Perceived aggression: A re-evaluation of the bandura modeling paradigm

Two experiments were performed to examine whether naive observers would label as aggressive the destructive behavior of a child following the latter's observation of an adult model. Experiment I provided descriptions of a standard modeling paradigm to female introductory psychology students. One of four sequences was presented in written form: the model and child were both constructive or destructive, the model was destructive and the child was constructive, or vice versa. The results showed that imitative destructive behavior by a child was not perceived as aggressive or bad, whereas the same pattern of play was considered aggressive and bad in the absence of similar behavior by a model (p <.03). Experiment II replicated the design and results of the first study and accrued information regarding the legitimizing function of the model. Observers attributed responsibility for imitative destructive behavior to the model and believed the child considered the behavior as appropriate and likely to produce rewards. The present results question the generalizability of prior findings and the adequacy of the modeling paradigm for the study of aggression.