Blaming the same-sex victim in HIV-prevention messages: Further examination of the self-protective similarity bias

It has been demonstrated recently that men will judge their own (threat-relevant) personalities and sexual practices as safer than another man's if that man's HIV status is believed positive compared to negative or is unknown (Gump & Kulik, 1995). The present experiment was designed to expand our understanding of the moderators and mediators of this recently documented "self-protective similarity bias." College students (N= 150) participated in a 2 (Sex of Participant) x 2 (Sex of Model) x 3 (Serostatus: Positive, Negative, Unknown) x 2 (Threat Relevance of Item) factorial design with repeated measures on the last factor. Results indicated that the similarity bias specifically occurs with same-sex models. Analyses of self-ratings and model ratings suggest that the similarity bias was more a function of "blaming" or devaluing the victim than of inflated ratings of the participant's own safety characteristics. Finally, although this bias reduced perceived personal susceptibility and was specific to same-sex models, intentions to adopt safer sexual practices were raised by all HIV-positive models regardless of sex concordance.