Folkman and Lazarus (1988) found that situational variations in emotion were
significantly predicted by a combination of problem- and emotion-focused strategies. Together,
the results of these studies confirm that individuals employ a complex combination of coping
efforts aimed at both altering the source of stress (i.e., problem-focused coping) and regulating
their emotions (i.e., emotion-focused coping).
Sport psychology research. Support for the multidimensionality of coping in sport has
been demonstrated both qualitatively (Giacobbi et al., 2004; Gould et al., 1993) and
quantitatively (Crocker & Graham, 1995; Grove et al., 1997; Hammermeister & Burton, 2001;
Kim & Duda, 2003; Ntoumanis & Biddle, 1998; Ntoumanis et al., 1999). In their often-cited
study, Gould and colleagues examined coping by conducting semi-structured interviews with 17
former U.S. national champion figure skaters. Eight general coping dimensions emerged from
content analysis of the interviews: (a) rational thinking and self-talk, (b) positive focus and
orientation, (c) social support, (d) time management and prioritization, (e) precompetitive mental
preparation and anxiety management, (f) training hard and smartly, (g) isolation and deflection,
and (h) ignoring the stressor. Indicative of the complexity of coping, skaters employed coping
strategies from at least two of these eight dimensions irrespective of the source of stress.
Furthermore, in cognitive-motivational-relational terms, the eight coping dimensions served both
problem-focused (e.g., training hard and smartly) and emotion-focused (e.g., isolation and
deflection) functions. Therefore, the qualitative results reported by Gould and colleagues clearly
support the cognitive-motivati onal-relational view that coping is a complex, multidimensional
process. Subsequently, researchers employing quantitative methodologies have confirmed the
multidimensionality of coping in sport across athletic populations and stressors (Crocker &
Graham, 1995; Grove et al., 1997; Hammermeister & Burton, 2001; Kim & Duda, 2003;
Ntoumanis & Biddle, 1998; Ntoumanis et al., 1999).
Dynamics of coping
General psychology research. As mentioned previously, C-M-R theory (Lazarus, 1991b,
1999) adopts the process-oriented view that coping efforts constantly change both within and
between stressful situations. In addition, Folkman and Lazarus (1985) recommended that in order
to adequately examine coping as a process, researchers should make repeated assessments of
how an individual actually copes with a specific stressful encounter. Several studies in the
general psychology literature that have followed these recommendations have supported the
view that coping is a dynamic process (David & Suls, 1999; Folkman & Lazarus, 1980; Folkman
& Lazarus, 1985; Levy-Shiff, Dimitrovsky, Shulman, & Har-Even, 1998; Schwartz & Stone,
For example, in the frequently cited study by Folkman and Lazarus (1985), participants
completed a modified form of the Ways of Coping checklist (Fokman & Lazarus, 1980) during
each of three stages of a college midterm examination. Coping was assessed (a) 2 days prior to
the exam (the anticipatory stage), (b) 5 days after the exam (the waiting stage), and (c) 5 days
after grades were announced (the outcome stage). Results revealed that problem-focused coping
was at its highest during the anticipatory stage and decreased significantly during the waiting
stage. In contrast, participants increasingly distanced themselves from the stressor (i.e., used
emotion-focused coping) during the waiting stage. Third, students' use of wishful thinking and
distancing significantly decreased during the outcome stage of the exam.
Folkman and Lazarus (1985) explained these results in terms of the varying adaptational
demands that each exam stage places on students. Specifically, they argued that increased
problem-focused coping during the anticipatory stage reflected the situational requirement that
students study in order to earn a good grade on their exam. In contrast, the 5-day period after the
exam required students to shift their focus from studying to simply waiting for their grades to be
posted. Therefore, students increasingly distanced themselves from the exam situation as a
means of emotional regulation. Folkman and Lazarus (1985) argued that, in order to regulate
their emotions during the waiting stage, students might either hope for a good outcome (i.e., use
wishful thinking) or distance themselves from the situation altogether. To summarize then, the
results reported by Folkman and Lazarus (1985) supported C-M-R theory (Lazarus, 1991b, 1999)
by showing that, as a stressful situation unfolds, differential efforts are required to cope with the
changing adaptational demands of the person-environment transaction.
Sport psychology research. Process-oriented research on the dynamics of coping is scarce
in the sport coping literature (See Gaudreau et al., 2002; Gaudreau et al., 2001; Giacobbi et al.,
2004, for exceptions). However, preliminary support for the view that coping changes as a
stressful encounter progresses was provided by Gaudreau and colleagues (2001), who performed
a sport-specific replication of the study by Folkman and Lazarus (1985). Specifically, these
authors assessed the coping efforts of 33 Canadian male youth golfers competing in a qualifying
tournament. Gaudreau and colleagues (2001) administered the MCOPE-10 (Eklund et al., 1998)
and two exploratory coping subscales (mental disengagement and positive reappraisal) to the
golfers 2 hours prior to the competition, 15 minutes after the competition, and 24 hours after the
competition. Multivariate analyses of variance (MANOVAs) conducted for each coping subscale
revealed that golfers' use of social support, suppression of competing activities, increased effort,
active coping/planning, wishful thinking, and behavioral disengagement changed across the three
phases of the competition, thereby supporting the process-oriented view that coping is dynamic.
Nevertheless, despite these promising findings, further research is clearly needed.
General psychology research. In hopes of identifying which situational coping strategies
are more adaptive than others, researchers in general psychology have attempted to link coping
efforts with important adaptational outcomes. Many of these studies have indicated that coping is
directly related to emotions (Bowman & Stern, 1995; Cochrane, & Slade, 1999; Folkman &
Lazarus, 1985, 1988; Hahn, 2000; Morris & Engle, 1981; Zeidner, 1995), psychological well-
being (Cochrane & Slade; Essex, Seltzer, & Krauss, 1999; Holland & Holohan, 2003; Meyer,
2001; Terry, Mayocchi, & Hynes, 1996; Wilkinson, Walford, & Espenes, 2000), burnout
(Koleck, Bruchon-Schweitzer, Thiebaut, Dumartin, & Sifakin, 2000) and academic performance
(Morris & Engle, 1981; Zeidner, 1995). These studies represent important advances in the coping
literature because they go beyond the mere description of coping by attempting to link coping
with important outcomes such as subj ective well-being, emotional outcomes, and satisfaction in
a given context. However, the above-cited studies were limited in several ways. First, Lazarus
(1999) argued that the effectiveness of a particular coping strategy is not necessarily reflected by
the important outcomes of the situation. For example, it is possible for individuals to cope
effectively without necessarily experiencing positive affect, goal achievement, or other desired
outcomes (Lazarus, 1999). Second, from a methodological standpoint, Lazarus (1995) pointed
out that retrospective one-shot studies of coping and its outcomes assess these constructs
simultaneously, and are therefore inadequate for concluding that the coping strategies caused the
Due to these analytical pitfalls, some researchers have examined relationships between
perceived coping effectiveness (i.e., the extent to which an individual perceives that his or her
situational coping efforts were effective) and important outcomes (Brauer, 2001; Iwasaki, 2003;
Jean, Paul, & Beatty, 1999). For example, in a recent longitudinal study of the ways that
university students cope with everyday stressors, Iwasaki found that perceived coping
effectiveness was negatively associated with mental illness, while related positively to
psychological well-being and stress reduction. However, as this is one of only a handful of
studies on perceived coping effectiveness in the general psychology literature, broad conclusions
cannot be made.
