Abstract

Two experiments were conducted to test a 3-level attitude change model proposed by Na(1998). Instead of "high involvement → high elaboration" hypothesis in ELM (elaboration likelihood model), the extended 3-level attitude change model assumes "moderate involvement → high elaboration → objective processing," similar to the framework of arousal-performance relationship. Not only low involvement but also extremely high involvement (and potentially high arousal level) intrinsic in strong attitudes may impede the objective processing. Thus, differential effect of strong and weak messages on attitude change is predicted only in moderate attitudes due to the optimum level of involvement and arousal. Experimental designs were 3(attitude strengths) x 2(message strengths) between-subjects factorial design for Experiment 1, and 3(attitude strengths) x 2(message strengths) x 2(ingroup vs. outgroup sources) for Experiment 2. Dependent variables were the amount of attitude change, the types of cognitive responses, and the evaluation of message persuasiveness. As predicted by the 3-level model, only the moderate attitude holders demonstrated greater attitude change when given a strong rather than a weak message in Experiment I and in the ingroup source data of Experiment 2. Weak attitude holders showed considerable attitude change and strong attitude holders showed resistance regardless of message strengths. When given a strong message (Experiment 1) or when given a strong message from an outgroup source (Experiment 2), a boomerang effect was observed beyond a simple resistance. In both the experiments, counter-message arguments were relatively more generated by strong attitude holders, pro-message arguments more by moderate attitude holders, and indifference appeals more by weak attitude holders, supporting the 3-level model. Attitude strengths were highly correlated with integral affect as predicted. Moreover, it was also proved that the effect of initial conviction on the de-valuation of message persuasiveness was mediated by the strength of negative emotions (i.e., arousal level) agitated by the counter-attitudinal message.