The purpose of this study was to examine how positive and negative relationship characteristics and their interaction were predictive of global relationship happiness, psychological symptoms, and physical health in a large sample of older couples. The NSHAP Wave 2 partner data n = 955) were used to estimate both actor (within person) and partner (between person) effects using the Actor Partner Interdependence Model with Moderation. Global relationship happiness was predicted by the interaction of positive and negative characteristics, suggesting that more positive characteristics are only predictive of greater happiness in the presence of some negative characteristics. Male partners' depressive symptoms were predicted by the female partners' interaction of positive and negative characteristics, suggesting that negative characteristics were only associated with male partner depressive symptoms when positive characteristics were rated very low by their partner. Positive and negative characteristics were predictive of anxiety and stress but not their interaction, and only the male partner's perception of positive characteristics was predictive of the female partner's physical health. These results are discussed in the context of controversy over the measurement of relationship satisfaction in the field.
Keywords: dyadic adjustment; dyadic analyses; older couples; relationship satisfaction relationship quality.