Another way to assess coping effectiveness that has garnered far more research attention in
general psychology is Folkman' s (1992) goodness-of-fit approach. As stated previously,
Folkman proposed that the effectiveness of a particular coping strategy depends upon whether or
not it is appropriate for the adaptational demands of the encounter, i.e., the extent to which an
individual's secondary appraisal of control over the stressor matches his or her situational coping
strategies. Specifically, Folkman argued that problem-focused coping is most effective in
situations amenable to change (i.e., within personal control) while emotion-focused coping is
most effective in uncontrollable situations.
Studies examining the relationship between coping strategies and secondary appraisal of
control have generally supported Folkman's (1992) predictions (David & Suls, 1999; Folkman &
Lazarus, 1980; Folkman & Lazarus, 1985; Folkman, Lazarus, Dunkel-Schetter, et al., 1986;
Forsythe & Compas, 1987; Martin, 1993; Peacock & Wong, 1996; Reese et al., 1997; Wilson,
Stelzer, & Bergman, 1995). For example, in their study on the coping efforts of community-
residing adults, Folkman and Lazarus (1980) also examined the influence of situational
controllability appraisals on the coping process. After reporting a stressful event and completing
the Ways of Coping checklist (Folkman & Lazarus, 1980), participants answered in a binary
fashion four questions related to their situational appraisals: (a) Is this situation one that you
could change or do something about?; (b) Is this situation one that must be accepted or gotten
used to?; (c) Is this situation one that you needed to know more about before you could act?; and
(d) Is this situation one in which you had to hold yourself back from doing what you wanted to
do? Results revealed that situations appraised as controllable (i.e., something constructive could
be done or more information was needed) were characterized by increased problem-focused
coping, while emotion-focused coping increased in situations appraised as uncontrollable (i.e.,
must be accepted or had to hold back).
Subsequent studies by Reese et al. (1997) and Peacock and Wong (1996) have also
confirmed that situational coping strategies are related to an individual's appraisal of control
over a situation. Specifically, in a longitudinal study, Reese and colleagues found that
participant' s perception of control over negative life events was positively associated with
problem-focused coping across assessments. In contrast, participants who perceived a lack of
control over negative life events were more likely to employ emotion-focused coping.
Research relating goodness-of-fit to outcome-related variables has provided inconsistent
support for Folkman' s (1992) predictions. Specifically, several studies have indeed demonstrated
the predicted relationships for both problem- and emotion-focused coping (Conway & Terry,
1992; Park, Folkman, & Bostrom, 2001; Sorgen & Manne, 2002; Vitaliano, DeWolfe, Maiuro,
Russo, & Katon, 1990). For example, Sorgen and Manne tested goodness-of-fit in a study on the
coping efforts of children who have cancer. They found that (a) problem-focused coping was
positively associated with perceptions of control, (b) emotion-focused coping was negatively
associated with perceptions of control, and (c) goodness-of-fit was negatively related to
However, several other studies have demonstrated the predicted relationships for only one
of the coping functions, (Osowiecki & Compas, 1999; Zakowski, Hall, Cousino, & Baum, 2001).
For example, Osowiecki and Compas found that only problem-focused coping interacted with
perceived control to influence anxiety/depression symptoms, while Zakowski and colleagues
found that only emotion-focused coping interacted with perceived control to impact stress.
Still other studies have failed to demonstrate either of the predicted relationships (Masel,
Terry, & Gribble, 1996; O'Rourke & Cappeliez, 2002; Roberts, 1995). For instance, in an
examination of the relationships between perceived control, psychological symptomatology, and
coping with daily hassles, Roberts found that while participants tended to match coping with
perceived control, goodness-of-fit did not affect symptomatology. Therefore, the extant general
psychology literature appears to be equivocal with regard to Folkman' s (1992) goodness-of-fit
Sport psychology research. The dominant methodology for examining coping
effectiveness in sport has been to relate coping efforts to important sport outcomes. These studies
have found problem-focused coping to be associated with positive affect (Crocker & Graham,
1995; Gaudreau et al., 2002; Ntoumanis & Biddle, 1998; Ntoumanis et al., 1999), athletic career
satisfaction (Kim & Duda, 2003), sport enjoyment (Kim & Duda, 2003), successful transition to
university by freshmen athletes (Giacobbi et. al., 2004), and desire to continue in sport (Kim &
Duda, 2003). Emotion-focused coping has been found to relate to negative affect (Crocker &
Graham, 1995; Gaudreau et al., 2002; Ntoumanis & Biddle, 1998; Ntoumanis et al., 1999),
athletic career dissatisfaction (Kim & Duda, 2003), lack of sport enj oyment (Kim & Duda,
2003), and lack of desire to continue in sport (Kim & Duda, 2003).
Interestingly, and reflective of a maj or limitation in the sport coping literature, the few
studies that have examined the association between coping and athletic performance have found
that neither problem- nor emotion-focused coping are associated with athletic performance,
whether measured objectively or subj ectively (Dugdale et al., 2002; Pensgaard & Duda, 2003).
For example, Pensgaard and Duda studied the coping efforts of 61 Nordic athletes competing at
the 2000 Sydney Olympic Games. Coping was measured by the COPE (without the turning to
religion subscale; Carver et al., 1989), while performance was measured obj ectively by the
athletes' placing in the Games, and subjectively by performance satisfaction. Results revealed
that coping was associated with neither obj ective nor subj ective performance.
Although unable to demonstrate a relationship between coping and Olympic performance,
Pensgaard and Duda (2003) did find such a relationship when assessing the extent to which the
Olympians perceived their coping efforts to be effective. Specifically, hierarchical regression
analyses revealed that high perceived coping effectiveness positively predicted both obj ective
and subjective result.
