Fifty-six male and 61 female university students participated in a 2 × 2 × 2 design with task correctness-incorrectness, group agreement-disagreement and sex of subject constituting the classification factors. A modified Crutchfield apparatus served to manipulate a subject's prior experience of correctness and agreement, and to measure subsequent conformity.

Correctness and group agreement were seen to interact to produce varying degrees of perceived competence relative to a simulated group. Relative competence mediated conformity such that (a) on the basis of both experimentally manipulated competence and perceived competence, subjects who were less competent than the group manifested more conformity than subjects who were more competent than the group. This finding replicates the Ettinger et al. (1971) study; (b) subjects who perceived themselves and the group as equally competent conformed more than subjects who perceived themselves as either more competent or less competent than the group. This result was explained in terms of reciprocity; (c) no sex differences were found and possible explanations for this result were discussed; (d) no conformity differences between suspicious and unsuspicious subjects suggested the possibility that suspicious subjects were role playing.