Journal Articles
Abstract:
Personality traits are typically measured with single numbers that summarize self-beliefs or frequencies of certain behaviors. Trait manifestation in behavior, however, can be highly variable within individuals across time and situations. This variability gives rise to several aspects of trait expression that single-number summaries do not capture, such as average levels of variability and median or maximum levels of trait expression. The current research examined (1) whether people’s self-concepts include variability in trait expression, (2) whether people can accurately describe not only their average level of trait expression, but their entire distributions of trait expression via self-report, and (3) what implications this variability might have. In four studies, participants reported substantial amounts of variability in trait expression using a new measure of traits. This new instrument demonstrated discriminant, convergent, and predictive validity, as well as internal consistency and clear factor structure. Several parameters of participants’ self-reported distributions of trait expression strongly predicted corresponding parameters of actual behavioral distributions, and higher variability was related to long-term negative affect and low self-concept clarity. Self-reported variability was not related to several alternative measures of variability. Results indicate that variability in trait expression is part of people’s self-concepts, that people can report this variability, and that this variability might have important implications.