Abstract
Until recently, the traditional Inverted-U hypothesis had been the primary model used by sport psychologists to describe the arousal-performance relationship. However, many sport psychology researchers have challenged this relationship, and the current trend is a shift toward a more "multidimensional" view of arousal-anxiety and its effects on performance. In the current study, 104 college-age participants performed a simple response time task while riding a bicycle ergometer. Participants were randomly assigned to one of eight arousal groups (between 20 and 90% of heart rate reserve) and were told they were competing for a cash prize. Prior to the task, the Competitive State Anxiety Inventory-2 and Sport Anxiety Scale (SAS) were administered to assess the influence of cognitive and somatic anxiety. As hypothesized, regression analysis revealed a significant quadratic trend for arousal and reaction time. This accounted for 13.2% of the variance, F change (1, 101) = 15.10, p < .001, in performance beyond that accounted for by the nonsignificant linear trend. As predicted by the Inverted-U hypothesis, optimal performance on the simple task was seen at 60 and 70% of maximum arousal. Furthermore, for the simple task used in this study, only somatic anxiety as measured by the SAS accounted for significant variance in performance beyond that accounted for by arousal alone. These findings support predictions of the Inverted-U hypothesis and raise doubts about the utility theories that rely on differentiation of cognitive and somatic anxiety to predict performance on simple tasks that are not cognitively loaded.