Abstract
According to the CaRFAX model, rumination is one of the key underlying mechanisms of overgeneral autobiographical memory (OGM). The association between rumination and OGM is well established in clinical populations, but this relationship is not robust in nonclinical samples. A series of null findings is reported in the current paper. Additionally we followed up on recent findings suggesting that a state of rumination needs to be active in order to detect a relationship between trait-rumination and OGM. Secondary school students (N= 123) completed questionnaires assessing trait-rumination and depressive symptoms as well as two autobiographical memory tests (AMTs), one before and one after a self-discrepancy induction. This induction should trigger state-rumination, which would subsequently promote the retrieval of general rather than specific memories. Trait-rumination failed to predict increases in OGM. We did find, however, that higher BDI-II scores were positively related to an increase in OGM following the induction. This adds to the growing body of evidence that OGM reactivity might be more important than baseline memory specificity.
Acknowledgments
Jorien Smets is research assistant of the Research Foundation-Flanders (FWO). We also gratefully acknowledge the support of the FWO to James Griffith (GP.035.11N).
Notes
1All analyses described in this paper were repeated using the proportion of specific and categoric memories, which ignored omissions in the calculation of denominator of specificity scores. The pattern of results was the same.
2Following Raes et al. (Citation2012) we also performed two mixed ANOVAs. RRS scores were split at the median, resulting in one group of 62 participants who scored below (RRS<43) and one group of 60 participants with scores above (RRS>=43). First, the number of specific memories was used as dependent variable, ruminative scores (high vs low) as between-participants variable, and time (AMT1, AMT2) as a within-participants variable. There were no significant main or interaction effects. Second, number of categoric memories was used as dependent variable, ruminative scores (high vs low) as between-participants variable, and time (AMT1, AMT2) as a within-participants variable. There were again no significant main or interaction effects.
3We have also checked whether depressive symptoms moderated the relationship between trait-rumination and post-induction OGM or change in OGM, but analyses showed that this was not the case.