Content uploaded by Jennifer Cumming
Author content
All content in this area was uploaded by Jennifer Cumming
Content may be subject to copyright.
JOURNAL OF APPLIED SPORT PSYCHOLOGY, 16: 318–332, 2004
Copyright © Association for Advancement of Applied Sport Psychology
ISSN: 1041-3200 print / 1533-1571 online
DOI: 10.1080/10413200490517986
Motivational Profiles and Psychological Skills Use
within Elite Youth Sport
CHRIS HARWOOD
Loughborough University
JENNIFER CUMMING
University of Birmingham
DAVID FLETCHER
University of Wales Institute
This study investigated associations between achievement goal orientations and reported psy-
chological skill use in sport. Five hundred seventy three elite young athletes completed the Per-
ceptions of Success Questionnaire (POSQ; Roberts, Treasure, & Balague, 1998) and the Test
of Performance Strategies (TOPS; Thomas, Murphy, & Hardy, 1999). Cluster analysis revealed
three distinct goal profile groups: Cluster 1—Higher-task/Moderate-ego (n=260); Cluster
2—Lower-task/Higher-ego (n=120); and Cluster 3—Moderate-task/Lower-ego (n=119).
A MANOVA revealed a significant multivariate effect, Pillai’s Trace =.11, F(16, 1076) =
3.75, p=.001, η2=.05, with post hoc tests determining that higher-task/moderate-ego ath-
letes reported using significantly more Imagery, Goal setting, and positive Self-talk skills
when compared with Lower-task/Higher-ego and/or Moderate-task/Lower-ego athletes. These
findings are discussed with respect to the potential role that achievement goals play in the
application and development of psychological skills in youth sport.
Recent research investigating motivated behavior within the physical domain (i.e., sport,
physical education and exercise) has predominantly been conducted from a social cognitive
perspective (Roberts, 2001). This approach places an emphasis on an individual’s cognitions
to explain the reasons for their behavior in different social contexts. Across the sport and
physical activity domains, the bulk of this research has focused upon lower level or non-
elite participation (for recent reviews, see Dishman, 1994; Duda, 2001; Duda & Hall, 2001;
Ntoumanis & Biddle, 1999; Roberts, Treasure, & Kavussanu, 1997). While this line of inquiry
has considerably advanced our understanding of motivation in these settings, some scholars
(e.g., Hardy, Jones, & Gould, 1996) have highlighted a dearth of scientific information relating
Received 23 January 2003; accepted 1 July 2003.
Address correspondence to Chris Harwood, PhD, Sport and Exercise Psychology Research Group,
School of Sport and Exercise Sciences, Loughborough University, Loughborough, Leicestershire, LE11
3TU. E-mail: C. G. Harwood@lboro.ac.uk
318
MOTIVATIONAL PROFILES AND PSYCHOLOGICAL SKILLS 319
to elite performance. As a result, sport psychologists have often had to tentatively speculate
and make ‘educated guesses’ when considering how individual differences in achievement
motivation may influence the behavior of high-level athletes. The aim of this study was to
contribute to the bridging of this gap in our knowledge.
Nicholls’ achievement goal theory (Nicholls, 1984, 1989) formed the theoretical and con-
ceptual framework of the present investigation. Applied extensively around the world within
sport and exercise psychology research, a central tenet of Nicholls’ approach relates to the
different ways individuals construe their level of competence. Within the context of sport,
achievement motivation is conceptualized as a function of the subjective meaning that a per-
former assigns to success or failure. Thus, as Duda and Hall (2001) remark, the comparisons
performers make to formulate their perceptions of competence will likely affect their choice
to invest in a specific activity, the amount of effort they expend in the activity, and their level
of persistence when confronted by a challenge to perform the activity.
Further to the aforementioned variations in behavior, Nicholls (1984, 1989) argued that
individuals initiate different achievement goal states as a direct consequence of the personal
meaning they assign to achievement. In essence, these goal states represent the transient view
of achievement one holds for a specific activity or task. According to Nicholls, two achieve-
ment goals termed ‘task involvement’ and ‘ego involvement’ can operate in any achievement
situation. A task-involved goal (i.e., a state of task involvement) is activated when an individual
focuses on the development of competence, where their sense of achievement is self-referenced
and subjectively associated with personal mastery, progress and self-improvement. In contrast,
an ego-involved goal (i.e., a state of ego involvement) is triggered when an individual focuses
on the demonstration of superior competence, where their sense of achievement and personal
adequacy is normatively referenced and associated with showing a superior capacity of ability.
Even though it is these goal states that form the situational determinants of behavior in
Nicholls’ achievement goal theory, researchers have typically studied individuals’ disposi-
tional goal orientations, representing the trait-like tendencies to be task or ego-involved in
achievement situations. In actual fact, the majority of research over the past decade has cen-
tered on a broader examination of the beliefs, cognitions, affective responses, and values that
characterize athletes according to their levels of task and ego orientation (see Duda, 2001;
Duda & Hall, 2001; Duda & Whitehead, 1998; Roberts, 2001, for detailed reviews). As goal
orientations represent orthogonal, as opposed to a bipolar, cognitive schema (cf. Nicholls,
1989), recent research has recognized that individuals are capable of being high, moderate or
low in both orientations in combination. Subsequently, in contrast to early research that tended
to examine the correlates of task and ego orientation separately, momentum has gathered for
research that has incorporated goal profiling as a technique that allows the investigator to study
the cognitive-behavioral patterns of individuals with different combinations (i.e., levels) of task
and ego orientation (see Cumming, Hall, Harwood, & Gammage, 2002; Fox, Goudas, Biddle,
Duda, & Armstrong, 1994; Hodge & Petlichkoff, 2000; Roberts, Treasure, & Kavussanu, 1996;
Walling & Duda, 1995; Wang & Biddle, 2001; White, 1998).
Two consistent messages appear to have emerged from this overall body of research. Firstly,
in support of Nicholls’ (1984) original theorizing, task orientation tends to be associated
with positive psychological responses. For example, correlational studies have found positive
relationships between task orientation and corresponding levels of enjoyment (Hom, Duda, &
Miller, 1993; Kim & Gill, 1997), intrinsic motivation (Duda, Chi, Newton, Walling, & Catley,
1995; Ntoumanis, 2001) and adaptive moral values and behaviors (Carpenter & Yates, 1997;
Duda, Olson, & Templin, 1991; Dunn & Dunn, 1999). These findings support the call for
practitioners to “do whatever possible to make sure an athlete’s task orientation is robust”
(Duda, 2001, p. 163).
