Capitalization interactions, in which partners share positive events, typically produce positive relationship outcomes (Gable, Gonzaga, & Strachman, 2006). However, the limiting conditions of these interactions have not been examined. In this study, 101 dating couples discussed a positive event in the life of each partner. Ratings of perceived responsiveness were made by both the romantic partner who disclosed a positive event and his/her responding partner. Trained observers then rated each videotaped interaction. More avoidantly attached responders reported being and were rated by coders as less responsive, particularly if their disclosing partners were more anxiously attached. Avoidantly attached individuals also underestimated their partners' responsiveness relative to observer ratings. Anxious responders underestimated their own responsiveness when their disclosing partners were more avoidantly attached. These results indicate that insecurely attached individuals are relatively less likely to be responsive and to perceive responsiveness in capitalization interactions than are more securely attached individuals. This is especially true when highly anxious and highly avoidant individuals are relationship partners.

There are many ways in which the provision of social support can be ineffective. Recent research suggests that the benefits of support may be maximized when it is provided invisibly. What remains unknown, however, is whether invisible support reflects the skillful behavior of support providers or recipients' blissful unawareness, as well as how invisible support is delivered during spontaneous social interactions. We hypothesized that both providers' skillful behavior and recipients' unawareness are necessary for invisible support to be effective, and we sought to document what effective invisible support looks like. Eighty-five couples engaged in a videotaped support interaction in the lab. Support recipients whose partners provided more invisible practical and emotional support (coded by observers) but who reported receiving less support experienced the largest preinteraction-to-postinteraction declines in negative emotions. In the case of practical invisible support, the combination of more support and less awareness of that support also predicted increases in self-efficacy. These results indicate that invisible support is a dyadic phenomenon.

Individual differences in empathic accuracy (EA) can be assessed using daily diary methods as a complement to more commonly used lab-based behavioral observations. Using electronic dyadic diaries, we distinguished among elements of EA (i.e., accuracy in levels, scatter, and pattern, regarding both positive and negative moods) and examined them as phenomena at both the day and the person level. In a 3-week diary study of cohabiting partners, we found support for differentiating these elements. The proposed indices reflect differing aspects of accuracy, with considerable similarity among same-valenced accuracy indices. Overall there was greater accuracy regarding negative target moods than positive target moods. These methods and findings take the phenomenon of "everyday mindreading" (Ickes, 2003) into everyday life. We conclude by discussing empathic accuracies as a family of capacities for, or tendencies toward, accurate interpersonal sensitivity. Members of this family may have distinct associations with the perceiver's, target's, and relationship's well-being.

This paper provides a roadmap for integrating a dyadic framework into individual-level models of behavior change. The findings suggest that data from both partners and relationship quality are important to consider when trying to understand and change health-related behavior such as physical activity. The results broaden the potential applications of the TPB as well as our understanding of how romantic partners might influence important health-related practices. (PsycINFO Database Record