THE PERCEPTION OF TRAIT DIMENSIONS AS A FUNCTION OF SOCIAL INTERACTION

Abstract

The present study was concerned with the process of forming impressions in the context of an actual, direct relationship. Two major issues were examined: (1)Do subjects generate traits containing specific interpersonal information; and (2)Is the content of the traits that are attributed tied to a given interactional context, so that traits are differentially attributed as a function of the specific interaction. Three interactional dimensions were proposed, the sentiment, competency, and power interactional dimensions, defined by Hershkowitz (1980). These are hypothesized to correspond to three types of trait dimensions, those of sentiment, competency, and power. Subjects generated traits in response to a stimulus person of their own choice, with whom they were engaged in a particular kind of relationship, comprised mainly of one of the interactional dimensions. These traits were then rated by another group of subjects, regarding the extent to which they reflect the above trait dimensions. Results dealing with the trait ratings indicated that subjects agreed on the extent to which each trait dimension was reflected in the traits, and that the trait dimensions did not seem to differ with regard to the ease with which they could be rated. Furthermore, most traits conveyed information mainly relevant to one trait dimension. Findings related to the trait generation task showed that trait information is differentially utilized, so that it reflects information specific to the particular relationship involved. Sentiment traits were attributed in each relationship, whereas power traits were selectively used. Results are discussed in terms of the social nature of all types of relationships, the role of liking in maintaining the interaction, and the constraints on making power attributions in our society. The meanings conveyed by trait labels and their probable definition via social convention is also discussed. Implications for the functions of trait attribution in terms of the stability and predictability of behavior are proposed, along with suggestions for future research.