Group influences on success expectancies regarding social influence attempts

This experiment concerned the generality of the discussion-induced shift toward pessimism found in some previous studies but not found in others. Male university students (N = 95) were presented with one of two fictitious situations in which a protagonist wanted to gain compliance from another person (an employee wants a promotion; a parent wants his child to study more). For each of 16 compliance-gaining behaviors (e.g., promise, warning, invocation of need) the subjects had to indicate, on a 7-point scale, the likelihood that the protagonist would be successful ("objective response mode H ) or they had to indicate how well they personally could perform the respective influencing behavior ("subjective response mode H). All subjects first made a success-expectancy estimate. In the group condition (24 triads) participants discussed the judgmental issue and then again made a private estimate on each of the 16 items in turn. For the subjects in the control condition (N = 23) group discussion was replaced by further individual thought and note-taking. Discussion-induced shifts toward pessimism were found in the group conditions (p < .01 and .10 in the subjective and objective conditions). Since there was a (modest) overall initial tendency toward the pessimistic pole, this shift can in part be considered as group polarization, thus attesting to the generality of the group polarization phenomenon. (Secondary analyses bearing on the explanation of this polarization are presented.) In addition, secondary analyses show that pessimistic shift also occurred in those discussion cases where there was no initial leaning toward pessimism. The theoretical explanation of these shifts is discussed