In two studies, we examined how individual differences in causal uncertainty (CU), causal importance (CI), and initial attitudes affected the processing of a persuasive message that contained causal or non-causal arguments. We predicted that high CU individuals' doubts about their causal understanding of events would be activated when they were presented with counterattitudinal arguments. When these individuals also placed a high value on causal understanding (high CI), they should scrutinize any available causal explanations. As a result, they should be more persuaded by strong compared to weak causal arguments. In support of these predictions, we found in two studies that high CU/high CI participants were more persuaded by strong compared to weak counterattitudinal causal arguments. Mediational analyses in Study 2 revealed that high CU/high CI participants were more persuaded by strong causal arguments because they were more confident in them. Implications for the CU model and persuasion processes are discussed.Social psychologists long have been interested in how people understand the causes of events in the social world. Early attribution theorists argued that in order to predict and control their social environments, people attribute behavior and events to stable, underlying causes (Heider, 1958). This early work and subsequent research has provided us with a wealth of information about attribution processes (for a review see Gilbert, 1998) and the critical role they play in many different domains.
44Social Cognition, Vol. 26, No. 1, 2008, pp. 44-65 Stephanie J. Tobin, Department of Psychology, University of Houston; Gifford Weary, Department of Psychology, Ohio State University.This article is based on a doctoral dissertation submitted by Stephanie Tobin to the Ohio State University. We would like to thank committee members Russell Fazio and Richard Petty for their valuable feedback on this research. We also would like to thank members of the 2002-2004 Group for Attitudes and Persuasion and the Weary Lab for their comments on this research and Pablo Briñol and two anonymous reviewers for their comments on an earlier version of this paper. A partial report of the data was presented at the 2004 Society for Personality and Social Psychology meeting.