Researchers also have found relationships between perceived coping effectiveness and
important sport outcomes other than performance (Kim & Duda, 2003; Ntoumanis & Biddle,
1998; Ntoumanis et al., 1999). For example, in a study of the mediating effect of perceived
coping effectiveness, Ntoumanis and Biddle (1998) administered six subscales of the MCOPE
(Crocker & Graham, 1995) and the PANAS (Watson et al., 1988) to 356 British university
athletes. In addition, Ntoumanis and Biddle (1998) instructed the athletes to rate the extent to
which they perceived each of the six MCOPE (Crocker & Graham) sub scales to be effective.
Results showed that the perceived coping effectiveness of seeking social support, venting of
emotions, and behavioral disengagement mediated the impact of these coping strategies on
affect. Specifically, athletes reported higher positive and lower negative affect when they
perceived seeking social support as effective. Similarly, when perceived as effective, behavioral
disengagement was related to higher positive affect. Finally, when athletes judged venting of
emotions to be effective, they experienced lower negative affect than when it was ineffective.
Ntoumanis and Biddle's (1998) finding that perceived coping effectiveness mediated the
relationship between coping and affect is important for two reasons. First, the amount of
explained affect variance was significantly increased for the three coping strategies (seeking
social support, venting of emotions, and behavioral disengagement) when perceived coping
effectiveness was included in the analyses. Second, and more importantly, the findings of
Ntoumanis and Biddle support the cognitive-motivational-relational view that problem-focused
coping is not necessarily more adaptive than emotion-focused coping. Specifically, these authors
found that the emotion-focused coping strategies of behavioral disengagement and venting of
emotions had a positive influence on affect when they were perceived as effective. Stated
differently, Ntoumanis and Biddle found that both problem- and emotion-focused coping can
enhance affect, provided that an athlete perceives these coping efforts as effective. Nevertheless,
because few studies have examined perceived coping effectiveness in sport, its contribution to
empirical knowledge remains incomplete. Therefore, to address this limitation of the extant
literature, I assessed perceived coping effectiveness, and examined its influence on affective
Research on coping effectiveness is also necessary with respect to Folkman's (1992)
goodness-of-fit hypothesis. The few studies in sport examining the relationship between coping
and secondary appraisal of control have only partially supported the goodness-of-fit hypothesis
(Dugdale et al., 2002; Hammermeister & Burton, 2001; Haney & Long, 1995; Kim & Duda,
2003). For example, Kim and Duda assessed the manner in which 318 Division I university
athletes perceived control over and coped with psychological difficulties in competition. Results
revealed that an athlete's appraisal of control over a stressor was positively related to active
coping strategies (i.e., planning, social support). Partial support for the goodness-of-fit
hypothesis was also provided by Hammermeister and Burton, who found that endurance athletes
who appraised low control over potential race performance threats and employed emotional
social support to cope with these threats experienced significantly greater precompetitive
cognitive state anxiety as measured by the CSAl-2 (Martens, Burton, Vealey, Bump, & Smith,
1990). Stated in cognitive-motivati onal -relati onal term s, these authors found that when an athl ete
perceives that he or she can do relatively little to change a threatening situation, and responds
with emotion-focused coping, a negatively-valenced emotion (i.e., anxiety) is likely to be
In one of the most sophisticated studies in the sport coping literature on the topic, Haney
and Long (1995) examined both outcome and goodness-of-fit related indicators of coping
effectiveness. A total of 178 basketball, soccer, and field hockey players participated in a two-
round skills competition relevant to their sport (i.e., free throws and penalty shots). Athletes
completed a three-item measure of perceived control over performance stress 5 minutes prior to
both rounds. In addition, they reported their use of engagement and disengagement coping in
response to performance stress 5 minutes after both rounds. Haney and Long proposed a model
of coping effectiveness whereby, for both rounds of competition, perceived control would predict
in-competition coping behaviors, which would in turn predict performance for that round.
Furthermore, in their proposed model, performance in round 1 was hypothesized to predict
control appraisal prior to round 2. Path analysis of the data supported goodness-of-fit predictions
only with respect to disengagement coping. Specifically, for round 1, athletes who perceived
high pre-round control used less disengagement coping and performed better. In addition,
athletes who performed well in round 1 perceived even more control prior to round 2.
Furthermore, these athletes also employed less disengagement coping and performed better in
round 2. In contrast, no significant relationships were found between perceived control and
engagement coping in either of the rounds.
Although the three studies discussed above (Hammermeister & Burton, 2001; Haney &
Long, 1995; Kim & Duda, 2003) offered support for the goodness-of-fit approach, strong
conclusions cannot be drawn for at least two reasons. First, neither Kim and Duda nor Haney and
Long measured problem- and emotion-focused coping as defined by the C-M-R theory of
emotion (Lazarus, 1991b, 1999). It is possible that relationships between control and coping as
assessed in an engagement/di engagement (or approach/avoidant) framework are dissimilar to
those predicted by C-M-R theory. Second, all three of these studies assessed coping within a
competitive performance context, potentially limiting the athletes' coping options. Due to the
unique ego-involving nature of competition, athletes might be either (a) less likely to appraise
their performance stress as uncontrollable, or (b) less likely to avoid/disengage from their
performance. If this contextual effect did occur, then it is understandable that the demonstrated
relationships were not completely consonant with C-M-R theory (Lazarus, 1991b, 1999).
Therefore, to address the above limitations in my study, I assessed secondary appraisal of
control, problem-focused coping, and emotion-focused coping in the context of a non-
Summary of research on the process aspects of coping
From the preceding review of literature in general and sport psychology, several
conclusions can be drawn regarding the process aspects of coping. First, studies have
consistently shown that coping is a complex, dynamic process in that individuals (a) employ both
problem- and emotion-focused coping within the same situation (Carver et al., 1989; Crocker &
Graham, 1995; Folkman & Lazarus, 1980; Giacobbi et al., 2004; Gould et al., 1993), and (b)
change the ways they cope as the stressful situation progresses (Folkman & Lazarus, 1985;
Gaudreau et al., 2001). Second, the effectiveness of coping efforts in a given stressful encounter
has been shown to relate to (a) important cognitive, emotional, and behavioral outcomes
(Crocker & Graham, 1995; Folkman & Lazarus, 1985; Gaudreau et al., 2002; Zeidner, 1995), (b)
the extent to which an individual perceives his or her coping to be effective (Iwasaki, 2003;
Pensgaard & Duda, 2003), and (c) the extent to which situational coping efforts are
adaptationally compatible with perceived situational control (Kim & Duda, 2003; Park et al.,
2001). Finally, and perhaps most importantly, research over the past 20 years has clearly shown
that the C-M-R theory of emotion (Lazarus, 1991b, 1999) is an empirically viable theoretical
framework for sport psychology researchers (Gould, 1996).
