This study considered the validity of Petty and Cacioppo's (1981, 1986aJ 1986b) Elaboration Likelihood Model of Persuasion (ELM) in the context of an attempt to enhance favourable attitudes toward AIDS victims and homosexuals. The relative impact of "central" versus "peripheral" route processing of a dispassionate written message and a threat appeal was examined. It was hypothesized that a favourable advocacy would, when processed centrally, induce attitude change which is relatively more persistent over time than if it were processed peripherally. A significant gender difference in attitudes toward both AIDS victims and homosexuals was also hypothesized and a high positive correlation between these attitudes was predicted. Attitudes were measured using the "Kite Homosexual Attitude Scale" (Kite & DeauxJ 1986), and the "Attitudes to AIDS Victims Scale", an instrument devised by the author for the purposes of this study.

A significant improvement in subjects' attitudes toward AIDS victims and homosexuals was recorded over the four weeks between assessments. Gender was found to be a powerful predictor of attitude. Consistent with expectations, males were significantly less favourable in their attitudes toward both AIDS victims and homosexuals than females. Scores on the two attitude measures were highly correlated. (p=.7775). In all other respects the findings were inconsistent with the experimental hypotheses and the adequacy of the ELM was called into question as a result. Because of the limitations of the study, however, the issue was inconclusive. The use of threat appeals as a preventive health strategy was discounted.