Observation and Performance of Aggression

Abstract

Although the frustration-aggression hypothesis has focused typically on aggressive behavior, the primary purpose of the present experiment was to investigate the effect of frustration on observation of either aggressive or nonaggressive models when an individual can choose which model to observe. A total of 103 second- and third-grade males participated in a factorial experiment manipulating frustration and modeling cues. The results disclosed that an aggressive model elicited more imitative aggression than simultaneously presented aggressive and cooperative models (multiple models). Both of these conditions, in turn, produced more aggression than a cooperative-model condition or a no-model condition. Within the multiple-models condition, which is characteristic of many naturalistic settings, frustration produced greater observation of the aggressive model, greater recall of aggressive acts, and stronger attacks against another person. These data were interpreted in terms of an expanded frustration-aggression hypothesis: In social situations in which observers have a choice of heterogeneous models to observe, prior frustration increases the probability that an individual will observe, remember, and hence perform more acts of violence.