Despite these conclusions, several issues remain unresolved with respect to C-M-R theory
(Lazarus, 1991b, 1999). First, studies that systematically examine how individuals cope with
chronic stress are few and far between in the extant literature. The relative lack of research is
surprising because chronic stress has been linked to a variety of deleterious health outcomes
(e.g., heart disease, somatic complaints, burnout; http://www.apa.org). Second, despite research
suggesting that coping is linked to the particular context in which it occurs (e.g., work, family,
health, training, competition), this area of research has received minimal attention. Third, as
illustrated by the divergent approaches that researchers have taken (i.e., important outcomes,
perceived coping effectiveness, and goodness-of-fit), measurement of coping effectiveness
remains both conceptually and empirically unresolved. Finally, although several studies have
examined multiple aspects of C-M-R theory (Lazarus, 1991b, 1999) simultaneously, researchers
have yet to link antecedents of appraisal, cognitive appraisals, coping, and emotion within the
same study, as was suggested by Lazarus (1999).
As outlined in the preceding review, the C-M-R theory of emotion (Lazarus, 1991b, 1999)
has received considerable support in the extant sport psychology literature. However, several
aspects of the theory remain relatively unexplored. Specifically, few studies have examined the
ways that athletes appraise, cope with, and emotionally respond to chronic, non-competitive
stress. In addition, person variables other than the Big Five personality dimensions and trait
anxiety (e.g., achievement goal orientations) have received little attention. Furthermore, while
preliminary findings indicate that perceived coping effectiveness may be an important mediator
in the relationship between coping and important outcomes, this issue is by no means resolved,
and therefore requires further exploration. Finally, few studies have simultaneously examined all
four maj or constructs of C-M-R theory (i.e., antecedents of appraisal, appraisal, coping, and
emotion; Lazarus, 1991b, 1999). Therefore, in my study, I addressed each of these limitations in
an attempt to refine scientific knowledge regarding the viability of Lazarus' s (1991Ib, 1999)
theory in the sport domain.
Furthermore, several studies have identified managing athletic and academic time demands
as a maj or source of stress for student-athletes at various competitive levels (Giacobbi et al.,
2004; Ortez, 1997; Petrie & Stover, 1997; Tracey & Corlett, 1995). Of these studies, club sport
athletes have received the least attention. Therefore, in my study, I examined the manner in
which club sport athletes appraise, cope with, and experience affective responses to the stress of
managing athletic and academic time demands. Based on the C-M-R theory of emotion (Lazarus,
1991b, 1999), I tested the relationships between these constructs (See Figure 2-2) in a proposed
model informed by both theory (Duda, 1989; Roberts, 1992; Lazarus, 1991b, 1999) and previous
research (Crocker & Graham, 1995; Gaudreau et al., 2002; Hammermeister & Burton, 2001;
Haney & Long, 1995; Kim & Duda, 2003; Maier et al., 2003; Ntoumanis & Biddle, 1998;
Ntoumanis et al., 1999; Pensgaard & Duda, 2003) in general and sport psychology.
In the proposed model, achievement goal orientations represented person variables as
antecedents of appraisal. Primary appraisals of threat and challenge, along with secondary
appraisal of perceived control represented the appraisal variables in the proposed model. In
addition, the model included both task- and avoidance-focused coping efforts. Finally, the
perceived effectiveness of these task- and avoidance-focused coping efforts served as mediators
of the relationships between coping and the outcome variables in the model, positive and
negative affect. What follows is a discussion of research that formed the rationale for testing the
hypothesized relationships between these constructs in the proposed model.
Achievement Goal Orientations and Primary Appraisal
The influence of achievement goal orientations on primary appraisals of threat and
challenge have yet to be fully explored in the extant literature. Therefore, the relationships
between antecedents of appraisal and other constructs depicted in Figure 2 are based on the
theoretical predictions of both the C-M-R theory of emotion (Lazarus, 1991b, 1999) and
achievement goal theory in sport (Duda, 1989; Roberts, 1992). Specifically, athletes scoring
highly on a measure of task orientation are more likely to seek challenges (Roberts, 1992), and
therefore should perceive managing athletic and academic time demands as a challenge. In
contrast, athletes who report a strong ego orientation tend to exhibit maladaptive motivational
patterns, and therefore are more likely to perceive managing athletic and academic time demands
as threatening (Roberts).
Achievement Goal Orientations and Secondary Appraisal
With respect to the relationships between goal orientations and secondary appraisal of
control, the knowledge base is also incomplete. For example, Pensgaard (1999) administered
measures of goal orientation and perceived control to 19 Norwegian national athletes competing
in the 1996 Atlanta Olympic Games, and found that task orientation was associated with high
perceived control. However, due to the small sample size in Pensgaard's study, and the lack of
subsequent empirical corroboration, strong conclusions regarding the relationships between goal
orientations and perceptions of control cannot be made.
Therefore, in the proposed model I tested in my study, inclusion of these relationships was
predominately based on the theoretical predictions of C-M-R theory (Lazarus, 1991b, 1999) and
achievement goal theory in sport (Duda, 1989; Roberts, 1992). Specifically, consistent with C-
M-R theory (Lazarus, 1991b, 1999), task and ego orientations will be conceptualized as
antecedents of secondary appraisal of perceived control. Consistent with Roberts's (1992)
arguments, athletes adopting a strong task orientation have a differentiated view of ability, and
therefore should be more likely to appraise stressors as within their control. In contrast, athletes
who score high on a measure of ego orientation adopt an undifferentiated view of ability, and
therefore should be more likely to appraise stressors as outside of their control.
Achievement Goal Orientations, Coping, and Affect
Research in sport psychology is also limited with respect to the relationships between goal
orientations, coping, and affect. In one of the few studies on this topic, Ntoumanis et al. (1999)
examined whether coping is a mediator of the relationships between goal orientations and
coping. Three hundred fifty-six British university athletes completed the Task and Ego
Orientation in Sport Questionnaire (TEOSQ; Duda & Nicholls, 1992). In addition, these athletes
were instructed to recall a recent important competition in which they perceived threat or
challenge. Athletes' use of three emotion-focused coping strategies (seeking social support for
instrumental reasons, suppression of competing activities, and effort) and three problem-focused
coping strategies (seeking social support for emotional reasons, behavioral disengagement, and
venting of emotions) were assessed by subscales of the MCOPE (Crocker & Graham, 1995).
Distancing, another emotion-focused coping strategy, was assessed by its subscale on the revised
Ways of Coping checklist (Folkman, Lazarus, Dunkel-Schetter, et al., 1986). Finally, the
athletes' affective responses to the stressor were measured by the PANAS (Watson et al., 1988).