320 C. HARWOOD ET AL.
A second message is that although ego orientation has been frequently associated with
negative psychological qualities such as maladaptive moral behavior (Carpenter & Yates,
1997; Duda et al., 1991), it does not always lead to negative cognitive-motivational outcomes
when combined with a corresponding level of task orientation (e.g., a high-task/high-ego goal
profile). Indeed, as a growing collective of research, the findings from goal profiling studies
attest to the cognitive-motivational benefits of a moderate-to-high ego orientation provided
that it is combined with a corresponding level of task orientation. Hence, performers reporting
a ‘complementary balance’ of both the desire to demonstrate superior abilities over others and
to progress and develop through personal mastery are more motivated to engage in tasks that
maximize achievement (Hodge & Petlichkoff, 2000).
Despite these advances in the data handling and analysis of goal orientations, this area is
characterized by a paucity of research addressing the applied role that such combinations may
play in regulating the use of, or engagement in psychological skill behaviors associated with
practice and competition. This lack of understanding was of central interest to the current
investigation. At present, studies have provided researchers with an initial awareness of the
relationship between goal orientations and those self-reported psychological responses asso-
ciated with competition, such as multi-dimensional anxiety (Hall & Kerr, 1997; Ntoumanis &
Biddle, 1999; White, 1998; White & Zellner, 1996), cognitive interference (Hatzigeorgardis
& Biddle, 1999; Newton & Duda, 1993) and coping (Ntoumanis, Biddle, & Haddock, 1999).
However, our understanding of how individual differences in motivation relate to the use of
strategies/skills that function to control such psychological responses is very limited. For ex-
ample, research shows that task orientation tends be negatively associated with symptoms of
anxiety and negative thinking and positively associated with adaptive coping efforts. However,
we do not know if or how goal orientations relate to the prevalence of those basic psycho-
logical strategies (e.g., relaxation, positive self-talk, imagery, or goal setting) that applied
researchers and practitioners believe regulate such responses (Hardy et al., 1996; Williams,
2001).
Twoexceptions are provided by Cumming, Harwood, and colleagues who recently initiated
a research program examining the impact of motivational profiles on athletes’ use of imagery
(Cumming et al., 2002; Harwood, Cumming, & Hall, 2003). Harwood and colleagues concep-
tualized the link between goal orientations and imagery use based on the proactive function
that imagery may serve in the achievement of an athlete’s personal goals (i.e., task and ego) in
sport. They argued that within their search for a sense of competence, athletes with higher lev-
els of task and ego orientation would engage in greater imagery use, with actual image content
differing as a function of goal orientation. In the first of these studies (Cumming et al., 2002),
results of a cluster analysis (N=105) revealed that Canadian provincial level swimmers with
amoderate-task/high-ego goal profile reported being engaged more frequently than other pro-
file clusters in motivational specific imagery (i.e., imaging themselves demonstrating superior
abilities in comparison to others) and motivational general-mastery imagery (i.e., imagining
the development of their abilities).
The second investigation (Harwood et al., 2003), using an elite British youth sport sample
(N=290), provided cross-cultural support, but reinforced the importance of being high in
both orientations in terms of distinguishing the higher frequency of imagery use from the
moderate and low task and ego goal profiles. In addition, Harwood and colleagues (2003) also
found that task and ego goal orientations were related to separate functions of imagery, and
these functions seemed to be consistent with the characteristics typically associated with each
dispositional goal orientation. An ego orientation was positively related to imaging successful
demonstration of skills in comparison to other athletes (motivational specific imagery), whereas
a task orientation was positively related to imaging the development and execution of skills
MOTIVATIONAL PROFILES AND PSYCHOLOGICAL SKILLS 321
(cognitive specific imagery), strategies (cognitive general imagery), and personal mastery, such
as self-confidence, mental toughness, and focus (motivational general-mastery imagery).
Takentogether, these findings demonstrate the important role that dispositional goal orienta-
tions may play in both the behavioral investment and subsequent development of sport-related
imagery skills. However, while these studies have begun to shed light on the conceptual and
applied links between achievement goals and reported imagery use, there remains the need to
further examine how goal orientation profiles regulate athletes’ engagement in a wider range of
psychological skills and strategies. This knowledge is of particular importance to practitioners
working within youth sport populations, who may have the means to shape optimal motiva-
tional profiles in young performers at a time when achievement goals are most malleable and
receptive to socialization or social learning influences (Harwood & Swain, 2002). This is cer-
tainly a worthy avenue of interest if the core social-psychological attributes (i.e., achievement
goals, beliefs, values) of a young athlete facilitate psychological skills use beyond simply the
systematic and external delivery of mental skills training programs.
In their study, Harwood and his colleagues (2003) presented conceptual links between goal
orientations and psychological strategy use. They noted that in an athlete’s personal search
for a sense of competence, individuals with different goal orientations will vary cognitively
on what has to be done to achieve their goals. Based on their findings, Harwood et al. sug-
gested two alternative perspectives that athletes may adopt that may influence their use of
psychological skills. These two perspectives are an investment perspective or a functional
perspective, and only continued research will establish which perspective is most applicable.
From an investment perspective, achievement goal theory would predict that athletes higher
in task orientation, compared with those lower in task orientation, would invest in greater use
of each psychological strategy in order to maximize their opportunities for mastery and ideal
learning and performance states. Such opportunities for developing competence would be less
salient to those high in ego orientation because from an investment perspective, they might
view engaging in structured psychological strategies to devalue more natural demonstrations
of superior ability. This is similar to Nicholls’ (1984) original thinking that, for an ego-involved
athlete, effortful engagement in strategies might serve to reinforce a lack of natural ability or
adequacy, particularly in terms of strategy use in practice.
In elite level sport, however, effort investment is virtually a natural requirement if one seeks
to demonstrate personal improvement (task) and normative superiority (ego). Accounting for
this specific population, it is perhaps pragmatic to appreciate a more functional perspective
and consider that aspiring athletes are likely to engage in those psychological strategies that
will service the achievement of their goals. In this respect, there would be no reason why an
elite athlete high in ego orientation might not invest in goal setting, imagery, positive self-
talk, and relaxation use if those strategies function to facilitate his or her ultimate goal of
superior adequacy compared to others. Taking this line, one would expect no differences to
emerge between psychological skills use in practice and competition at least as a function of
achievement goal orientation.
