Sharing reinforcement contingencies with a model: A social-learning analysis of similarity effects in imitation research

Tested the hypothesis that the degree to which an observer and model are subject to similar reinforcement contingencies is a potent determinant of imitation. In Exp I, 128 3rd graders participated in a premodeling phase in which the correspondence between the consequences received by them and an adult for performing similar and diverse responses was manipulated. Ss then watched the model perform new responses that incurred reward, punishment, or no reinforcing consequence. On a test for imitation of these responses, Ss who had experienced reinforcing contingencies similar to the model's imitated more of the model's responses that had incurred either reward or no consequence, but fewer of the responses that had incurred punishment, than Ss whose contingencies had been uncorrelated with, or opposite to, the model's. Exp II (48 different 3rd graders) indicated that the tendency of the former group to imitate responses incurring no consequence persisted when Ss could not infer the lack of consequence to imply reward. It is suggested that in naturalistic settings people associate similarity to a model with sharing reinforcement contingencies with the model, and this process might mediate effects of similarity on imitation. (22 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).