Trait Encoding in Behavior Identification and Dispositional Inference

Abstract

Subjects were presented with behavioral vignettes that either suggested traits in the actors or suggested that the behavior was caused by the situation. A word-fragment completion procedure was used to determine whether suggested traits were activated during processing of the vignettes. One issue of interest was whether suggested traits were activated both in the context of behavior identification and in the context of inferences about actors. To distinguish between these two forms of activation, fragments of implied traits were presented either with or without the relevant actor label. It was reasoned that if a trait is linked to an actor, the actor cue should facilitate trait completion. Subjects' orienting task was also manipulated: half were asked to form impressions of the actors and half to allocate causality to the actor or to the situation. Trait priming proved to be common for subjects instructed to form impressions of actors, even in conditions that did not elicit dispositiona' inferences. Actor cues were effective in aiding trait completion for these subjects, but only when covariation information implied a person attribution. There was no evidence of trait priming following the causal allocation task In the context of Trope's distinction between behavior identification and dispositional inference, it is argued that these results show that trait concepts are involved in the identification of behavior independently from their involvement in dispositonal inferences.