In drawing the various points in this introduction together, our goal for the present study
wastodevelop an initial understanding of the relationships between the varying motiva-
tional profiles of elite youth sports performers and their reported use of basic psychological
skills/strategies. More specifically, the purpose of this investigation was to examine individ-
ual differences in elite youth athletes’ use of goal setting, relaxation, self-talk, and imagery
skills in practice and competition as a function of their dispositional goal orientations. Based
upon these conceptual notions and the limited empirical support within high level samples
(Cumming et al., 2002; Harwood et al., 2003; Hodge & Petlichkoff, 2000), we decided to
adopt the more functional perspective and hypothesized that performers with higher levels
322 C. HARWOOD ET AL.
of both goal orientations in combination would invest in and report greater psychological
skills/strategy use, regardless of both strategy and context (i.e., practice and competition),
when compared to athletes lower in task and/or ego orientations. We envisaged that the find-
ings would allow us to establish not only the level of validity of such a perspective, but also the
possibility of strategy-specific relationships with achievement goals that are only speculative
at the present time.
METHOD
Participants
A total of 573 athletes (male =174, female =395, unreported =4) elite young athletes
participated in this study with an average age of 17.6 years (range =14–20 years; SD =1.6).
These athletes were recruited over a two-year period from a series of national governing
body high level training camps that were organized through a joint initiative between Nike,
The Institute of Youth Sport and the Youth Sport Trust in Great Britain. Only athletes who
had attained a national level designated by the national governing body (i.e., national sports
association) were selected for these camps. The athletes competed in a broad range of sports,
which included badminton (n=120), basketball (n=15), field hockey (n=49), goal ball
(n=35), lacrosse (n=35), netball (n=15), rugby union (n=101), squash (n=10), soccer
(n=71), track and field (n=43), triathlon (n=15), and volleyball (n=30). A breakdown
of the sample according to sport and gender is reported in Table 1. Parental consent for
participation was granted prior to each training camp through the Youth Sport Trust.
Measures
Perceptions of Success Questionnaire (POSQ)
The POSQ (Roberts, Treasure, & Balague, 1998) is a 12-item measure of task and ego
orientation with six item statements comprising each subscale. Each participant responded to
Tabl e 1
Breakdown of Sample (N=573) According to Sport and Gender
Sport Males (n)Females (n)
Individual Sports
Badminton 66 53
Squash 4 6
Track and Field 23 20
Triathlon 9 5
Team Sports
Basketball 0 15
Goalball 16 19
Field Hockey 23 25
Lacrosse 0 34
Netball 0 15
Rugby Union 19 82
Soccer 14 57
Volleyball 0 64
Total 174 395
Note: Table1excludes the 4 participants who did not report their gender.
MOTIVATIONAL PROFILES AND PSYCHOLOGICAL SKILLS 323
the stem: ‘When playing my sport, I feel most successful when.. . . ’ Items on the Task scale
include ‘I overcome difficulties’ and ‘I perform to the best of my ability,’ while items on the
Ego scale include ‘I win’ and ‘I outperform my opponents.’ Participants responded on 5-point
Likert scales ranging from 1 (strongly disagree)to5(strongly agree). Both the Task and Ego
subscales of the POSQ have demonstrated acceptable internal consistency across a variety of
samples with mean alpha coefficients of 0.81 and 0.82 respectively (Duda & Whitehead, 1998).
Test of Performance Strategies (TOPS)
Thomas, Murphy, and Hardy (1999) developed the TOPS as an assessment of psychological
skills used by athletes in practice and competition contexts. The TOPS is a 64-item question-
naire developed to measure psychological skills used by athletes in various sport situations.
Specifically, within its 16 subscales, it examines activation, relaxation, imagery, goal setting,
self-talk, automaticity, emotional control, and negative thinking/attentional control skills dur-
ing competition and practice settings. Four items represent each subscale, with participants
rating the frequency of each item on a scale anchored by 1 (never)to5(always). Selected ex-
amples of items from the competition subscales include “I am able to relax if I get too nervous
at competition” for Relaxation, and “I talk positively to myself to get the most out of competi-
tion” for Self-talk. Example items from the Practice subscales include “I visualise successful
past performances” for Imagery, and “I set goals to help me use practice time effectively” for
Goal setting.
Within its 16 subscales, the TOPS allows researchers to investigate four of the most com-
monly cited and established psychological strategies (i.e., Goal setting, Imagery, Relaxation,
and Self-talk) for both practice and competition contexts (Fletcher & Hanton, 2001; Hardy
et al. 1996; Vealey, 1994; Weinberg & Gould, 2002; Williams, 2001). Although participants
in the present study completed the questionnaire in its entirety, only the eight subscales as-
sessing the four key psychological skills in practice and competition were included in the data
analysis. This selection is in line with Fletcher and Hanton’s (2001) study that also targeted
the most common and basic psychological strategies/skill use as opposed to all psychological
skill outcomes. Initial analyses of the psychometric properties underpinning the TOPS are
encouraging (see Hardy, Murphy, & Thomas, 1997; Thomas, Murphy, & Hardy, 1999) and
sport psychologists have successfully utilized it not only in applied settings, but also in their
empirical work (e.g., Fletcher & Hanton, 2001). Thomas et al. (1999) reported Cronbach alpha
coefficients ranging from .72 (Imagery use in practice) to .81 (Self-talk in practice) for the 8
subscales used in subsequent analyses.
Procedure
All participants attended one of a series of high performance summer training camps oper-
ated by the Youth Sport Trust for their particular sport. Each participant had been infor med prior
to their camp about the general purpose of the investigation, the assurance of strict confidential-
ity and the importance of providing honest responses. The questionnaires were administered
in pre-planned break sessions under the supervision of one or two research assistants and/or
an informed coach during the course of their training camp experience.
RESULTS
Preliminary Analyses
Reliability Analysis
The internal consistency of the items representing each of the different constructs measured
in the study was determined by calculating Cronbach’s alpha. Adopting a criterion of .70, the
324 C. HARWOOD ET AL.
internal consistency was determined to be acceptable for the task and ego subscales of the
POSQ, with values of 0.87 and 0.81 respectively. For the eight TOPS subscales employed in
this study, coefficients alphas were good. Specifically, these included relaxation in practice
(α=.75), Relaxation in competition (α=.80), Imagery in practice (α=.75), Imagery in
competition (α=.82), Goal setting in practice (α=.79), Goal setting in competition (α=
.79), Self-talk in practice (α=.75), and Self-talk in competition (α=.77).
