A behavioural analysis of "choking" in self-paced skills

Abstract

This thesis is about "choking" in self-paced skills. Choking refers to ''the occurrence of inferior performance despite striving and incentives for superior performance" (Baumeister and Steinhilber, 1986, p. 361). Self-paced skills are skills in which performance is initiated by the athlete. This research set out to investigate the cause of choking in self-paced skills within the theoretical framework of behaviour analysis. The main focus of the research relates to the distinction between behaviour under the control of verbal antecedents (rule-governed behaviour) and behaviour that is shaped by its consequences (contingency-shaped behaviour). It was originally hypothesised that the insensitivity of rule-governed behaviour to changes in the contingencies of reinforcement could he beneficial in situations where these changes led to greater performance pressure. Specifically, it was predicted that performance under the control of verbal antecedents would be less susceptible to choking. In the first experiment, no support was found for the hypothesis and, furthermore, rule-governed performance appeared to be inferior to contingency-shaped performance in the early stages of acquisition. In light of these results, and after a detailed examination of the behaviour analysis distinction between these two forms of behaviour, evidence was presented which suggested that verbal control of the topography, or form, of behaviour would be likely to disrupt performance in self-paced skills. In subsequent experiments, it was found that using simple verbal cues was an effective means of preventing choking under pressure. It was hypothesised that the function of these cues was in preventing reinvestment of too many technical instructions in the moments before performance initiation. The assumptions upon which the reinvestment theory of choking is based were also examined with results providing general support for the theory but also suggesting that it needs to be refined to account for verbal antecedents that do not disrupt performance.