Measurement of Dispositional Affect: Construct Validity and Convergence with a Circumplex Model of Affect

A circumplex model of affect has recently gained tentative acceptance by prominent researchers investigating the structure of dispositional affect. The present study examines the issue of construct validity of several existing measures of affect and their fit with the circumplex model. A traditional multitrait-multimethod matrix approach is utilized first, followed by a confirmatory factor analytic approach. These analyses demonstrate that the data collected using these four scales are characterized by generally good concurrent and discriminant validity. Furthermore, the data are in partial agreement with the proposed circumplex model of affect. Issues for the further development of the circumplex model and the use of these measures are discussed. Recent years, 1999 in particular, have witnessed a flurry of scholarly activity addressing the structure of affect. Long after Russell (1980) introduced his circumplex model of affect, the structure of affect still provokes lively discussion and disagreement. Unfortunately, a commensurate body of literature addressing the measurement of affect has not accompanied the debate over the structure of affect. The present study examines four measures of dispositional affect in the context of a circumplex model of mood that has been the recipient of support from several diverse sources.