Descriptive Statistics
Means and standard deviations were calculated for each dependent variable (e.g., subscales
of the POSQ and TOPS), and are presented in Table 2 for the entire sample. In general, these
athletes reported having a higher task (M=4.53, SD =.49) than ego orientation (M=3.60,
SD =.90). In addition, these athletes reported using Goal Setting in competition (M=3.46,
SD =.85) more than any other psychological skill. In comparison, Relaxation in practice
(M=2.15, SD =.75) was the skill least used by these athletes.
Gender and Sport Type Differences
Two separate MANOVAs were conducted to examine whether any gender or sport type dif-
ferences existed in the dependent variables measured in the present study. In the first MANOVA,
gender served as the independent variable, and a significant multivariate effect was found,
Pillai’s Trace, =.14, F(9, 536), =9.45, p<.001, η2=.14. According to Cohen (1988), guide-
lines for interpreting an eta square value (η2)isthat .01 indicates a small effect, .06 indicates a
moderate effect, and .14 indicates a large effect. Therefore, our finding that the η2=.14 indi-
cates that 14% of the total variance in achievement goal orientations is accounted for by gender
differences and this can be classified as a large effect. Using a Bonferroni adjustment (.05/num-
ber of dependent variables) to control for Type 1 errors when making multiple comparisons, an
alpha level of .005 was adopted (Vincent, 1999). Further univariate analysis revealed a signif-
icant effect for the ego subscale, F(1, 544) =42.08, p<.001, η2=.07, which indicated that
males (M=3.94, SD =.78) repor ted having a higher ego orientation than females (M=3.43,
SD =.90). A significant effect was also found for the relaxation in competition subscale,
Tabl e 2
Means and Standard Deviations for the POSQ and TOPS Subscales
Measure MSD
Goal orientation
Task 4.53 .49
Ego 3.60 .90
Psychological skills
Relaxation (practice) 2.15 .75
Relaxation (competition) 3.20 .82
Imagery (practice) 2.91 .90
Imagery (competition) 3.13 .95
Goal setting (practice) 3.24 .84
Goal setting (competition) 3.46 .85
Self-talk (practice) 3.24 .82
Self-talk (competition) 3.28 .89
Note: Goal orientations are assessed on a 5-point Likert scale (1 =strongly disagree,
5=strongly agree), psychological skill usage is assessed on a 5-point Likert scale (1 =never,
5=always).
MOTIVATIONAL PROFILES AND PSYCHOLOGICAL SKILLS 325
F(1, 544) =18.93, p<.001, η2=.03, which indicated that males (M=3.41, SD =.84) re-
ported engaging in this activity more frequently than females (M=3.09, SD =.79). Finally,
a significant effect for the self-talk in competition subscale, F(1, 544) =10.29, p<.001,
η2=.02, which also indicated that males (M=3.49, SD =.91) reported engaging in this
activity more frequently than females (M=3.19, SD =.87).
To examine for differences according to sport type, each sport was first classified as being
either an individual (e.g., track and field) or team sport type (e.g., soccer). It must be noted,
however, that although certain individual sports do contain team elements (e.g., relay events
in track and field, doubles event in badminton), these were still classified as individual for
the present study (Weinberg, Butt, Knight, Burke, & Jackson, 2003). A second MANOVA
was then conducted with sport type serving as the independent variable, and a significant
multivariate effect was found, Pillai’s Trace =.13, F(9, 540) =9.04, p<.001, η2=.13. The
only significant univariate effect found was for the ego subscale, F(1, 548) =23.65, p<.001,
η2=.05, which indicated that individual sport athletes (M=3.88, SD =.76) reported having
a higher ego orientation than team sport athletes (M=3.44, SD =.91).
Cluster Analysis
Rather than using the traditional mean- or medium-split procedures, goal profile groups
were created using a cluster analysis. This procedure has recently gained popularity in the
area of achievement goal orientation as the method used for classifying sample participants
according to their task and ego orientation scores (Cumming & Hall, in press; Cumming et al.,
2002; Harwood et al., 2003; Hodge & Petlichkoff, 2000; Wang & Biddle, 2001). An advantage
to using a cluster analysis over more traditional methods is that it provides the researcher
with the opportunity to examine different solutions, and then select the solution that best fits
the data (Hodge & Petlichkoff, 2000). The main purpose of the cluster analysis is to group
respondents to a questionnaire in such a way that individuals clustered in the same group are
very similar according to some predetermined selection criteria (e.g., task- and ego-orientation
scores). The resulting solution should contain a number of cluster groups that exhibit within-
cluster homogeneity while maximizing between-cluster differences (Hair, Anderson, Tatham,
& Black, 1995).
Following the steps outlined by Hair et al. (1998), all of the dependent measures were first
standardized using z scores (mean of 0 and a standard deviation of 1). Next, the univariate
and multivariate distributions of all variables were inspected for normality and missing data.
Because outliers will distort the solution that emerges from the cluster analysis (Hair et al.,
1998), any case identified as being an outlier (n=74) was deleted from the data, thereby
reducing the overall sample size to 499. Goal profile groups were then generated through
a combination of both hierarchical and non-hierarchical cluster analysis procedures. First, a
hierarchical cluster analysis using a Ward’s method of linkage and a squared Euclidean distance
was employed to identify the number of cluster groups that should be formed by the present
data. The Ward’s method was chosen to be the algorithm used for developing for cluster groups
because it will minimize the within cluster differences and avoid problems with forming long,
snake-like chains found in other methods (Aldenderfer & Blashfield, 1984). In addition, the
squared Euclidean distance was used as a similarity measure because it is the recommended
one to use with a Ward’s method of clustering (Aldenderfer & Blashfield, 1984). Inspection
of the dendogram, which is a graphical representation of the results of the hierarchical cluster,
suggested that a 3- or 4-cluster solution might exist in the data. However, the agglomeration
schedule revealed that a larger increase in the agglomeration coefficient occurred from a
326 C. HARWOOD ET AL.
Tabl e 3
Characteristics of Cluster Membership
Demographics POSQ scores
Gender Sport type Task Ego
Clusters nMale Female Individual Team MSD z MSD z
Higher-task/Moderate-ego 260 79 178 79 181 4.85 .16 .66 3.90 .63 .35
Lower-task/Higher-ego 120 47 73 62 58 4.16 .28 −.76 4.10 .48 .57
Moderate-task/Lower-ego 119 26 93 23 96 4.33 .42 −.40 2.45 .49 −1.28
three-cluster to a two-cluster solution. Therefore, it was concluded that a 3-cluster solution
best fitted the data (Hair et al., 1998).
