How does intergroup contact reduce prejudice? Meta-analytic tests of three mediators

Recent years have witnessed a renewal of interest in intergroup contact theory. A meta-analysis of more than 500 studies established the theory’s basic contention that intergroup contact typically reduces prejudices of many types. This paper addresses the issue of process: just how does contact diminish prejudice? We test meta-analytically the three most studied mediators: contact reduces prejudice by (1) enhancing knowledge about the outgroup, (2) reducing anxiety about intergroup contact, and (3) increasing empathy and perspective taking. Our tests reveal mediational effects for all three of these mediators. However, the mediational value of increased knowledge appears less strong than anxiety reduction and empathy. Limitations of the study and implications of the results are discussed. Copyright # 2008 John Wiley & Sons, Ltd. Recent decades have witnessed an intense renewal of research and theoretical interest in Allport’s (1954) intergroup contact hypothesis. Hundreds of papers have appeared since the publication of Hewstone and Brown’s (1986) Contact and Conflict in Intergroup Encounters. This work has advanced the hypothesis into a developed theory (Brown & Hewstone, 2005; Pettigrew, 1998), shown its applicability to a wide variety of groups and settings, and firmly established its basic contention that intergroup contact typically diminishes intergroup prejudice (Pettigrew & Tropp, 2006). These results, combined with their direct policy implications, have encouraged a concentrated research effort to understand the processes involved and how this established effect can be maximized (Brown & Hewstone, 2005). A meta-analysis of 515 studies involving a quarter of a million participants in 38 nations estimates the mean effect size between contact and prejudice as a correlation coefficient of .21. Greater intergroup contact typically corresponds with lower levels of intergroup prejudice, and 94% of the studies reveal an inverse relationship between contact and prejudices of many types. Moreover, the more rigorous experimental studies yield a stronger mean effect of r1⁄4 .33 (Pettigrew & Tropp, 2006). Contact theory has now been extended in new directions. Pettigrew and Tropp’s (2006) meta-analytic work shows that contact effects hold equally well for groups other than races and ethnicities for whom contact theory was originally developed. Additionally, researchers have begun to examine moderators of contact’s effects. For example, contact in the form of cross-group friendships (Levin, van Laar, & Sidanius, 2003; Paolini, Hewstone, Cairns, & Voci, 2004; Pettigrew, 1997) or structured under Allport’s optimal conditions (Pettigrew & Tropp, 2006) typically promotes greater reductions in intergroup prejudice. Additionally, affective indices of prejudice tend to yield stronger contact effects than such cognitive indices as stereotypes (Tropp & Pettigrew, 2005a). And the effects of contact are significantly stronger for majority group members than for members of minority status groups (Tropp, 2007; Tropp & Pettigrew, 2005b). artment of Psychology, University of California, Santa Cruz, Santa Cruz, CA 95064, USA. red to the June 2006 conference of the Society for the Psychological Study of Social Issues held in Long s, Ltd. Received 30 November 2006 Accepted 16 January 2008 How does intergroup contact reduce prejudice? 923 MEDIATORS OF THE CONTACT-PREJUDICE EFFECT These consistent and widespread findings raise a central question: just how does intergroup contact reduce prejudice? A number of different processes have been proposed and tested in the research literature. Consider the three most commonly tested mediational processes.