Developing and Testing an Integrated Model of Choking in Sport

In general, choking has been defined as the occurrence of sub-optimal performance under pressure (Baumeister and Showers, 1986). The most widely accepted theory of choking is what I have labeled the 'automatic execution model' (Baumeister, 1984), the basic premise being that choking occurs due to the inhibition of well-learned or automatic skills. The present dissertation was designed to test the automatic execution model in a sport context. Beyond this, I was interested in broadening the research paradigm by concentrating on choking as a process. The five inter-related studies contained in this dissertation tested predictors of choking, perceptions of pressure, coping processes, the automatic execution model, and gender differences. The general aim of this dissertation was to produce an integrated model of choking in sport. The five studies used to examine choking had a total of 89 competitive basketball players as participants (M = 20.01 years old, SD = 2.12). Sixty-four participants took part in studies, 1, 2, 3 and 5. The remaining 25 participants took part in study 4 only. All participants completed a series of basketball free throws in a low-pressure (LP) condition and a high-pressure (HP) condition. Pressure was manipulated by videotaping performance, using an audience, and offering financial incentives. In Study 1, dispositional self-consciousness (S-C) and trait anxiety (A-trait) were tested as potential predictors of choking. Participants (N = 66) completed the Self-Consciousness Scale (SCS; Fenigstein, Scheler, & Buss, 1975) and the Sport Anxiety Scale (SAS; Smith, Smoll, & Schutz, 1990). A correlational analysis and a hierarchical multiple regression analysis supported the hypothesis that athletes high in S-C and somatic A-trait were susceptible to choking under pressure. In Study 2, the effects of manipulated pressure were examined for possible changes in perceived state anxiety (A-state) and subsequently performance. Participants (N = 64) completed the modified Competitive State Anxiety Inventory-2 (CSAI-2; Jones & Swain, 1992) prior to performing in the LP condition and the HP condition. A multivariate analysis of variance (MANOVA) and a correlational analysis showed that participants experienced increased intensity of somatic and cognitive A-state in the HP condition, but the direction (facilitative/debilitative) of somatic A-state and cognitive A-state did not change significantly. Correlation analysis between A-state and performance under pressure was also not significant. In addition, a one-way analysis of covariance (ANCOVA) showed an absence of significant differences in performance between participants who had a negative A-state and participants who had a positive A-state. Study 3 was designed to examine the relationship between coping styles and choking susceptibility. Participants (N = 64) completed the Coping Style Inventory for Athletes (CSIA; Anshel & Kaissidis, 1997). As stated above participants also completed the CSAI-2 prior to performing in the LP and HP conditions. Correlation and hierarchical regression analyses supported the hypothesis that approach coping strategies would result in heightened A-state and performance decrements under pressure. In Study 4, the automatic execution model was tested by investigating possible differences in the occurrence of choking as a function of task characteristics. Participants (N = 25) completed a running task (effort-dominant task) and a free throw shooting task (skill-dominant task) in a LP condition and a HP condition. As hypothesized, a series of t tests showed that the HP condition resulted in improved performance for the running task but decrements in performance for the free-throw shooting task. Study 5 was used to investigate potential gender differences in choking susceptibility. Participants (male = 46; female = 18) had already completed the SCS, SAS, CSIA, and CSAI-2. A one-way MANCOVA showed there were no gender differences in somatic and cognitive A-state changes from the LP condition to the HP condition. Furthermore, a one-way ANCOVA showed no gender differences in performance from the LP condition to the HP condition. In addition, a correlational analysis showed that S-C was more likely to affect performance for females and A-trait was more likely to affect performance for males. Based on the results of these five studies, previous choking research, and the theoretical choking framework, a proposed choking process model is presented and discussed. This proposed choking process model includes susceptibility factors, perception of pressure, coping process, task characteristics, and skill levels