How suspense and surprise enhance subsequent memory: the case of the 2016 United States Presidential Election

ABSTRACT

We examined whether the retroactive enhancement effect – i.e., improved memory accuracy for event details occurring before a surprising moment – would be present in participants’ memory for details in their private lives following a surprising and suspenseful public event. To equate event type across participants, we selected when they first learned the outcome of the 2016 US Presidential Election. Participants self-divided into those who viewed the outcome as positive, negative, or neutral, while we experimentally divided them into those whose memory was assessed 6 or 12 months post-election. We assessed their accuracy for details surrounding the election and their phenomenological experience of learning the outcome, including emotional tension, our operationalisation of suspense. We found participants’ memory characteristics were strongly related to their level of tension and shock, irrespective of valence. We also observed uniformly high accuracy regarding details about the weather participants experienced in their ZIP codes on election day. While these results intimated about the presence of retroactive enhancement, Experiment 2 examined the effect more directly by comparing participants’ memory for the 2016 Election with two other politically-relevant events that provoked less tense reactions. The results revealed retroactive enhancement is dependent upon experiencing a surprising moment amidst a suspenseful event.

KEYWORDS:

• Retroactive enhancement effect
• event memory
• autobiographical memory
• emotional valence
• tension/suspense
• flashbulb memory

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We also asked participants in the Short Delay conditions questions about the precipitation and approximate temperature in their ZIP codes on the day before the election. However, as these questions did not produce any significant results with regard to either of the two emotion variables (Valence and Tension), they will not be discussed further.

Additional information

Funding

This research was supported by grants from the Danish National Research Foundation (DNRF89) and the Danish Council for Independent Research (DFF-6107-00050). Special thanks to Daniel Munkholm Møller for helping manage data collection via Amazon Mechanical Turk.