Validation of the Cluster Solution
A non-hierarchical cluster analysis (e.g., K-means cluster) then followed to validate the
3-cluster solution. Using the cluster centers (e.g., mean score of clustering variables in a
particular cluster) resulting from the hierarchical analysis as the seed points, a K-means cluster
analysis created new cluster groups. A three-cluster solution was again determined to be the
best fit, based on the similarity between the final cluster centers resulting from the K-means
solution to those in the hierarchical analysis, and the interpretability of the solution. The means,
standard deviations, and standardized scores for the three clusters are presented in Table 3. We
then tested the stability of a three-cluster solution by performing a second K-means cluster
analysis with a random selection of 67% of the sample. Results of this second cluster analysis
indicated that over 85% of the sample was correctly re-classified confirming the stability of
the three-cluster solution.
Interpretation of the Cluster Solution
Goal profile groups were interpreted as being higher or lower on the two goal orientations
using a criterion z score of ±.5 (Hodge & Petlichkoff, 2000; Wang & Biddle, 2001). Ac-
cordingly, Cluster 1 contained athletes (n=260) with a Higher-task/Moderate-ego profile,
Cluster 2 contained athletes (n=120) with a Lower-task/Higher-ego profile, and Cluster 3
contained athletes (n=119) with a Moderate-task/Lower-ego profile. A MANOVA was then
calculated to confirm that significant differences existed between the cluster groups on their
Task- and Ego-orientation scores. A significant multivariate effect was found for Goal ori-
entations, Pillai’s Trace =1.06, F(4, 992) =306.63, p<.001, η2=.55, with an observed
power of 100%. Significant univariate effects were found for both Task, F(2, 496) =315.93,
p<.001, η2=.56, and Ego orientation, F(2, 496) =330.60, p<.001, η2=.57. For Task
orientation, post hoc tests revealed that athletes in Cluster 1 (Higher-task) had a significantly
higher task orientation than athletes in Cluster 3 (Moderate-task), who in turn had a signif-
icantly higher task orientation than athletes in Cluster 2 (Lower-task). For Ego orientation,
post hoc tests revealed that athletes in Cluster 2 (Higher-Ego) had a significantly higher Ego
orientation than athletes in Cluster 1 (Moderate-ego), who in turn had a significantly higher
Ego orientation than athletes in Cluster 3 (Lower-ego). In sum, the cluster groups accounted
for over 55% of the variance in achievement goal orientations, which is considered to be a
large effect size (Cohen, 1998). In addition, significant differences between the groups at the
post hoc level on both Task and Ego orientation scores confirmed the appropriateness of the
labels given to the groups during the interpretation stage of the cluster analysis.
MOTIVATIONAL PROFILES AND PSYCHOLOGICAL SKILLS 327
Cluster Membership
Following validation and interpretation of the cluster groups, the next step was to describe
the characteristics of each cluster based on data not included in the cluster procedure (Hair
et al., 1998). Because our preliminary analyses had suggested that gender and sport type
differences existed in the variables used to cluster the data, two separate Chi square tests
were performed to examine whether any such differences existed in cluster membership. A
significant result, χ2(2) =8.43, p=.02, for gender revealed that the percentage of males and
females were not evenly distributed across cluster groups. Upon examination of the clusters,
it appeared that Cluster 3 (Moderate-task/Lower-ego) contained the highest ratio of females
(78.2%) to males (17.1%) whereas Cluster 2 (Lower-task/Higher-ego) contained the lowest
ratio of females (60.8%) to males (39.2%). A significant result was also found for sport
type, χ2(2) =39.11, p<.001, which again indicated that the percentage of individual and
team sport members were not evenly distributed across cluster groups. Cluster 3 contained
a higher ratio of Team Sport (80.7%) to Individual Sport athletes (19.3%). However, Cluster
2 contained a higher ratio of Individual Sport (51.7%) to Team Sport (48.3%) athletes. A
complete breakdown of the cluster membership according to gender and sport type is reported in
Tabl e 3 .
Goal Profile Differences in Psychological Skill Usage
Because our preliminary analyses revealed differences in psychological skill usage accord-
ing to gender, a multivariate analysis of covariance (MANCOVA) was calculated to examine for
differences among the cluster groups after adjusting the means of the TOPS subscales for differ-
ences in gender (Tabachnick & Fidell, 1996). The standardized scores, unstandardized means,
and standard deviations for the three clusters on each of the TOPS subscales are presented
in Table 4. The covariate was significant, Pillai’s Trace =.07, F(8, 484) =4.43, p<.001,
η2=.07. A main effect was also found for the cluster groups, Pillai’s Trace =.10, F(16,
970) =3.29, p<.001, η2=.05. Similar to the preliminary analyses, a Bonferroni adjust-
ment (.05/# of dependent variables) was again employed resulting in an alpha level of .00625
(Vincent, 1999). Univariate analyses of the main effect revealed significant results for imagery
in practice, F(2, 491) =10.90, p<.001, η2=.04, imagery in competition, F(2, 491) =
9.54, p<.001, η2=.04, goal setting in practice, F(2, 491) =11.23, p<.001, η2=.04, goal
Tabl e 4
TOPS Scores for Goal Profile Groups
1. Hi-T/Mod-E 2. Lo-T/Hi-E 3. Mo-T/Lo-E
Psychological skill MSDzMSDz MSDz
Relaxation-P 2.20 .77 .01 2.08 .69 −.01 2.09 .68 −.11
Relaxation-C 3.27 .81 .12 3.21 .73 .00 3.08 .82 −.23
Imagery-P 3.0823 .91 .18 2.75 .85 −.18 2.68 .82 −.21
Imagery-C 3.3023 .94 .16 3.04 .93 −.01 2.86 .90 −.25
Goal Setting-P 3.432.83 .16 3.04 .76 −.29 3.23 .76 −.01
Goal Setting-C 3.6523 .81 .17 3.32 .86 −.15 3.31 .77 −.21
Self-talk-P 3.4323 .74 .22 3.03 .83 −.25 3.13 .84 −.24
Self-talk-C 3.4523 .83 .20 3.19 .85 −.13 3.06 .92 −.30
Note: Superscript23 denotes post-hoc significant differences: 2 =Significantly higher mean in comparison to Cluster 2;
3=Significantly higher mean in comparison to Cluster 3.
