Using causal persuasive arguments to change beliefs and teach new information: The mediating role of explanation availability and evaluation bias in the acceptance of knowledge.

Abstract

In this article processes by which causal arguments effect change in established beliefs were explored. The hypothesis that explanation availability mediates belief change in response to causal arguments was tested in 2 experiments. Persuasive communications used causal evidence, noncausal (statistical) evidence, or both to argue that AIDS is not transmissible by casual contact. Results supported the authors' hypothesis. Causal arguments produced the greatest belief change, with the effect mediated by explanation availability. Causal arguments were also less subject to evaluation bias, consistent with a cognitive interpretation of the biased assimilation phenomenon (in which evidence in favor of one's position is evaluated more favorably than evidence in opposition). Experiment 2 replicated the basic effects with dependent measures (including a measure of behavioral commitment) obtained 3 weeks after presentation of the communications. These results suggest that a causal component strengthens the educational potential of persuasive arguments. Persuading an audience to accept new information seems particularly difficult when audience members can explain why they hold their current beliefs. The persuasion task then goes beyond simply convincing an audience of the correctness of the new information; persuasion must also overcome forces that maintain established beliefs. For example, in teaching social psychology, we have seen students resist ideas that challenge their common sense understanding of social behavior. Of course, common sense is derived from life experience and is rich with explanations. Researchers studying learning in children have observed a similar phenomenon. In several studies, incorrect prior knowledge interfered substantially with the learning of correct scientific information (Alvermann & Hague, 1989; Alvermann,