Thinking about choking? Attentional processes and paradoxical performance.

When pressure to perform is increased, individuals commonly perform worse than if there were no pressure ("choking under pressure"). Two mechanisms have been proposed to account for this effect--distraction (cognitive load), wherein pressure distracts attention from the task, and self-focus, wherein attention shifts inward interfering with performance. To distinguish between these two competing explanations, the current experiment manipulated pressure by offering performance-contingent rewards. For half the participants, cognitive load was increased by requiring participants to count backward from 100. Additionally, adaptation to self-awareness was manipulated by videotaping half the participants during practice trials. Results show that pressure caused choking when participants were not distracted and had not been adapted to self-awareness. This effect was attenuated when cognitive load was increased or when self-awareness adaptation had occurred. These results support self-focus mediated misregulation as the mechanism for choking and disconfirm the distraction hypothesis. The paradoxical performance effect, or "choking under pressure," has been defined as the occurrence of suboptimal performance despite incentives for optimal performance (Baumeister, 1984). Although several previous studies have demonstrated the detrimental effects of performance pressure on a wide range of tasks, few have sought to examine the underlying causes for these effects, and none have allowed the development of an adequate understanding of the cognitive processes responsible for choking. The current study examined the effects of performance pressure on task performance in an attempt to distinguish between the two dominant mechanisms proposed to account for the occurrence of choking--the distraction and self-focus models. In addition, the efficacy of interventions designed to alleviate choking was examined.

One of the earliest attempts to understand the relationship between arousal and performance was HullSpence drive theory, which received considerable empirical support when applied to simple monoresponse learning tasks. This theory was soon extended (Spence & Spence, 1966) to include complex performance behaviors. These dominant response models (e.g., Broen & Storms, 1961; Landers, 1980; Spence & Spence, 1966) propose that increases in arousal (or drive) increase the probability that the dominant response will occur. If the task is well learned, the dominant response is the correct or successful behavior. Thus, increasing pressure to perform well, thereby increasing drive, will elevate the level of performance. If the task is poorly learned, however, the dominant response is failure or error, and a decrease in performance will result from enhanced drive.

This version of drive theory fails to account adequately for all instances of choking. In addition to the abundant anecdotal evidence (missing an easy putt on the 18th green when a large wager hangs on the outcome), there is extensive empirical evidence contradicting this formulation of drive theory. In an archival study by Baumeister and Steinhilber (1984), it was shown that increases in drive often resulted in a decrease in the dominant response. These researchers found that in the championship series in professional baseball and basketball, the home team is significantly more likely to lose the decisive game than it is to lose earlier home games in the series. At this level of competition, the dominant response is successful performance; thus, these findings do not support drive theory. These researchers also found a similar pattern of results for individual performances. Performance effects seemingly inconsistent with a dominant response facilitation explanation have been found in other tasks as well, including attentional tasks (McNamara & Fisch, 1964), maze completion tasks (Seta, Paulus, & Risner, 1977), and test anxiety research, in which the performance of cognitive tasks has been shown to suffer as a result of increased drive (Paul & Eriksen, 1964; Wine, 1971).

Drive theory, then, does not seem to provide an adequate explanation for the relationship between pressure and performance.

Attention and Overmotivation

A modified drive theory, proposed to account for paradoxical performance, has been termed the inverted-model (Broen & Storms, 1961; Duffy, 1962; Easterbrook, 1959). This model asserts that increasing drive will increase performance up to a certain optimal point, but after this point is reached, increases in drive will lead to performance decrements. To account for this phenomenon, Easterbrook (1959) postulated that the range of cues used in a task is reduced by increases in drive. At low to moderate levels of drive, this range reduction enables the actor to screen out task-irrelevant cues, resulting in increased task concentration and performance. However, at high levels of drive, the range of attentional focus becomes overly narrow, screening out task-relevant cues as well, thus reducing performance quality. This model is a departure from pure drive theory and may be more appropriately considered to fall under the rubric of attentional theories.

Attentional theories may offer the most promise for understanding the cognitive processes underlying poor performance under heightened incentive, and thus for developing interventions to alleviate choking. They are based on the assumption that successful task performance relies on the ability to attend to task-relevant cues, processes, and behaviors while ignoring task-irrelevant information. Interference with these attentional processes results in poorer performance. Two predominant variants of attentional models have been applied to the choking phenomenon--the distraction and the self-focus models. Distraction models hold that some stimuli cause a diversion of attention from task- relevant information, thereby reducing performance (Baumeister & Showers, 1986). Self- focus models propose that through a process of heightened self- awareness, the performance response is inhibited, either through a reduction in response speed (Bond, 1982) or through an interference in the automaticity of the response (Kimble & Perlmuter, 1970; Zimring, 1983). Because both the distraction and the self- focus models of choking can account for paradoxical performance effects, the dilemma becomes one of developing a way to distinguish between them. The implications of each model must first be explored to find a point of divergence on which different empirical outcomes can be predicted.

Self-focused attention and choking. Self-focus models of choking are anchored in the tenets of …