Similarity revisited: A comparison of similarity—attraction versus dissimilarity—repulsion

Abstract

Byrne, Clore & Worchel's (1966) similarity‐attraction hypothesis and Rosenbaum's (1986a) dissimilarity‐repulsion hypothesis offer very different explanations concerning similarity phenomena (people with similar attitudes become friends, lovers, etc.). Byrne has long maintained that similarity itself leads to greater attraction. Rosenbaum has recently challenged that position and offers evidence that individuals are initially repulsed by dissimilarity, and then seek out similar others. Two studies were designed directly to compare the variability of each hypothesis using a variation of Byrne's phantom other' technique. After filling out an attitude survey, subjects were privately asked to include or exclude similar or dissimilar others (manipulated in surveys supposedly completed by the other subjects in the experiment but actually constructed by the experimenter) for a group activity. Both the inclusion/exclusion decision and the order in which the decisions were made comprised the measure of attraction. Results of both studies indicate that subjects tend initially to include similar others and then exclude dissimilar others, thus supporting a similarity‐attraction hypothesis. Instructions concerning the group activity were also varied across studies, without affecting the pattern of results. The implications of the method and findings for similarity‐attraction research are discussed.