Memory of the UK’s 2016 EU Referendum: The effects of valence on the long-term measures of a flashbulb memory.

Abstract

Emotional public events, relative to non-emotional ones, are typically remembered more accurately, more vividly and with more confidence. Such memories are referred to as flashbulb memories. However, the majority of previous studies on this phenomenon have focused on negative public events and less is known about positive ones. The current study examined whether positive and negative public events were remembered in a similar manner by assessing individuals' memory for the time when they learned the results of the UK's 2016 Referendum on its European Union (EU) membership. Participants included UK residents who voted to 'leave' the EU in the referendum and found the event highly positive, UK residents who voted to 'remain' in the EU and found the event highly negative, and US residents who did not vote and found the event neutral. Data from a total of 845 participants were assessed at four time points over the course of 16 months. Growth curve modelling showed that differences in memory between Remain voters (who reported the highest levels of negative emotion) and Leave voters (who reported the highest levels of positive emotion) emerged over time. Specifically, Remain voters maintained higher levels of memory consistency than Leave voters, while Leave voters maintained higher levels of memory confidence than Remain voters. These results indicate that positive and negative public events are remembered differently, such that negative valence enhances memory accuracy, while positive valence results in overconfidence.

Remembering Election Night 2016: Subjective but not objective metrics of autobiographical memory vary with political affiliation, affective valence, and surprise.

“…(Note that although Negative Affect significantly predicted memory quantity when added to the Affiliation model, this was driven by the quadratic term, not the linear term, of this predictor, suggesting that memory quantity may have increased with emotion intensity, but did not differ with valence.) Interestingly, information quantity and consistency significantly increased over time instead of decreasing, in contrast to prior observations from the autobiographical memory literature (Kensinger & Schacter, 2006; Raw et al, 2020; Talarico & Rubin, 2003). When interpreting our consistency results relative to prior work, however, it is important to note that we only measured consistency at T2 and T3, relative to T1; global consistency scores at both T2 ( M = .58, SD = .21) and T3 ( M = .61, SD = .22) are well below a perfect score of 1, indicating both memory decline relative to T1, and an increase from T2 to T3.…”

“…First, Trump supporters reported higher memory confidence and belief that the event occurred the way they remember than Clinton supporters. Given that the election outcome was generally considered a positive emotional event by Trump supporters, this observation is consistent with a growing consensus that memory confidence may be greater for positive autobiographical events compared with negative (Holland & Kensinger, 2012; Kensinger & Schacter, 2006; Raw et al, 2020), which may be attributable to a greater willingness to endorse remembered items under positive versus negative affective states (Bowen et al, 2020; Levine & Bluck, 2004). In addition to memory confidence, a number of additional differences in subjective memory experience were observed as a function of political affiliation.…”

“…Although many investigations of autobiographical memories elicited by public events have characterized negatively-valenced memories, a limited number of studies have been able to characterize autobiographical memory for a single public event perceived as positive or negative by different individuals while controlling for other event properties. These studies include investigations of memories for major sporting event outcomes as a function of team affiliation (Botzung, Rubin, Miles, Cabeza, & LaBar, 2010;Breslin & Safer, 2011;Kensinger & Schacter, 2006) as well as other studies of highly significant political events, such as the 2008 American presidential election (Holland & Kensinger, 2012) and the UK's 2016 EU referendum ('Brexit'; Raw, Rorke, ELLIS, Murayama, & Sakaki, 2020). These investigations have yielded mixed evidence regarding the effect of valence on autobiographical memory: while some studies report higher memory consistency for negative events and higher memory confidence for positive events (Holland & Kensinger, 2012;Kensinger & Schacter, 2006;Raw et al, 2020), other studies have reported greater memory consistency and confidence for positive versus negative events (Botzung et al, 2010;Breslin & Safer, 2011) .…”

“…These studies include investigations of memories for major sporting event outcomes as a function of team affiliation (Botzung, Rubin, Miles, Cabeza, & LaBar, 2010;Breslin & Safer, 2011;Kensinger & Schacter, 2006) as well as other studies of highly significant political events, such as the 2008 American presidential election (Holland & Kensinger, 2012) and the UK's 2016 EU referendum ('Brexit'; Raw, Rorke, ELLIS, Murayama, & Sakaki, 2020). These investigations have yielded mixed evidence regarding the effect of valence on autobiographical memory: while some studies report higher memory consistency for negative events and higher memory confidence for positive events (Holland & Kensinger, 2012;Kensinger & Schacter, 2006;Raw et al, 2020), other studies have reported greater memory consistency and confidence for positive versus negative events (Botzung et al, 2010;Breslin & Safer, 2011) . Notably, two of these studies investigated valence effects in memories for a common event (the 2004 baseball playoff final between Boston Red Sox and New York Yankees (Breslin & Safer, 2011;Kensinger & Schacter, 2006)) and revealed differing results as a function of retrieval interval, with higher memory consistency for the game as a negative event when retrieved at ~6 months (Kensinger & Schacter, 2006), and higher memory consistency for the game as a positive event when retrieved at ~4 years, potentially because of increased memory rehearsal, in a separate participant group (Breslin & Safer, 2011).…”