SEM was employed to examine the complex relationships between these variables. In
accordance with their predictions, Ntoumanis and colleagues found that problem-focused coping
mediated the relationship between task orientation and positive affect, while emotion-focused
coping mediated the relationship between ego orientation and negative affect. Because these
results are both consistent with theoretical predictions, and were derived through highly
sophisticated statistical analyses (i.e., SEM), the proposed model in my study included (a) task-
focused coping as a mediator of the task orientation/positive affect relationship, and (b)
avoidance-focused coping as a mediator of the ego orientation/negative affect relationship.
Primary Appraisal, Secondary Appraisal, and Coping
Folkman (1992) predicted that problem-focused coping will predominate in situations
appraised by individuals as challenging and controllable, while emotion-focused coping
predominates in situations appraised as threatening and uncontrollable. Research in both general
and sport psychology has generally supported this prediction. In contrast, few studies in the
extant literature have simultaneously examined the relationships between threat and challenge
primary appraisals, secondary appraisal of control, and problem- and emotion-focused coping.
One exception in the general psychology literature was a study by Portello and Long
(2001), who found that perceived threat and low perceived control were associated with the use
of disengagement coping. In the sport coping literature, Hammermeister and Burton (2001)
found that perceived threat, low perceived control, and emotion-focused coping (i.e., emotional
social support) were predictors of endurance athletes' competitive state anxiety.
Despite the relative lack of research examining all three of these constructs simultaneously
(i.e., primary appraisal, secondary appraisal, and coping), evidence does exist supporting the
component relationships between (a) primary and secondary appraisal, and (b) secondary
appraisal and coping (Campbell & Jones, 2002; Dugdale et al., 2002; Ferguson et al., 2000;
Haney & Long, 1995; Kerig, 1998; Kim & Duda, 2003; Solomon, Mikulincer, & Benbenishty,
1989). For example, perceived control has been shown to relate positively to challenge appraisals
(Campbell & Jones) and negatively to threat appraisals (Ferguson et al.). Furthermore, research
has generally shown that problem-focused coping is related to high perceived control, while
emotion-focused coping is related to low perceived control (David & Suls, 1999; Folkman, 1984;
Folkman & Lazarus, 1980, 1985; Folkman, Lazarus, Dunkel-Schetter, et al., 1986; Martin, 1993;
Peacock & Wong, 1996; Reese et al., 1997).
For example, Folkman and Lazarus (1980) found that situations appraised as controllable
(i.e., something constructive could be done or more information was needed) were characterized
by increased problem-focused coping, while emotion-focused coping increased in situations
appraised as uncontrollable (i.e., they must be accepted or one had to hold back). Similarly,
Reese et al. (1997) and Peacock and Wong (1996) found that participant' s perception of control
over negative life events was positively associated with problem-focused coping across
assessments, while participants who perceived a lack of control over negative life events were
more likely to employ emotion-focused coping strategies. In sport, findings by Haney and Long
(1995) and Kim and Duda (2003) have confirmed the general psychology research.
Therefore, based on the theory and research presented in this section, I proposed that club
sport athletes who appraise managing athletic and academic time demands as challenging and
controllable are more likely to employ task-focused coping. In contrast, athletes who appraise
this stressor as threatening and uncontrollable are more likely to employ avoidance-focused
Primary Appraisal and Affect
In the C-M-R theory of emotion, Lazarus (1991b, 1999) predicts that challenge appraisals
cause positive emotions (e.g., hope) while threat appraisals cause negative emotions (e.g.,
anxiety). As specific positive and negative emotions are viewed as subsets of a broader affective
construct (Ekkekakis & Pettruzello, 2000), one can infer that the predictions of C-M-R theory
(Lazarus, 1991b, 1999) regarding emotion can extend to affect as well. Therefore, I hypothesized
that challenge appraisals will positively predict positive affect, while threat appraisals will
positively predict negative affect.
In addition to the theoretical bases for the relationships between primary appraisals and
affect predicted in Figure 2, empirical support has been provided by several studies in the general
psychology literature (Dopke & Milner, 2000; Kuiper, McKenzie, & Belanger, 1995; Maier et
al., 2003; Pierce, Lydon, & Yang et al., 2001). For example, Kuiper and colleagues (1995) found
that challenge appraisal was positively related to positive affect, while studies by Pierce and
colleagues (2001) and Dopke and Milner (2000) revealed that threat appraisal was positively
related to negative affect.
Although these studies individually supported one of the two predicted primary appraisal-
affect relationships (i.e., challenge/positive affect and threat/negative affect), only recently have
both relationships been demonstrated in the same study. Specifically, utilizing both experimental
and survey methodologies, Maier and colleagues (2003) measured the primary appraisals (i.e.,
threat and challenge), affect, and cardiovascular reactivity of 56 male participants while they
completed a computerized mental arithmetic task. In support of their hypotheses, positive affect
was significantly predicted by challenge appraisal (R2 = .44), while negative affect was
significantly predicted by threat appraisal (R2 = .32). Therefore, in my study, I predicted that
student-athletes who perceive the difficulty of managing athletic and academic time demands as
challenging will experience greater positive affect. In contrast, athletes who perceive this stressor
as threatening will experience greater negative affect.
Coping, Perceived Coping Effectiveness, and Affect
Research in sport psychology has repeatedly demonstrated that problem-focused coping is
positively related to positive affect, and that emotion-focused coping is positively related to
negative affect (Anshel & Anderson, 2002; Crocker & Graham, 1995; Gaudreau et al., 2002).
For instance, Gaudreau and colleagues administered the MCOPE-10 (Eklund et al., 1998), an
exploratory positive reappraisal subscale, and the PANAS (Watson et al., 1988) to 62 Canadian
male golfers competing in a regional golf tournament 2 hours before, 1 hour after, and 24 hours
after the competition. Results revealed that problem-focused coping (i.e., active coping/planning,
increased effort, positive reappraisal, seeking social support, and suppression of competing
activities) was positively related to positive affect at all three phases of the competition. In
contrast, emotion-focused coping responses (i.e., venting of emotions, humor, and behavioral
disengagement) were positively associated with negative affect at each phase of the competition.
However, recent evidence has expanded the problem-focused coping/positive affect and
emotion-focused coping/negative affect relationships to include perceived coping effectiveness
as an important mediating variable. Specifically, empirical support for this refinement was
provided by Ntoumanis and Biddle (1998), who found that perceived coping effectiveness
mediated the coping/affect relationship. Stated differently, according to the results of Ntoumanis
and Biddle, both problem-focused coping and emotion-focused coping resulted in either positive
or negative affect, depending on whether the athlete perceived their problem- and emotion-
focused coping as effective for dealing with stress. Therefore, in my study perceived coping
effectiveness was viewed as a mediator of the coping/affect relationship. Specifically, if an
athlete perceives his or her coping to be effective, these coping efforts will result in higher
positive affect and lower negative affect. In contrast, if an athlete perceives his or her coping to
be ineffective, these coping efforts will result in lower positive affect and higher negative affect.
