Research that includes measures of affect has often used measures of affective state, subjective well-being, and affective disposition interchangeably. The present study examined the relationships between three widely-used measures of these constructs. Examination of coefficient alpha and test–retest reliabilities indicated that there are temporal differences between the constructs. Confirmatory factor analysis models indicated that affective state correlates with subjective well-being, which in turn correlates with affective disposition. Results indicate that the constructs are related but not equivalent, and that researchers should ensure they use the most valid instrument for the construct they wish to measure.
In recent years, affect has played a prominent role in psychological, educational, human resources, and consumer behavior research (Barone et al., 2000, Efklides et al., 1997, George, 1991, Gencoz, 2002, Judge & Locke, 1993, Lee & Sternthal, 1999, Levin & Stokes, 1989 and Varma et al., 1996). For this reason, many researchers have included measures of affect in their studies.
Psychology, in particular, has enjoyed a continuing debate concerning the dimensional and temporal aspects of affect. Specifically, affect has been examined frequently through operationalizations such as: the multidimensional circumplex model (Feldman Barrett & Russell, 1998 and Remington et al., 2000); the bipolar positive and negative affect model (Gencoz, 2002, Gomez et al., 2000 and Watson et al., 1988); bipolar intensity and frequency (Simonsson-Sarnecki et al., 2000 and Schimmack & Diener, 1997); unidimensional affective disposition (Connolly & Viswesvaran, 2000 and Judge, 1992); and temporally-based long-term and short-term affect (Watson, 1988b).
The choice of instruments to measure affect has not always been construct-valid, however, due to a lack of clarity in distinguishing between different aspects of affect. Our study offers distinctions between three temporally-based constructs of affect by operationalizing affect as: unidimensional affective disposition; bipolar positive and negative affective state; and bipolar positive and negative subjective well-being. Using a repeated-measure design, we examine the relationships between these three temporally-based constructs of affect and we examine the long-term stability of these constructs.
