Self-discrepancy and reduced autobiographical memory specificity in ruminating students and depressed patients

Abstract

This work uses self-discrepancies as a unifying framework to understand the relationship between autobiographical memory specificity (AMS) and components of rumination (reflection and brooding). Rumination can be triggered by the awareness of a discrepancy between one's current state of being and one's desired state of being. Such discrepancies may partly underlie the phenomenon of reduced AMS (a.k.a. overgeneral memory), which is a phenomenon of great clinical importance especially for depression. The aim of the present studies was to experimentally investigate the impact of a self-discrepancy manipulation on AMS in a student sample (Study 1) and in a depressed sample (Study 2). Results failed to reveal a direct effect of a self-discrepancy manipulation on AMS, but in both studies, actual–ideal discrepancies interacted with the reflection component of rumination with respect to memory specificity. In the self-discrepancy condition, there was a negative association between reflection and change in AMS: Higher reflection was associated with a greater decrease in AMS from pre- to post-testing. The results of these two studies suggest that the interplay between components of rumination and self-discrepancy has an effect on memory specificity.

Keywords:

• Autobiographical memory
• Self-discrepancy
• Rumination
• Depression
• Overgeneral memory

Acknowledgments

This research was supported by a Research Program of the Research Foundation-Flanders (FWO) (G.0339.08) awarded to FR. We also gratefully acknowledge the support of the FWO to JWG (GP.035.11N).

Notes

¹We also conducted analyses with post-manipulation AMS scores (specific and categoric) as the dependent variable and pre-manipulation scores as a covariate. Neither analysis was significant, ps > .12.

²Because our self-discrepancy manipulation also resulted in an increase in sadness, which was unexpected, we repeated all regression analyses using the change in the VAS for sadness as covariate. This was not a significant predictor in any model, and the key interactions were still significant. In addition, based on the request of an anonymous reviewer, we repeated our regression analyses using post-manipulation AMS scores (specific and categoric) as the dependent variable; and pre scores as covariates. Across these four models, the key interactions were not significant, ps > .15.

³As with Study 1, analysing only post-manipulation scores with pre-manipulation scores as a covariate did not change the results, ps > .35.

⁴As with Study 1, using post-manipulation AMS scores (specific and categoric) as the dependent variable and pre-manipulation scores as a covariate, did not change the results. The key interaction was not significant in any model examined, ps > .067.

⁵We performed a regression analysis with change in the VAS on focus on self-discrepancy as the dependent variable, standardised Reflection scores and group as the predictors in the first block and the interaction term in the second block. We failed to find a moderation effect of reflection on the change in focus on self-discrepancy (B=9.72, t=1.31, p=.20 for Study 1 and B= − 1.86, t= − 0.13, p=.90 for Study 2).