328 C. HARWOOD ET AL.
setting in competition, F(2, 491) = 11.21, p<.001, η2=.04, self-talk in practice, F(2, 491) =
12.54, p<.001, η2=.05, and self-talk in competition, F(2, 491) =9.59, p<.001, η2=.04.
Follow-up Tukey HSD post hoc tests were then calculated to determine whether the means
for reported psychological skill usage varied significantly across the three cluster groups. The
results indicated that athletes in Cluster 2 (Lower-task/Higher-ego) and Cluster 3 (Moderate-
task/Lower-ego) did not significantly differentiate from each other on their reported use of the
four psychological skills in practice and competition. The post hoc tests did reveal, however, that
athletes in Cluster 1 (Higher-task/Moderate-ego) reportedly used significantly more imagery
in practice and competition, goal setting in practice and competition, and self-talk in practice
and competition than athletes in Cluster 2 (Lower-task/Higher-ego). In addition, athletes in
Cluster 1 (Higher-task/Moderate-ego) reported using significantly more imagery in practice
and competition, goal setting in competition, and self-talk in practice and competition than
athletes in Cluster 3 (Moderate-task/Lower-ego). Cohen’s D effect sizes were calculated to
assess the practical significant of the results, and these ranged from .30 to .68 in magnitude.
Following Cohen’s (1988) guidelines, where .1 is a small effect, .25 is a medium effect and .40
or higher is a large effect, these values were interpreted as being medium to large effects.
DISCUSSION
The primary purpose of the present study was to examine whether individual differences in
achievement motivation was associated with differential levels of reported psychological skill
use in practice and competition. By acquiring an extensive sample of elite youth athletes from
abroad cross section of sports, however, the investigation also sought to be informative to
sport psychologists in general, particularly those working as practitioners within youth sport
populations.
The mean responses for task and ego orientation were somewhat higher than those sum-
marized by Duda and Whitehead (1998), but perhaps not surprising given the elite nature of
the sample and their heavy investment in achievement oriented activities and experiences. A
cluster analysis procedure was used to classify the athletes in the present study according to
their achievement goal orientations. Three cluster groups emerged from the analysis, and these
groups were labelled according to a z-score criterion of +.5 (Hodge & Petlichkoff, 2000; Wang
& Biddle, 2001). As a result, these groups were labeled as being lower, moderate, or higher
in their respective Task- and Ego-orientation scores. Representing three very different goal
profiles, athletes in Cluster 1 had a Higher-task/Moderate-ego profile, athletes in Cluster 2 had
aLower-task/Higher-ego profile, and athletes in Cluster 3 had a Moderate-task/Lower-ego pro-
file. As pointed out by Hodge and Petlichkoff (2000), however, these labels are simply created
in relation to the z score, and might not correspond with the actual strength of the goal orienta-
tion when viewed as an absolute mean value. When examining the unstandardized means and
standard deviations of each group, it appears that while the range between a lower (M=2.45)
to higher ego (M=4.33) orientation is quite large, the absolute difference between a lower
task (M=4.16) and higher task ( M=4.85) orientation cluster label is only 0.69 of the unit
from 4 (agree)to5(strongly agree). In other words, the “lower” label given to the task score
in Cluster 2 actually represents a high task orientation when considered in absolute sense. It
is important to note, however, that significant differences were found between the task scores
of each cluster group suggesting that the “higher” task score in Cluster 1 was indeed higher
than the “moderate” task score in Cluster 3, which in turn was higher than the “lower” task
score in Cluster 2. Although cluster analysis is preferable to mean or median split techniques
in allowing truly distinct groups to emerge from that specific data set, the typical positive
MOTIVATIONAL PROFILES AND PSYCHOLOGICAL SKILLS 329
skewness of task orientation continues to force researchers to accept some difficult discrim-
inations. Therefore, although this study supports the sensitivity of cluster analysis in aiding
researchers to detect reported behavioral differences in motivational profiles, it is important
to appreciate that cluster labels are only sample-specific, and should not be judgemental of
individuals in objective terms.
The membership of each cluster group was profiled for differences in gender and sport type.
Asignificant chi square test revealed that the distribution of males and females throughout the
cluster groups was consistent with preliminary analyses revealing differences in achievement
goal orientations. Although males and females could be found in each cluster group, there was
a bias towards a larger proportion of the total male sample to be in the moderate to higher
ego-orientation clusters. Similarly, there was a bias towards a larger proportion of the total
individual sport type sample to be in the moderate to higher ego-orientation clusters. The three
cluster groups that emerged in the study also could be distinguished by their use of different
psychological skills. A Higher-task/Moderate-ego profile emerged as the most adaptive pattern
with respect to psychological skill usage, given that athletes with this profile reported greater
use of Imagery, Goal Setting, and Self-talk in practice and competition as compared to athletes
with Moderate-task/Lower-ego profile. In addition, this group distinguished itself from athletes
with a Lower-task/Higher-ego profile by not only engaging in greater Imagery use, Goal Setting
and Self-talk in competition, but also Imagery and Self-talk in practice. When considering the
meaningfulness of these differences, effect sizes were moderate to large in magnitude, and
indicated that 30 to 60% of the variance in psychological skill usage in practice and competition
could be accounted for by the cluster membership. Moreover, these findings are comparable to
those of Harwood et al. (2003) who also reported effect sizes moderate to large in magnitude
for explaining differences in imagery use according to cluster membership among elite young
athletes.
In relation to our hypotheses, the findings generally support the importance of a high task
orientation both from investment and functional perspectives. Even with such small absolute
differences, the higher the task orientation element ‘in situ’ the greater the reported skill use,
except for Relaxation that emerged as a limited strategy across all cluster groups. Further,
the reported use of Goal Setting in practice did not differ between the Higher-task/Moderate-
ego cluster and the Moderate-task/Lower-ego cluster. This perhaps suggests the adaptive role
that task orientation plays in helping the athlete to target achievements in practice, despite
possessing lower motivation in normative terms.