I tested three hypotheses in my study. First, club sport athletes who score highly on a
measure of task orientation will appraise the chronic stress of managing athletic and academic
time demands as challenging and within their control, which will result in increased use of
problem-focused coping to deal with the challenge. Second, club sport athletes who score highly
on a measure of ego orientation will appraise the chronic stress of managing athletic and
academic time demands as threatening and outside their control, and will thus employ emotion-
focused coping to deal with the threat. Finally, perceived coping effectiveness will mediate the
relationship between coping and affect such that, for all club sport athletes, coping efforts
perceived as effective will result in increased positive affect, while those perceived as ineffective
will result in increased negative affect.
Figure 2-1. Predicted relationships between the maj or constructs of the cognitive-motivational-
relational theory of emotion.
Figure 2-2. Proposed path model of the relationships between goal orientations, primary
appraisal, secondary appraisal, coping, perceived coping effectiveness, and affect.
Dotted lines represent negative relationships.
Participants and Procedure
With assistance from the Coordinator of Sport Clubs, I contacted the presidents of 24 sport
clubs at a large university in the Southeast United States, and received permission to recruit
participants at a scheduled meeting or practice. During recruitment, I introduced potential
participants to the purpose of the study, informed them of what their participation would entail,
and assured them that their responses would be anonymous and confidential. By providing their
institutionally approved informed consent, 295 club sport athletes (154 women and 141 men)
volunteered to participate in this study.
Upon consenting to participate, I gave 80 (47 women and 33 men) of the 295 volunteers
unique numerical codes that allowed one-time access to a secure website containing the full
measurement protocol, and instructed these participants to access the website as soon as possible
after experiencing a situation in which they had difficulty balancing their athletic and academic
demands. Upon entry into the website, participants recalled, briefly described, and indicated their
appraisals and coping responses to the situation. The decision to employ this same-day
retrospective design was based on the recommendation of Smith et al. (1999), who found
retrospective bias in the assessment of situational coping strategies. Such a design represented an
improvement over the typical methodology employed by sport coping researchers whereby
coping is assessed several months subsequent to the adaptational encounter (Crocker & Graham,
1995; Giacobbi & Weinberg, 2000; Ntoumanis & Biddle, 1998; Ntoumanis et al., 1999).
I invited the remaining 215 participants (107 women and 108 men) to complete the
measurement protocol via traditional paper-and-pencil administration at the time of recruitment.
Because these participants completed the surveys at the time of recruitment, and therefore were
unable to wait until they experienced a situation in which they had difficulty balancing their
athletic and academic demands, the time frame referenced for the stressor was 7 days. While this
did not meet the recommendations of Smith and colleagues (1999), it nevertheless was an
improvement over the typical 6-month time frame pervasive in the extant sport coping literature.
Finally, in order to ensure the confidentiality of their responses, I placed participants' completed
survey packets in a box, which was then sealed and stored in a secure location.
I computed a Multivariate Analysis of Variance (MANOVA) to examine possible method
effects. This test was not significant [Wilk' s 32 = .53, F (19, 32) = 1.52, p = .145], indicating that
subsequent analyses need not distinguish between participants who completed the measurement
protocol online and those who completed the paper and pencil protocol.
The overall response rate was 83.1% (245/295), with 37.5% (30/80) of the volunteers
invited to complete the measurement protocol via the study website doing so, and 100.0%
(215/215) of the participants who volunteered to complete the paper-and-pencil measures doing
so. Of the 245 surveys completed by participants (125 from women and 120 from men), 9 were
discarded because of excessive missing data, and 13 were discarded because the participant
indicated that they had not experienced the stressor. Therefore, the final sample included 223
athletes(122 women and 101 men) from the following sport clubs: Crew, Cuong Nhu, Cycling,
Equestrian, Fencing, Handball, Judo, Men's and Women's Lacrosse, Men's and Women's
Rugby, Men's and Women's Soccer, Synchronized Swimming, Tennis, Triathlon, Men's and
Women's Ultimate Frisbee, Underwater Hockey, Men's and Women's Volleyball, Men's and
Women's Water Polo, and Wrestling.
The remaining demographic characteristics of the final sample are presented in Table 3-1.
On average, participants were 20 years old and had approximately four years of organized
competitive experience in their sport. In addition, they appeared to be heavily involved in
academics given that they devoted approximately three times as many hours to school than they
did to club sports while keeping a semester courseload of over 13 credit hours. The vast maj ority
of participants were Caucasian, while the sample was evenly distributed with respect to class
level. Finally, it should be noted that while I obtained my sample in a non-random manner, its
demographic characteristics were nevertheless representative of the broader university student
population with respect to gender and ethnicity.
The first page of the measurement protocol assessed basic demographic information
related to each participant's age, gender, race/ethnicity, and university classification (e.g.,
freshman, sophomore, etc.). The demographics questionnaire (See Appendix A) also requested
that participants indicate (a) the sport they played, (b) the approximate number of hours per week
they spent in sport-related activities (e.g., training, practicing, competing, traveling), (c) the
approximate number of hours per week they spent in school-related activities (e.g., attending
class, studying, getting assistance from a tutor, taking exams), (d) the number of years they had
been participating in their sportss, and (e) the number of credit hours in which they were
The Task and Ego Orientation for Sport Questionnaire (TEOSQ)
The TEOSQ (Duda & Nicholls, 1992; See Appendix B) is a 13-item instrument comprised
of a 7-item task orientation (TO) subscale and a 6-item ego orientation (EO) subscale. I
requested respondents to think of a time when they felt most successful in sport, and indicate on
a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree) their agreement with items
reflecting task-oriented and ego-oriented criteria. I calculated subscales scores for TO and EO by
summing their respective item responses. The subscales of the TEOSQ have been found to
demonstrate acceptable test-retest (Duda & Nicholls, 1992) and internal consistency reliability
(Duda & Whitehead, 1999). In addition, the orthogonal two-factor solution of the TEOSQ has
been repeatedly validated through confirmatory factor analysis (CFA; Chi and Duda, 1995;
Guivernau & Duda, 1994). Examples of items on the TO sub scale are, "I feel most successful in
sport when I work really hard," and "I feel most successful in sport when I learn something that
is fun to do." Sample items from the EO sub scale are, "I feel most successful in sport when the
others can't do as well as me," and "I feel most successful in sport when I score the most points."
