Researchers usually assume the effect of dispositional variables on outcomes as stable over time. We disagree with this dominant assumption (i.e., Fixed Model) and argue that if dispositional variables lead to different extent of reactions between two time waves, it will have both indirect and direct effect on outcomes at time 2 (i.e., Continuous Model). It is still possible that dispositional variables may moderate the relationship between outcome variables over time (i.e., Interacting Model). Using the effect of four objective variables (i.e., gender, age, education, and tenure) and four abstract trait emotions (i.e., happiness, sadness, depression, and disgust) on job satisfaction as examples, we illustrate that these three possibilities may exist, depending on the particular dispositional variable.
