Repulsion or Attraction? Group Membership and Assumed Attitude Similarity Fang Fang Chen and Douglas T. Kenrick Arizona State University Three studies investigated group membership effects on similarity–attraction and dissimilarity–repulsion. Membership in an in-group versus out-group was expected to create initially different levels of assumed attitude similarity. In 3 studies, ratings made after participants learned about the target’s attitudes were compared with initial attraction based only on knowing target’s group membership. Group membership was based on political affiliation in Study 1 and on sexual orientation in Study 2. Study 3 crossed political affiliation with target’s obnoxiousness. Attitude dissimilarity produced stronger repulsion effects for in-group than for out-group members in all studies. Attitude similarity produced greater increments in attraction for political out-group members but not for targets with a stigmatic sexual orientation or personality characteristic. What is the link between attitude similarity and attraction, and how is that link affected by group membership? There is strong support for the general statement that we like those who are like us (Byrne, 1961; Newcomb, 1956; Tesser, 1993). And there is evi- dence that we like those with whom we share group memberships (Brewer, 1979; Mullen, Brown, & Smith, 1992; Rokeach, 1960; Tajfel & Turner, 1986). But what happens when a member of an in-group is discovered to disagree with one’s opinions or when a member of an out-group is found to share one’s beliefs? Consider the case of Margaret Thatcher, the “Iron Lady” of the British Conservative party. Thatcher developed a reputation for intoler- ance of other Conservatives who disagreed with her. On the other hand, although Thatcher’s right-wing policies included strong op- position to the Soviet Union, she came to regard Mikhail Gor- bachev, the leader of the Soviets, as a friend (Thatcher, 1993). In this article, we investigate the possibility that the influence of similarity and dissimilarity on attraction depends on whether the person being evaluated is regarded as a member of one’s in-group or as an out-group member. Two Hypotheses Regarding Attitude Similarity and Attraction Two somewhat different views have been advanced with regard to the role of attitude similarity– dissimilarity in interper- sonal attraction: the similarity–attraction hypothesis and the dissimilarity–repulsion hypothesis. The similarity–attraction hy- pothesis proposes that attitude similarity promotes interpersonal attraction (Byrne, 1971; Newcomb, 1956; Singh, 1974). According to this view, attraction is a positive function of the extent to which two individuals share beliefs about important topics. A substantial amount of evidence has been found to support this hypothesis (e.g., Byrne, 1961, 1971; Byrne & Nelson, 1965; Clore & Baldridge, 1968; Condon & Crano, 1988; Singh, 1974; Tesser, 1993). For example, Byrne and Nelson (1965) asked participants to rate a stranger and varied the stranger’s proportion of similar attitudes at four levels (1.00, .67, .50, and .33). They found a linear relation- ship between similar attitudes and attraction. On the other hand, the dissimilarity–repulsion hypothesis (Rosenbaum, 1986a) argues that the relationship between attitude similarity and attraction is instead a relationship between attitude dissimilarity and repulsion. It is not so much that attitude similarity leads to liking as that dissimilarity leads to disliking. According to Rosenbaum (1986a), most studies of attitude similarity have not included an adequate baseline or no-attitude-information control condition, which is necessary for one to determine whether simi- larity enhances attraction or dissimilarity decreases attraction. And the studies that did include such a condition did not statistically test the difference between the control and the similarity condi- tions. Consequently, those studies failed to provide the evidence to support the assumption that attitude similarity leads to attraction; attraction could instead be explained in terms of dislike for those whose attitudes are dissimilar. To support this dissimilarity–repulsion hypothesis, Rosenbaum (1986a) conducted two experiments using Byrne’s (1971) para- digm. In those experiments, participants were shown the opinions of a stranger who was either similar or dissimilar to them. In addition, Rosenbaum included a control condition in which ratings were made in the absence of attitude information. In the first experiment, participants were shown photographs of people along with either no attitude information, .20 similar attitudes, or .80 similar attitudes. Attraction ratings based on .80 similar attitudes did not differ significantly from the ratings based on the photo- graph and no attitude information. However, ratings of the target with .20 similar attitudes were significantly lower. The second study was a naturalistic field experiment conducted at the Iowa Fang Fang Chen and Douglas T. Kenrick, Department of Psychology, Arizona State University. We thank Stephen G. West, Delia S. Saenz, and Steven L. Neuberg for their comments on this article. We also express our appreciation to our research assistants, Terri Corder, Erin Steffin, Ryan Kalleitner, Andy Delton, and Liz Lightner, for conducting the experiments. Correspondence concerning this article should be addressed to Fang Fang Chen, Department of Psychology, Arizona State University, Tempe, Arizona 85287-1104. E-mail: xiyu@asu.edu Journal of Personality and Social Psychology Copyright 2002 by the American Psychological Association, Inc. 2002, Vol. 83, No. 1, 111–125 0022-3514/02/$5.00 DOI: 10.1037//0022-3514.83.1.111 111