Trait enactments as density distributions: The role of actors, situations, and observers in explaining stability and variability.

Abstract

The purposes of this article were to determine (a) whether the high consistency of individual differences in average aggregated behavior is because of actors' personalities or because of the consistency in the situations those actors encounter; and (b) whether the surprisingly high within-person variability in trait enactment is a real phenomenon corroborated by observers, or merely in individuals' heads. Although traits are supposed to describe what individuals are like in everyday life, little evidence exists about the enactment of trait content in everyday life. Past experience-sampling studies have found both highly variable and highly consistent trait enactment, but were restricted to self-report data and to naturally occurring situations. The current study used experience-sampling in controlled lab environments with 97 targets and 183 observers to address these shortcomings. Targets attended hour-long lab sessions 20× each and observers rated targets' behavior. Parameters of distributions were highly consistent (rs ∼ .80), revealing that actors were responsible for consistency, not situations. Nonetheless, observer ratings revealed that most variability in trait enactment was within-person, confirming that even when people put it on the line in ways that affected others, they still varied rapidly in the traits they enacted. In the face of 2 historically vexing objections to traits, this article supports the density distributions model of traits and argues that trait conceptualizations must accommodate large within-person variability.