These results however also lend support to the functional role that an ego orientation may
play with respect to reported skill use. Firstly, possessing a moderate ego orientation, alongside
their high task orientation, certainly did not deprive participants in Cluster 1 of employing Goal
Setting, Imagery, and Self-talk to the higher levels in practice and competition. Moreover,
although task orientation appears to be the dominant party in Cluster 1, the lack of difference
in skill use reported between Clusters 2 and 3 gives the reader an indication that highly ego-
oriented athletes will engage in skills to help them achieve their goals. Despite a lower level
of task orientation, the Higher-ego/Lower-task cluster reported using similar levels of all four
skills to the Lower-ego/Moderate-task cluster.
A number of pressing research questions have emerged as a result of these findings. Firstly,
given that task and ego orientations represent orthogonal combinations, how much of a role
does a participant’s task orientation have in encouraging a propensity for psychological skills
use in practice (a self-regulatory characteristic that would be indicative of athletes focused on
skill mastery)? Secondly, how much responsibility lies at the feet of an ego orientation for the
functional application of skills in competition and should practitioners be encouraging ego ori-
entation in moderation? These findings offer interesting and controversial avenues to explore.
330 C. HARWOOD ET AL.
Thirdly, given that researchers have some indication of the nature of imagery use (Cumming
et al., 2002; Harwood et al., 2003), what type of goal setting and self-talk do athletes with
moderate to high task and ego orientations actually use in practice and competition? Although
we can establish scientific differences in reported use per se, the TOPS does not give prac-
titioners a sufficient indication of the exact quality and nature of internal dialogue and goal
setting practices that characterise different motivational types.
Practitioner and Researcher Recommendations
In sum, it is hoped that this study promotes the applied importance of achievement motiva-
tion characteristics in the development of young athletes. The manner in which a young athlete
defines success and failure and approaches achievement situations is linked to the strategies
and skills that he or she may view as fundamental to the attainment of subjective success and
favourable perceptions of competence. Given that the need to learn and practice psycholog-
ical skills has long been acknowledged in the sport psychology literature (Harris & Harris,
1984; Williams, 2001), national governing bodies and practitioners are encouraged to focus
upon education, delivery style and implementation of psychological skills training programs
in elite youth sport. However, they should also focus upon developing the conditions that can
maximize the development of a high task orientation, as well as paying careful attention to the
adaptive regulation of an athlete’s ego orientation. This clearly extends to shaping an optimal
motivational climate around the developing young athlete (Harwood & Swain, 2001) incorpo-
rating the behavioral education of coaches, parents, and those significant others who can exert
a task orienting influence on the individual.
For applied researchers, a weakness of this study is the fact that it provides no more than
a cross-sectional snapshot of the achievement motivation–psychological skills relationship.
Forexample, it was impossible to determine whether reported psychological skill use was
more naturally associated with a particular motivational type, or whether skills that had been
introduced or taught to athletes were embraced, internalized, and adhered to more greatly by
a particular athlete goal profile. Further, given the item content of the TOPS, we could not
clearly identify those participants who have knowledge and experience of systematically using
psychological strategies from those who report using them more naturally. Additionally, the
effect of psychological skill use on the quality of goal orientations admonishes researchers
not to overlook the possibility of a cyclical relationship. As much as individual motivational
differences may impact upon the initial adoption and adherence to psychological skill use, the
application and development of cognitive skills such as Goal Setting and Self-talk may in turn
effect an individuals definition of achievement and criterion view of success or failure.
A deeper understanding of the motivational mechanisms underpinning these relationships
is certainly a challenge worthy of well designed longitudinal or intervention research. Investi-
gating whether athletes with particular motivational profiles actually adhere to and persevere
with newly learned skills that facilitate mastery and/or normative goal attainment would be
one example of the sophisticated research that is needed when answering this question.
CONCLUSION
In closing, as with the achievement context of the classroom, the exam hall, or the business
presentation, young athletes in sport face several behavioral choices that will contribute to their
levels of achievement. They have a choice of whether or not to invest in psychological skills
(training) and strategies; how hard they work at them; and whether they adhere to them and
persist when facing challenges to their perceptions of ability (e.g., during performance slumps).
MOTIVATIONAL PROFILES AND PSYCHOLOGICAL SKILLS 331
These motivated behaviors of choice, effort, adherence, and persistence in psychological skills
are important in elite youth sport and the right motivational conditions may make a difference
to the quality of psychological skills required not merely to survive, but more importantly,
to flourish in high-pressure competitive environments. It is hoped that by being able to draw
upon an extensive sample of high-level young athletes, researchers and practitioners have a
greater awareness not only of reported current psychological skills usage, but also of the role
that achievement motivation and social psychology may play in the behavioral investment of
such skills.
REFERENCES
Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis.Newbury Park, CA: Sage.
Carpenter, P., & Yates, B. (1997). Relationship between achievement goals and the perceived purpose
of soccer for semiprofessional and amateur players. Journal of Sport and Exercise Psychology, 19,
302–312.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Erlbaum.
Cumming, J., & Hall, C. (in press). The influence of goal orientation on self-efficacy in exercise. Journal
of Applied Social Psychology.
Cumming, J., Hall, C., Harwood, C., & Gammage, K. (2002). Motivational orientations and imagery use:
A goal profiling analysis. Journal of Sports Sciences, 20, 127–136.
Dishman, R. (1994). Advances in exercise adherence. Champaign, IL: Human Kinetics.
Duda, J. L. (2001). Achievement goal research in sport: Pushing the boundaries and clarifying some
misunderstandings. In G. C. Roberts (Ed.), Advances in motivational in sport and exercise (pp. 129–
182). Champaign, IL: Human Kinetics.
Duda, J. L., Chi, L., Newton, M. L., Walling, M. D., & Catley, D. (1995). Task and ego orientation and
intrinsic motivation in sport. International Journal of Sport Psychology, 26, 40–63.
Duda, J. L., & Hall, H. K. (2001). Achievement goal theory in sport: Recent extensions and future
directions. In R. Singer, H. Hausenblas, & C. Janelle (Eds.), Handbook of sport psychology (pp. 417–
443). New York: Wiley.
