This study examined the effects that varying subjects’ response sets on the Test of Attentional and Interpersonal Style (TAIS) had on predictive validity. Fifteen elite divers took the TAIS under two response conditions. Initially they answered the items without a specific set or comparison group in mind. The second time they were instructed to compare themselves with other elite divers. It was hypothesized that by telling divers who to compare themselves to and by providing them with a situation-specific response set, this would increase the correlations between performance and TAIS scores. The implications and consequences of response set manipulations are discussed.