Abstract

This study examined the long-term temporal stability and predictive utility of trait Positive Affect and Negative Affect scales. Participants were assessed initially as undergraduates: They rated how they felt generally (general sample) or during the past year (year sample). They were retested on a general affect measure and on scales assessing current depression and anxiety approximately 6 (general sample) or 7 (year sample) years later. By this time, all of them had graduated from college and most were working full-time. Negative Affect scores declined significantly over the study. Nevertheless, the Negative and Positive Affect scales both displayed a significant, moderate level of stability. Moreover, initial scores on both scales correlated significantly with measures of current symptoms that were completed several years later. Thus, trait affect scales were substantially stable--and maintained significant predictive power--even across extended time spans.