Memory for media: Investigation of false memories for negatively and positively charged public events

Abstract

Despite a large body of false memory research, little has addressed the potential influence of an event's emotional content on susceptibility to false recollections. The Paradoxical Negative Emotion (PNE) hypothesis predicts that negative emotion generally facilitates memory but also heightens susceptibility to false memories. Participants were asked whether they could recall 20 “widely publicised” public events (half fictitious) ranging in emotional valence, with or without visual cues. Participants recalled a greater number of true negative events (M=3.31/5) than true positive (M=2.61/5) events. Nearly everyone (95%) came to recall at least one false event (M=2.15 false events recalled). Further, more than twice as many participants recalled any false negative (90%) compared to false positive (41.7%) events. Negative events, in general, were associated with more detailed memories and false negative event memories were more detailed than false positive event memories. Higher dissociation scores were associated with false recollections of negative events, specifically.

Acknowledgements

The authors would like to thank the Social Sciences and Humanities Research Council of Canada (SSHRC) and National Sciences and Engineering Research Council of Canada (NSERC) who funded this project through grants to the first author.

Notes

¹A series of Pearson correlation matrices examined the extent to which mean characteristic ratings related to the recall of each emotional, public event. Frequency of recall of true negative, false positive, and false negative events was not significantly correlated with ratings of arousal, plausibility, or global or personal importance (ps>.05). Only plausibility was significantly related to the recollection of true positive events (r=.92, p<.05). No other characteristic ratings were significantly related to the recollection of true positive events, ps>.05.

²Given that only 21 participants reported details of at least one each of true positive, true negative, false positive, and false negative events, we used expected cell mean replacement to allow for a more comprehensive analysis. Expected cell values were calculated by adding the participant's mean across veracity×valence conditions to the condition mean across all participants, minus the grand mean. After expected cell mean replacement for individuals missing only one cell of data (e.g., didn't recall any false positive events and therefore didn't provide any details of these events), N increased to 53 for this analysis. It should be noted that a parallel repeated-measures ANOVA, without cell replacement (N=21), also revealed a main effect of veracity, F(1, 17) = 4.96, p<.05, such that memories for true events (M=4.63; SD=2.25) were associated with a greater number of details than false events (M=3.50; SD=2.49).