Abstract
Two studies were conducted to examine the behavior effects of discrepant role playing, and to test the proposition that the production of aversive consequences is necessary for the dissonance-predicted, inverse relationship between amount of inducement and subsequent evaluative change to be obtained. In the first experiment, subjects were enticed for either a low or high monetary incentive to advise three confederates to engage in a dull task for a considerable amount of time by asserting that the task is enjoyable. As expected, subjects evidenced the dissonance-predicted effect only when they were successful in convincing their colleagues. Subjects were also found to engage in the dull task for a longer time after performing a successful act of discrepant role playing. A second study was designed to assess the relative importance of success and aversive consequences in producing the dissonance-predicted effect. All subjects learned that they were successful in convincing their colleagues of the altitude-discrepant position, but for only half did the successful speech produce aversive consequences. The results clearly demonstrate that aversive consequences rather than success is the key determinant of dissonance-predicted evaluative changes. The data further indicate that discrepant role playing is not effective in producing attitude-related behavior change. The results are discussed in terms of dissonance theory, self-perception theory, and response contagion theory.