On the inferential epistemics of trait centrality in impression formation

We provide a novel, inferential, account of the trait centrality phenomenon. We suggest that a trait possesses the property of ‘‘centrality’’ to the extent that it is subjectively deemed to imply other traits. Five studies explore four central elements of this view. First, trait relations can be stored as unidirectional rules (‘‘if X then Y’’ but not necessarily ‘‘if Y then X’’). Second, the strength of individuals’ lay inference rules determines the effect of traits on impressions. Third, situationally manipulating the strength of lay inference rules influences the impact of traits on impressions. Fourth, the impact of an inference rule is reduced when it is difficult to discern the inference rule and when processing resources are limited. Copyright # 2009 John Wiley & Sons, Ltd. Forming impressions of persons constitutes a key aspect of social interaction, as beliefs about others drive to a considerable degree the interpersonal decisions and actions one undertakes. For example, an impression that a politician is principled may prompt one to lend her or him one’s vote, a supposition that a lawyer is articulate may lead one to choose her or him as a defender, and an opinion that an athlete is motivated may determine her or his choice as an inductee for a team. Prior research has suggested that in forming impressions, people organize a large amount of behavioral evidence around a few central traits (Asch, 1946; Goldman, Cowles, & Florez, 1983; Kelley, 1950; Sedikides & Skowronski, 1993; Watkins & Peynircioglu, 1984; Wyer, Bodenhausen, & Srull, 1984). In this sense, Solomon Asch’s (1952, p. 208) distinction between ‘‘qualities that furnish the key to a person and those that are subsidiary’’ remains critically important for understanding impression formation. Though past research has amply documented that different traits may vary in their degree of centrality, the question has remained what, exactly, contributes to a trait’s centrality. Asch (1946) originally posited that trait centrality resides in the traits themselves, representing a property of their specific contents. However, follow-up research has challenged this position, and suggested that trait centrality emerges out of the specific relations between co-occurring traits (e.g., principled–trustworthy, articulate–intelligent). This research highlighted the relational, rather than essentialist, nature of the phenomenon (Wishner, 1960; Zanna & Hamilton, 1972). We use principles derived from Lay Epistemic Theory (Kruglanski, 1989) to shed light on the phenomenon of trait centrality, and the representation of trait relations in memory more generally. In brief, we propose that information that a target of evaluation possesses a given trait functions as evidence for inferences about this target’s other traits and properties. In this sense, a trait is central to the extent it implies other traits, thereby affording generalized social judgment. Three elements of this proposition are of present interest: (1) The structure of the inference rule that links a specific trait to ent of Psychology, University of Maryland, College Park, MD 20742, USA. s, Ltd. Received 16 May 2008 Accepted 20 July 2009 Trait centrality 1121 other traits to enable a judgment; (2) the notion of subjective relevance, which refers to the strength with which a trait is believed to imply other traits; and (3) the difficulty of gleaning the subjectively relevant evidence from the environment. We present six studies that attest to the unidirectional ‘‘if. . . then’’ structure of the mental representation of trait relations, and the role of lay inferences in the trait centrality phenomenon. We also show that the ability to use trait information as evidence for a given impression depends on task demands and individuals’ available processing resources.