Implied evaluation and the similarity-attraction effect.

Abstract

In Experiment I, manipulations of similarity (high or low), implied evaluation (positive, negative, or none), sex of other, and physical attractiveness of other (high or low) produced main effects for similarity, implied evaluation, and physical attractiveness as well as a similarity by implied-evalu ation interaction. The similarity-dissimilarity effect was greater in the no-evaluation condition than in the positive- and negative-evaluation conditions. It was further found that in the no-evaluation condition, subjects perceived that the similar other would like them more than the dissimilar other. These results were interpreted as supporting three theoretical models and as ruling out a ceiling effect interpretation. In Experiment II, there were manipulations of similarity (five levels), sex of subject, and sex of other. A division of subjects into those who perceived the other would either like or dislike them versus those subjects who perceived more neutral implied evaluations resulted in an interaction with similarity. Subjects who perceived the evaluative implications of similarity showed a more marked similarity-attraction effect. A parallel analysis for subjects who differentially perceived the anticipated rewards of future interaction did not produce a significant interaction with similarity. The overall results were discussed in terms of their bearing on the three theoretical models, and the data were interpreted as giving the most support to a model implying that similarity causes liking, liking causes implied evaluation, and implied evaluation causes further liking. This model was the only one that predicted that the effect of similarity upon implied evaluation would become nonsignificant when liking was used as