Evidence for a General Construct of Aggression

Abstract

Three analyses of published research were undertaken to assess whether diverse laboratory response measures that are intended to measure aggression reflect a common underlying construct. It was found that (a) alternative measures of physical aggression directed by the same subjects against the same target tend to intercorrelate positively within studies, (b) across studies, the correlations between effect-size estimates of physical and written aggression emitted by the same subjects are positive, and (c) physical and written aggressive responses are similarly influenced by theoretically relevant antecedent factors (e.g., personal attack and frustration). The consistent overall pattern of results supports the notion that aggression, defined as intent to harm, is a viable construct that possesses some degree of generality.