Valence Explains How and Why Positive Affects and Negative Affects Correlate: A Conceptual Replication and Extension of Diener et al.'s (1995) The Personality Structure of Affect

Diener et al. (1995) used a multimethod approach to test a hierarchical model of trait affect. The model suggests that specific trait affects are related to each other by two, distinct, but negatively correlated factors. We report the results of a conceptual replication study that addressed several limitations of Diener et al.'s (1995) study. We used three ethnically diverse samples which included a group of undergraduates along with both of their biological parents. As such, in terms of generalizability, we improved upon the original study which was limited to a student sample by also including middle-aged adults as targets. Most importantly, we included measures of hedonic tone to validate the interpretation of the higher-order factors as positive affect and negative affect. Also, we did not average informant ratings to model individual rating biases. Further, we used item-level indicators rather than item averages as indicators of basic affects. Our results confirm Diener et al.'s (1995) model and demonstrate that positive trait affect and negative trait affect are negatively correlated and account for the covariance among specific affects. We discuss the implications of these results in the context of personality theories that consider positive trait affect and negative trait affect as independent factors related to extraversion and neuroticism, respectively (Costa & McCrae, 1980). We argue that this model cannot account for the negative correlation between positive affect and negative affect and that further research is needed to locate affect within the Big Five model of personality.