Attachment and capitalizing on positive events

Abstract

Although previous work has examined how individual differences in attachment security affect support processes for negative events, little work has looked at how attachment security affects support for positive events (capitalization). In a 10-day diary study of romantic couples, we examined the association between individual differences in attachment security and perceptions of capitalization support. We also examined how attachment security moderated the relationship between capitalization support and daily emotions, relationship satisfaction, and life satisfaction. The results showed that stronger avoidance orientations were associated with reduced perceptions of partner responsiveness. Anxiety moderated the association between responsive support and daily outcomes; for those high (versus low) on anxiety, daily relationship and life satisfaction were more strongly tied to partners’ capitalization responses, and more negative emotions were experienced regarding those responses. Overall, insecure attachment was associated with mixed reactions to receiving responsive capitalization support.

Keywords:

• capitalization
• attachment style
• responsiveness
• daily diaries
• romantic relationships

Acknowledgements

This study was supported by a grant from the National Science Foundation (BCS-0444129) awarded to the second author.

Notes

1.In this data set individual is technically crossed with couple, not nested within couple. Barnett, Raudenbush, Brennan, Pleck, and Marshall (1995) recommend a two-level no intercept model in which data from both members of the pair are nested within the couple; coefficients of men and women are estimated separately but simultaneously at level 1 and couple is accounted for at level 2, as opposed to a three-level model which we used. However, because our couples did not necessarily share positive events on the same days (and often did not), using a two level model to analyze the current data would have been problematic because in such a two-level model only data from days that each member of the couple shared an event with one another would have been included in the model; all other days would be missing (e.g., she shared a positive event but he did not). This would result in a functional reduction in the data of all days in which only one member of the couple shared an event. Therefore, we treated the individual as nested within couple, not crossed with couple.

2.Some participants did not have a score on each variable (indicating that they did not report sharing any positive events across the week). In these cases, a couple’s score was calculated from only one member of the couple. For two couples, neither member reported sharing a positive event over the diary period (and thus no emotion composite could be calculated).

3.We also ran all the analyses assuming no dependence in the data, using each participants scores (N = 67) and the pattern of results were similar, although the relationship between the emotion composite and satisfaction with life was no longer significant (r(65) = −0.178, p = 0.149). The correlation between the emotion composite and relationship satisfaction remained significant (r(65) = −0.337, p = 0.005).