Self-Reported Variability and Enhancement of Prediction

Abstract

The present study examined the role of self-reported variability in facilitating the prediction of behavior. Anxiety and extraversion scores based on the 16 PF were obtained for 100 (50 male and 50 female) students in introductory psychology. Each participant also was asked to judge how variable he was in each trait and to judge how anxious he would be in each of 18 situations and how extraverted he would behave in each of 17 situations. Multiple regression analyses showed that self-reported variability enhanced prediction of experiences of anxiety in 13 of 18 situations but had minimal impact on enhancing prediction of extravertive behavior. The results, in contrast with other research, suggest that self-reported variability need not undermine prediction of behavior and support the use of aggregate criteria and predictors to facilitate prediction.