Consistent with previous research in both sport and general psychology, I instructed
participants to briefly describe a situation they had experienced in the preceding 7 days in which
they had difficulty managing their academic and athletic time demands. As will become evident
in the following pages, the situation they described served as the basis for their responses on the
appraisal and coping measures I used in my study.
Upon obtaining permission from the instrument' s second author (personal communication,
3/10/2005) I employed the threat and challenge subscales of the Stress Appraisal Measure
(SAM; Peacock & Wong, 1989) to measure primary appraisals (See Appendix C). I instructed
participants to indicate on a 5-point scale (1 = not at all; 5 = extremely) the extent to which each
of the eight items represented the stressful time management situation they previously described.
I calculated threat and challenge subscale scores by summing their respective item scores. Across
three validation studies, Peacock and Wong reported average internal consistency estimates of
.71 for the threat subscale and .73 for the challenge subscale. In addition, the convergent validity
of the SAM threat and challenge subscales of the SAM was supported by theoretically
meaningful correlations with measures of locus of control, psychological symptoms, and
dysphoric mood (Peacock & Wong). Items on the SAM include, "How threatening is this
situation?" from the threat sub scale and, "How eager am I to tackle this problem?" from the
I assessed secondary appraisal of control with a modified version of the three-item
personal control subscale of the Revised Causal Dimension Scale (CDSII; McAuley, Duncan, &
Russell, 1992; See Appendix C). Specifically, I instructed athletes to indicate on a 9-point scale
(with 9 representing the greatest level of each variable) the extent to which the stressful situation
they previously described was (a) "manageable by you," (b) "something you can regulate," and
(c) "something over which you have power." I calculated a personal control sub scale score by
summing its item scores. The personal control subscale has demonstrated adequate reliability in
both general and sport coping research (Dugdale et al., 2002; McAuley et al., 1992). For
instance, across their four initial validation studies, McAuley and colleagues reported an average
internal consistency reliability of .79. Similarly, in the sport coping literature, Dugdale and
colleagues (2002) reported a Cronbach's alpha of .80 for the three items.
The COPE Inventory
I used sub scales selected from the situational COPE inventory (Carver et al., 1989) to
measure the coping thoughts and behaviors of participants in my study. In its complete form, the
COPE (See Appendix D) is comprised of 15 conceptually distinct four-item subscales informed
by Lazarus's (1991b, 1999) C-M-R theory of emotion, Carver and Scheier' s (1985) model of
behavioral self-regulation, and previous empirical research findings. The 15 COPE subscales are
as follows: (a) active coping (ACT), (b) planning (PLAN), (c) suppression of competing
activities (SCA), (d) restraint coping (RES), (e) seeking social support for instrumental reasons
(SSI), (f) seeking social support for emotional reasons (SSE), (g) focusing on and venting of
emotions (VENT), (h) behavioral disengagement (BDIS), (i) mental disengagement (MDIS), (j)
positive interpretation and growth (PRG), (k) denial (DEN), (1) acceptance (ACC), (m) turning to
religion (REL), (n) humor (HUM), and (0) alcohol/drug use (ALC). In their initial development
of the COPE, Carver and colleagues found that these 15 COPE subscales loaded onto 4 second-
order factors. Task coping (TCOPE) was comprised of the ACT, PLAN, and SCA subscales.
Avoidance coping (ACOPE) consisted of the MDIS, BDIS, and DEN subscales. Emotion coping
(ECOPE) was comprised of the SSI, SSE, and VENT subscales. Finally, cognitive coping
(CCOPE) consisted of the RES, PRG, and ACC subscales.
For at least two reasons, I assessed only the subscales of the TCOPE (ACT, PLAN, and
SCA) and ACOPE (MDIS, BDIS, and DEN) second-order factors in this study. First, as
discussed previously, achievement goal theory (Duda & Nicholls, 1992) predicts that task-
oriented individuals believe their skill level is due to factors under personal control (e.g.,
practice), while ego-oriented individuals believe their skill level is due to facturs outside of
personal control (e.g., innate ability). In addition, C-M-R theory (Lazarus, 1999) predicts that
controllable stressful situations lead to active attempts to alter the stressor, while uncontrollable
stressors result in attempts to disengage from the stressful situation and focus instead on
ameliorating its negative emotional effects. Therefore, theory suggests that task-oriented
individuals are more likely to employ task coping strategies (i.e., ACT, PLAN, and SCA), while
ego-oriented individuals are more likely to employ avoidance coping strategies (i.e., MDIS,
BDIS, and DEN).
Second, Skinner, Edge, Altman, and Sherwood (2003) recently published an extensive
review of the extant literature on coping taxonomies and recommended that, in order to
demonstrate consistent and coherent links between coping and adaptation, researchers should
employ higher-order measures of coping that exhibit both within-family adaptational similarity
and between-families adaptational distinctiveness. To this end, they identified several
empirically and theoretically supported higher-order coping families that best met these criteria.
Two of these higher-order coping families, problem solving and escape, closely resemble the
task and avoidance second-order coping factors found by Carver and colleagues in their
development of the COPE. Specifically, problem solving strategies serve the adaptive function of
adjusting ones actions to be effective, and include instrumental action, planning, and
strategizing-strategies analogous to the ACT, PLAN, and SCA subscales (Skinner et al., 2003).
Escape strategies serve the adaptive function of escaping a noncontingent environment, and
include cognitive avoidance, behavioral avoidance, and denial--strategies analogous to the
MDIS, BDIS, and DEN subscales (Skinner et al., 2003). Therefore, in the context of the present
study the work of Skinner and colleagues suggests that, of the COPE subscales and second-order
factors, assessment of task and avoidance higher-order coping is most likely to illuminate
relationships between coping and affective responses to stress.
The six situational COPE sub scales I used in this study have been shown to represent
reliable and valid measures of their respective situational coping strategies in both general and
sport psychology with the exception of MDIS (Carver et al, 1989; Eklund et al., 1998).
Specifically, in their initial development of the COPE, Carver and colleagues calculated an
internal consistency estimate of reliability for MDIS of a = .45, while in the sport psychology
literature, Eklund et al. found the MDIS subscale to exhibit inadequate reliability (a = .55).
Despite these findings, I included the MDIS sub scale in the present study due to its potential
salience as a coping strategy that club sport athletes might employ to deal with the difficulty of
balancing their athletic and academic demands.
