Emotional memories are (usually) harder to forget: A meta-analysis of the item-method directed forgetting literature.

The current meta-analysis explored whether emotional memories are less susceptible to item-method directed forgetting than neutral memories. Basic analyses revealed superior memory for remember (R) than forget (F) items in both the neutral, M = 19.6%, CI95% [16.1, 23.1], and the emotional, M = 15.1%, CI95% [12.4, 17.7], conditions. Directed forgetting in either valence condition was larger for (a) words than for other stimuli; (b) recall than recognition tests; (c) studies that used recall prior to recognition testing; (d) shorter lists; and (e) studies that included buffer items. Direct comparison of the magnitude of the directed forgetting effect across neutral and emotional conditions within studies revealed relatively diminished directed forgetting of emotional items compared to neutral items, with an average difference of 4.2%, CI95% [2.0, 6.4]. However, the nature of this finding varied broadly across studies, meaning that whether - and to what degree - emotional memories are more resilient than neutral memories likely depends on the methodological features of the study in question. Moderator analyses revealed larger differences (a) in studies for which the emotional items were more arousing than the neutral items, and (b) when buffer items were included. Together, these findings suggest that emotional memories are often more resilient to intentional forgetting than neutral memories, although further research is necessary to characterize the circumstances under which these differences emerge.