تجزیه و تحلیل خوشه ای از اضطراب قبل رقابتی: ارتباط با کمال گرایی و اضطراب خصلتی

The purpose of this study was to examine whether athletes of different sports clustered in meaningful ways, based upon their intensity, direction and frequency of cognitive and somatic anxiety using hierarchical cluster analysis, and to compare the subgroups of athletes on trait anxiety, perfectionism and self-confidence. One hundred and sixty six male and female athletes completed the Sport-Multidimensional Perfectionism Scale, the Sport Anxiety Scale and the Competitive State Anxiety Inventory-2 Revised including direction and frequency scales. Results revealed five-clusters labelled “anxious facilitators”, “anxious debilitators”, “low anxious facilitators”, “low anxious debilitators” and “ruminator debilitators”. Clusters differed significantly on concentration disruptions, trait somatic anxiety, worry, concern over mistakes, perceived parental pressure, and intensity and frequency of self-confidence. The importance of considering all dimensions of anxiety simultaneously when examining the functional nature of the construct and the five-clusters are discussed.

The stressful nature of sport and the competitive environment place many demands on athletes (Jones, 1995). Accordingly, an area of research in sport psychology has been directed towards the emotional responses to such stressors and in particular the study of competitive anxiety (Martens et al., 1990 and Woodman and Hardy, 2001). A considerable amount of research has studied levels (i.e. intensity) of competitive anxiety, investigating the antecedents, the temporal patterning of the subcomponents (i.e. cognitive and somatic) and the anxiety performance relationship (Jones, 1995 and Woodman and Hardy, 2001). Although research on intensity has contributed to our understanding of competitive anxiety, additional dimensions have been proposed, such as directional perceptions (i.e. interpretation of the symptoms associated with competitive anxiety as being facilitative or debilitative towards performance; Jones & Swain, 1992). Support for the differentiation between intensity and direction of competitive anxiety has been provided in many studies examining a range of dispositional and situational variables (Jones, 1995 and Woodman and Hardy, 2001). In particular, elite performers interpret their anxiety symptoms as more facilitative than their non-elite counterparts (Jones, Hanton, & Swain, 1994). In addition to examining anxiety intensity and direction, several researchers (Cerin et al., 2000 and Hanton et al., 2004) suggest including a frequency component to the anxiety response (i.e. the amount of time spent attending to anxiety symptoms experienced concerning competition; Swain & Jones, 1993). Although the intensity of the anxiety symptoms may not change, a state in which worries (i.e. intensity) are occurring 5% of the time is very different from one in which they are occurring 90% of the time (Jones, 1995). The distinction between intensity and frequency has been provided in studies showing that the frequency of cognitive and somatic anxiety was more sensitive to changes over time than the intensity in the week leading up to competition (Hanton et al., 2004 and Thomas et al., 2004). Despite their respective strengths and weaknesses, previous studies investigated the bivariate relationships between anxiety responses and some other variables. Such an approach has neglected the multivariate nature of anxiety. It would be fruitful to provide new insights in the relationships between the three dimensions of anxiety by identifying subgroups of performers with unique or unexpected anxiety response patterns. The possibility of identifying different profiles of athletes based on the three dimensions of anxiety may provide a deeper understanding of athlete anxiety leading to implementation of more effective interventions to manage precompetitive anxiety. This is why the first aim of this study was to identify meaningful clusters of athletes based on their precompetitive intensity, direction and frequency of cognitive and somatic anxiety. A method that has been widely used to classify participants is cluster analysis ( Hair, Anderson, Tatham, & Black, 1998). Cluster analysis is a statistical process that uses multivariate techniques to group participants based on their characteristics (i.e. anxiety scores) by maximizing both the homogeneity of cases within a group and the heterogeneity between the clusters ( Aldenderfer et al., 1984 and Hair et al., 1998). In cluster analysis, using another variable than the one used to create the groups has proven an efficient technique for validating this type of analysis ( Aldenderfer et al., 1984). Thus, the second purpose of this study was to investigate whether the subgroups of athletes differ on trait anxiety, perfectionism and self-confidence. According to the interactional model of competitive stress, state competitive anxiety is thought to be determined by personal (trait anxiety, perfectionism) and situational factors (state self-confidence) through the process of cognitive appraisal (Cerin et al., 2000). Competitive trait anxiety is a particularly crucial factor to investigate as it directly affects the perception of threat, subsequently mediating the level of state anxiety (Smith, Smoll, & Shultz, 1990). Contrary to the suggestion of Hardy et al., 1996 and Hanton et al., 2004 observed no differences on anxiety interpretation as a function of trait worry and trait somatic anxiety groups (high vs. low). They showed that low trait concentration disruption performers perceived their anxiety symptoms as more facilitating than their high trait counterparts. Further studies are needed to determine if the results of Hanton, Mellalieu and Hall were sample specific (male soccer players) or general within other sports and gender. There is currently a lack of research examining the relationship between perfectionism and anxiety in sport (Hall, Kerr, & Matthews, 1998). Perfectionism has been conceptualized as a domain-specific stable multidimensional construct and is comprised of four dimensions in sport (Dunn et al., 2006). Concern over mistakes is the tendency to become overly concerned about personal mistakes and to view mistakes in sport as unacceptable. Personal standards correspond to the tendency to set high standards of personal performance in sport. Perceived parental pressure and perceived coach pressure reflect the tendency to perceive respectively parents and coaches as being overly demanding, and critical with respect to the athletes’ performance efforts. Previous research has shown several links between perfectionism and precompetitive anxiety intensity in sport ( Hall et al., 1998 and Koivula et al., 2002). However, perfectionism was conceptualized and measured as a global personality construct and not as a sport-specific construct and the anxiety dimensions of direction and frequency were not included in the perfectionism anxiety relationship. Self-confidence is one of the most important variables related to performance in sport (Hardy et al., 1996). Interestingly, Jones and his colleagues (Jones, 1995 and Jones et al., 1994) showed that anxiety direction correlated more strongly with self-confidence than anxiety intensity. Using a qualitative methodology, Hanton et al. (2004) showed that increases in competitive anxiety intensity were perceived as debilitating to performance in the absence of self-confidence and as facilitating to performance under conditions of high self-confidence. This finding does allow the proposition that self-confidence may, in some way, protect against potential debilitative anxiety effects (Hardy et al., 1996 and Hanton et al., 2004). In summarizing, the aim of our study was twofold: (1) firstly, we examined whether athletes clustered in meaningful ways based upon their intensity, direction and frequency of cognitive and somatic anxiety; (2) secondly, we investigated whether these subgroups of athletes differ in trait anxiety, perfectionism and self-confidence. We also examined whether these subgroups of athletes were confounded by individual difference variables such as age, type of sport and sex (Woodman & Hardy, 2001). Given the exploratory nature of this investigation, no specific hypotheses were advanced.