Does attitude certainty beget self-certainty?

This research explores the possibility that changes in attitude certainty can affect general self-certainty and, thus, have consequences that extend beyond the attitude domain. Across two studies, attitude certainty is manipulated using repeated attitude expression and attitude consensus paradigms. The implications of these manipulations are tested for feelings of general self-uncertainty (Study 1) and global selfdoubt about one’s abilities (Study 2). In each study, it is demonstrated that participants feel greater selfcertainty under conditions of high rather than low attitude certainty, but only when they view aspects of the attitude as central to their self-concept. 2008 Elsevier Inc. All rights reserved. Securing certainty is a fundamental goal in judgment formation (Kruglanski, 1989; Petty, Briñol, Tormala, & Wegener, 2007). This has been revealed in research on a variety of topics, including attitudes (Tormala & Rucker, 2007), the self (DeMarree, Petty, & Briñol, 2007; Wright, 2001), impression formation (Yzerbyt, Schadron, Leyens, & Rocher, 1994), stereotyping (Sechrist & Stangor, 2001), and mental health (Weary & Edwards, 1994). Particularly extensive attention has been paid to certainty in the attitudes domain. Whereas an attitude refers to one’s general evaluation of something (e.g., a product or policy), attitude certainty refers to the sense of conviction one has about that evaluation (Abelson, 1988; Gross, Holtz, & Miller, 1995). Attitude certainty has garnered considerable research attention primarily because of its important consequences for an attitude’s durability and influence. High certainty attitudes typically are more resistant to change (Bassili, 1996; Wu & Shaffer, 1987) and more predictive of behavior/choice (Bizer, Tormala, Rucker, & Petty, 2006) than their low certainty counterparts. The current research tests a novel consequence of attitude certainty. We explore the possibility that manipulations of attitude certainty can affect general self-certainty and, thus, have consequences that reach beyond the attitude domain. We define selfcertainty broadly to include the sense of clarity one has about one’s personality or self-concept and the sense of confidence one has about one’s general competence and abilities. Could becoming less certain of one’s attitude toward an issue like abortion increase one’s need for individualized feedback following a personality ll rights reserved. . test? Could becoming more certain of one’s attitude toward abortion reduce one’s doubt about one’s general abilities? On the face of it, these effects seem highly improbable—one’s attitude toward abortion is considerably different from one’s general self-views. However, research and theory suggest that people’s self-concepts are based to some extent on their perceptions of their own attitudes (e.g., Brown, 1991), particularly when the attitude is important or central to the self in some way (e.g., Correll, Spencer, & Zanna, 2004; Crano, 1995). Thus, one might expect attitude certainty to affect general self-certainty, but only when aspects of the attitude are central to the self-concept. Previous research has shown that there are multiple ways to manipulate attitude certainty. For instance, attitude certainty can be increased by repeated attitude expression as well as by learning that one’s attitude is supported by social consensus (e.g., Petrocelli, Tormala, & Rucker, 2007). We propose that these manipulations of attitude certainty can affect self-certainty, but only when aspects of the attitude are central to the self-concept. For instance, repeated attitude expression might increase selfcertainty, but only when the attitude is perceived to reflect one’s core values. Similarly, attitude consensus might affect self-certainty, but only when the consensus comes from an identity-relevant group. Two studies investigate these possibilities. Across studies, we varied attitude certainty using attitude expression (Study 1) and social consensus (Study 2) manipulations. To assess self-certainty, we used a measure of self-uncertainty (desire for personality feedback) and self-doubt about one’s competence and abilities. Our primary hypothesis was that when aspects of the attitude were viewed as central to their self-concept, participants would exhibit greater self-certainty in the high rather than low attitude certainty J.J. Clarkson et al. / Journal of Experimental Social Psychology 45 (2009) 436–439 437 condition. When aspects of the attitude were less self-central, we expected no effect of attitude certainty on self-certainty.