The similarity-attraction relationship: Is there an ideal one?

Abstract

It is proposed that we are attracted to similar individuals because: (1) such individuals are similar to our ideals and (2) such individuals are similar to ourselves. Previous studies have employed a similarity manipulation which has confounded these two components. When the components are separated, similarity to the ideal was expected to exert a major influence and similarity to the self a minor influence on attraction. A series of experiments orthogonally manipulated similarity to self and similarity to ideal with a new technique involving content-free dimensions. The results indicated a consistent main effect for ideal similarity on liking and no consistent main effect for similarity to the self. Finally, the results also indicated that given any two of the three variables—similarity to self, similarity to ideal, and own self to own ideal discrepancy—subjects seemed able to infer the other or third variable in a fairly logical fashion.