In its situational form, COPE items are phrased in the past tense to assess coping responses
in specific stressful encounters (e.g. "I made a plan of action"). Using this format, I instructed
participants to answer on a 4-point Likert-type scale (1 = I didn't do this at all; 4 = I did this a
lot) the degree to which they employed a given coping strategy to deal with the stressful time
management situation they previously described. I calculated subscale scores for ACT, PLAN,
SCA, MDIS, BDIS, and DEN by summing their respective item scores. In addition, I calculated
composite scores for TCOPE and ACOPE by summing the scores for their component subscales.
Perceived Coping Effectiveness Measure
Similar to the procedure employed by Ntoumanis and Biddle (1998), I measured perceived
coping effectiveness with 24 items corresponding to those on the COPE. Specifically, I requested
participants to indicate on a 7-point Likert-type scale (1 = not at all effective; 7 = very much
effective) the overall extent to which they believed their coping efforts were effective in dealing
with the stressful time management situation they previously described. For example, with
respect to the ACT item, "I concentrated my efforts on doing something about it," participants
responded to the question, "Overall, how effective was this in dealing with the stressful
situation?" I calculated subscale scores for the effectiveness of ACT, PLAN, SCA, MDIS, BDIS,
and DEN by summing their respective item scores,. In addition, I computed composite scores for
TCOPE and ACOPE by summing the scores for their component subscales.
The Positive Affect and Negative Affect Schedule (PANAS)
I used the Positive Affect and Negative Affect Schedule (PANAS; Watson et al., 1988; See
Appendix E) to measure the affective experiences of participants in my study. The PANAS
(Watson et al) is a 20-item instrument comprising two subscales that represent orthogonal
dimensions of affect: Positive Affect (PA) and Negative Affect (NA). The PA subscale consists
of 10 positively valenced emotions (e.g., excited, enthusiastic, inspired), while the NA sub scale
contains 10 negatively valenced emotions (e.g., distressed, hostile, irritable). Participants
indicated on a 5-point Likert-type scale (1 = very slightly or not at all; 5 = extremely) the extent
to which they experienced each of the listed emotions during the proceeding 7 days. I computed
PA and NA sub scale scores by summing their respective item scores. Watson and colleagues
reported that the PANAS exhibited acceptable factorial validity and internal consistency (a = .86
to 90 for PA; a = .84 to 87 for NA). In the sport coping literature, studies by Crocker and
Graham (1995), Ntoumanis and Biddle (1998), and Ntoumanis et al. (1999) have further
supported the internal consistency of both the PA (as = .89, .91, and .91 respectively) and NA
(as = .84, .83, and .83 respectively) sub scales.
I began the preliminary analyses by using the Expectation-Maximization (EM) algorithm
to replace any missing data that were retained in the final sample. In simple terms, the EM
algorithm proceeds through cycles in which it first replaces missing data with expected values
from a regression model of the non-missing data, and then maximizes the likelihood that the
values are correct by recalculating the regression model using the newly replaced values. Once a
full data set was obtained, I calculated the means and standard deviations for all items and their
scales. Furthermore, because the parameter estimation method used in this study assumes
multivariate normality in the data, I examined the distributional properties of each item. and
transformed any item that exhibited excessive non-normality into its natural logarithm.
Next, I computed a series of MANOVAs to determine whether gender, ethnicity, or class
level had an impact on the obtained results, and conducted follow-up univariate F-tests for any
significant multivariate findings. Finally, I calculated bivariate correlations to assess the
associations between demographic variables (age, years of competition, school hours per week,
or club sport hours per week).
Psychometric Analyses (The Measurement Model)
Within an SEM analytical framework, the proposed structural equation model should not
be tested until the measurement protocol is first demonstrated to be reliable and valid (Anderson
& Gerbing, 1988; Byrne, 1998). Following this recommendation, I performed a multi-construct
CFA on the measurement model using the maximum likelihood (ML) method of parameter
estimation provided in LISREL 8.53 (Joireskog & Soirbom, 1996). The measurement model was
comprised of first-order latent variables the first-order latent variable indicators, and
measurement error terms related to each indicator. Because each perceived coping effectiveness
indicator was embedded within the same item as its corresponding coping indicator (i.e., they
were two parts of the same question), it was reasonable to assume that the measurement errors
for the corresponding indicators would be related. Therefore, I specified the measurement model
to include two-way paths relating the error term of each coping indicator to the error term of its
corresponding perceived effectiveness indicator. Finally, to remove origin and scale
indeterminancy, I set the latent variable variances equal to 1.0 in the measurement model.
I evaluated the construct validity of the measurement protocol using the root mean square
error of approximation (RMSEA; Steiger, 1990), its associated confidence interval (CI), the test
of close fit (polose), and the model chi-square per degrees of freedom value (Q). A good-fitting
model produces (a) an RM SEA less than .05, with its entire CI below this value, (b) a failure to
rej ect the null hypothesis of the pelose test of close fit, and (c) a Q value of less than 3.0 (Browne
& Cudeck, 1992). Alternatively, a reasonable model has an RMSEA between .05 and .08, while
a value greater than .10 indicates unacceptably poor fit.
Due to the plethora of model fit indices currently available within an SEM framework, a
justification for using the RMSEA in this study is worth noting. Specifically, the RMSEA offers
several statistical and practical advantages over most of the other fit indices available in the SEM
literature. First, the RMSEA penalizes for lack of parsimony such that complex models do not
automatically fit better than simple models. Second, unlike other fit indices, the RMSEA is
unaffected by increases in sample size. Third, because the RMSEA is a point estimate, a CI can
be computed to measure its precision. Finally, its associated hypothesis test (polose) focuses on
close fit, rather than the impractical standard of exact fit assessed by the X2 goodness of fit test.
In addition to overall construct validity, analysis of the measurement model provided
evidence regarding the external consistency reliability, convergent validity, and discriminant
validity of the measurement protocol. External consistency reliability was supported if the latent
variables did not contain cross-loaded indicators. Convergent validity was supported if (a) the
factor loadings of a latent variable were significant and greater than .707, (b) the average
variance extracted (AVE) by a latent variable was greater than .50, and (c) no more than 10% of
its standardized fitted residuals exceeded |2.00| (Bagozzi & Yi, 1988; Fornell & Larcker, 1981).
Finally, discriminant validity was established if (a) the 95% CI for a latent variable's correlation
with another latent variable did not include 1.0, or (b) its AVE was greater than its squared
correlation with another latent variable (Anderson & Gerbing, 1988; Fornell & Larcher, 1981).
I used different procedures, depending on the type of latent variable, to assess internal
consistency reliability in this study. For the first-order latent variables other than coping, I
calculated Cronbach's alpha coefficient (a). For the first-order coping variables (ACT, PLAN,
SCA, MDIS, BDIS, and DEN), I followed the recommendations of Gerbing, Hamilton, and
Freeman (1994), who argued that when first-order latent variables serve as "indicators" of