Memory and Emotion: The Influence of Valence on Children's Memory for a Salient Event

Abstract

Remembering is a constructive process. Children's memories for events have been shown to conform increasingly over time to script-based expectations, stereotypes and suggestions. Effects of personal, socio-emotional factors on children's recall, however, are less well understood and often juxtapose highly arousing negative events (e.g., stressful medical procedures) with benign positive events (e.g., family gatherings). This research is unique because both the positive and negative groups experienced the same event, thus controlling for differences in event structure and salience.

Children were observed during the final game of an end-of-season soccer tournament. The event was scored for the presence or absence of central and peripheral components of the game, identified in previous research in consultation with peer experts. Recall was assessed immediately after the game and 6 weeks later, through interviews conducted at the field and in participants' homes. Extending previous work within this paradigm, the interview began with a free-recall component followed by elicited recall items. The protocol included misleading questions about plausible central and peripheral components. Participants' free-recall narratives about the event were coded for the proportion of central, peripheral, evaluative and mentalistic propositions in the narrative, and narrative cohesiveness.

Emotional valence of the event was defined initially as event outcome (won/lost). Although event outcome has been shown to correlate with post-game emotion, previous work suggests children's post-event descriptions of pre-game emotions are independent of outcome and related to correct rejection of misleading questions. Ratings of event salience, perceived individual and team performance, and the point at which the participant was confident of the outcome were obtained.

Narratives from participants in the positive condition included a greater proportion of cohesive devices than participants in the positive condition, and narratives contained a greater proportion of evaluative statements at the first interview but a greater proportion of mentalistic statements at the second interview. The proportion of statements about central and peripheral aspects of the event did not differ by time or outcome group. There was an outcome group by time interaction on elicited recall of present and absent features, where participants in the positive group did better on present feature questions and participants in the negative outcome group did better on correct denials of peripheral absent features, both at the first interview. Groups did not differ at the second interview. Participants who knew the outcome early in the event were less likely to correctly recall central event details than participants who did not know the outcome until the end, and there were few differences between the positive and negative outcome group when comparing participants who did not know the outcome until the end.

This research provides further evidence that emotion at the time of encoding is related to recall and change over time in children's memories, and highlights the importance of controlling event structure when comparing recall for positive and negative stimuli.