Two studies used a round-robin design to examine whether the observers made consensual judgments of targets' degree and quality of intergroup contact, and whether these consensual judgments were correlated with the targets' own self ratings, and moderated by the observability of the contact. Study 1 revealed projection/assumed similarity, with participants rating others as similar to themselves to a large extent, but also yielded evidence for the validity of whites' self-reports of direct, but not extended, intergroup contact with Asians, even when controlling for extraversion and perceived attitudes. Study 2 replicated the main results, using both Asians and Gay men as outgroups, and showed that participants' ratings discriminated between the two discrete outgroups, with measures of contact and attitude being only meaningfully related within, but not between, outgroups. Overall, these findings help to validate self-report measures of direct intergroup contact.

A correlational study investigated whether individual difference variables (public self-consciousness and social comparison) moderate the impact of extended cross-group friendship on attitudes towards two outgroups (Asians and gay men). Social comparison moderated the impact of extended crossgroup friendships on attitudes to both Asians (significantly) and gay men (marginally). There were no reliable effects of public self-consciousness. These results are consistent with findings that extended cross-group friendship works primarily when individuals attach importance to social norms, and that attitudes towards some outgroups (e.g., ethnic outgroups) are more influenced by norms than are other attitudes (e.g., attitudes towards gay men). We argue that the moderation effect of social comparison does not weaken the potential of extended contact to reduce prejudice, because intergroup encounters typically involve some degree of uncertainty and foster reliance on social norms.