Duda, J. L., Olson, L., & Templin, T. (1991). The relationship of task and ego orientation to sportsmanship
attitudes and the perceived legitimacy of injurious acts. Research Quarterly for Exercise and Sport,
62, 79–87.
Duda, J. L., & Whitehead, J. (1998). Measurement of goal perspectives in the physical domain. In J. L.
Duda (Ed.), Advances in sport and exercise psychology measurement (pp. 21–48). Morgantown,
WV: Fitness Information Technology.
Dunn, J. G. H., & Dunn, J. C. (1999). Goal orientations, perceptions of aggression, and sportspersonship
in elite male youth ice hockey players. The Sport Psychologist, 13, 183–200.
Fletcher, D., & Hanton, S. (2001). The relationship between psychological skills usage and competitive
anxiety responses. Psychology of Sport and Exercise, 2, 89–101.
Fox, K. R., Goudas, M., Biddle, S. J. H., Duda, J. L., & Armstrong, N. (1994). Children’s task and ego
goal profiles in sport. British Journal of Educational Psychology, 64, 253–261.
Hall, H. K., & Kerr, A. W. (1997). Motivational antecedents of precompetitive anxiety in youth sport.
The Sport Psychologist, 11, 24–42.
Hardy, L., Jones, G., & Gould, D. (1996). Understanding psychological preparation in sport: Theory and
practice of elite performers. Chichester, UK: Wiley.
Hardy, L., Murphy,S. M., & Thomas, P.R. (1997). Confir matory factor analysis of the Test of Performance
Strategies on North American athletes. Journal of Applied Sport Psychology, 9(Suppl.), S106.
Harris, D. V., & Harris, B. L. (1984). The athlete’s guide to sport psychology: Mental skills for physical
people. Champaign, IL: Leisure Press.
Harwood, C. G., Cumming, J., & Hall, C. (2003). Imagery use in elite youth sport participants: Reinforcing
the applied significance of achievement goal theory. Research Quarterly for Exercise and Sport, 3,
292–300.
332 C. HARWOOD ET AL.
Harwood, C. G., & Swain, A. B. (2002). The development and activation of achievement goals in tennis:
II. A player, parent and coach intervention. The Sports Psychologist, 16, 111–137.
Hatzigeorgeardis, A., & Biddle, S. (1999). The effects of goal orientation and perceived competence
on cognitive interference during tennis and snooker performance. Journal of Sport Behavior, 22,
479–501.
Hodge, K., & Petlichkoff, L. (2000). Goal profiles in sport motivation: A cluster analysis. Journal of
Sport and Exercise Psychology, 22, 256–272.
Hom, H., Duda, J. L., & Miller, A. (1993). Correlates of goal orientations among young athletes. Pediatric
Exercise Science, 5, 168–176.
Kim, B. J., & Gill, D. L. (1997). A cross-cultural extension of goal perspective theory to Korean youth
sport. Journal of Sport and Exercise Psychology, 19, 142–155.
Newton, M., & Duda, J. L. (1993). The relationship of task and ego orientation to performance-cognitive
content, affect and attribution in bowling. Journal of Sport Behavior, 16, 209–220.
Nicholls, J. G. (1984). Conceptions of ability and achievement motivation. In R. Ames & C. Ames (Eds.),
Research on motivation in education: Vol. 1 (pp. 39–73). New York: Academic Press.
Nicholls, J. G. (1989). The competitive ethos and democratic education. Cambridge, MA: Harvard
University Press.
Ntoumanis, N. (2001). Empirical links between achievement goal theory and self-determination theory
in sport. Journal of Sports Sciences, 19, 397–409.
Ntoumanis, N., & Biddle, S. J. H. (1999). A review of motivational climate in physical activity. Journal
of Sports Sciences, 17, 643–665.
Ntoumanis, N., Biddle, S. J. H., & Haddock, G. (1999). The mediatory role of coping strategies on
the relationship between achievement motivation and affect in sport. Anxiety, Stress & Coping, 12,
299–327.
Roberts, G. C. (2001). Advances in motivation in sport and exercise. Champaign, IL: Human Kinetics.
Roberts, G. C., Treasure, D. C., & Balague, G. (1998). Achievement goals in sport: The development and
validation of the Perceptions of Success Questionnaire. Journal of Sport Sciences, 16, 337–347.
Roberts, G. C., Treasure, D. C., & Kavussanu, M. (1996). Orthogonality of achievement goals and its
relationship to beliefs about success and satisfaction in sport. The Sport Psychologist, 10, 398–408.
Roberts, G. C., Treasure, D. C., & Kavussanu, M. (1997). Motivation in physical activity contexts: An
achievement goal perspectives. In M. L. Maehr & P. R. Pintrich (Eds.), Advances in motivation and
achievement, Vol. 10 (pp. 413–447). Greenwich, CT: JAI Press.
Thomas, P. R., Murphy, S. M., & Hardy, L. (1999). Test of performance strategies: Development and
preliminary validation of comprehensive measure of athletes’ psychological skills. Journal of Sports
Sciences, 17, 697–711.
Vealey, R. S. (1994). Current status and prominent issues in sport psychology interventions. Medicine
and Science in Sports and Exercise, 26, 495–502.
Vincent, W. J. (1999). Statistics in kinesiology (2nd ed.). Champaign, IL: Human Kinetics.
Walling, M. D., & Duda, J. L. (1995). Goals and their association with beliefs about success in and
perceptions of the purpose of physical education. Journal of Teaching and Physical Education, 15,
172–183.
Wang, C. K. J., & Biddle, S. J. H. (2001). Young people’s motivational profiles in physical activity: A
cluster analysis. Journal of Sport and Exercise Psychology, 23, 1–22.
Weinberg, R., Butt, J., Knight, B., Burke, K. L., & Jackson, A. (2003). The relationship between the use
and effectiveness of imagery: An exploratory investigation. Journal of Applied Sport Psychology,
15, 26–40.
Williams, J. M. (2001). Applied sport psychology: Personal growth to peak performance (4th ed.).
Mayfield, CA: Mayfield Publishing Company.
White, S. A. (1998). Adolescent goal profiles, perceptions of the parent-initiated motivational climate,
and competitive trait anxiety. The Sport Psychologist, 12, 16–28.
White, S. A., & Zellner, S. R. (1996). The relationship between goal orientation, beliefs about the
causes of sport success, and trait anxiety among high school, intercollegiate, and recreational sport
participants. The Sport Psychologist, 10, 58–72.
