The Least Likely of Times
Image credits: @neonbrand
PSYC 307: Cultural Psychology with Dr. Benjamin Cheung
Research papers aren't the easiest to read. They're written in complex language, filled with a hearty sprinkling of statistics, and often can seem quite confusing. This summary was created in an effort to summarize the main research in a way that's accessible to a wider audience. It lays out the process with maximum transparency, allowing the reader to form their own conclusions after critical evaluation.
In their paper, Morewedge et al. (2005) investigate the hypothesis that people tend to recall atypical instances that are unrepresentative of their general experiences, and relying on this to forecast future affective behaviour leads to overestimating the impact it can potentially have. This overestimation is fueled by the high availability of atypical aspects of events as well as a lack of clear understanding of the exact factors influencing the situation. They hypothesized that this effect (termed the ‘impact bias’) can be reduced through two mechanisms: by the conscious awareness of the atypical nature of the memory recalled and by recalling several distinct memories of similar events (as opposed to just one). Awareness of the atypicality of the event was to alert the participants of the ungeneralizable nature of the memory while recalling several instances was to induce regression to the mean affective state in a situation. They carried these out through three studies, with three conditions each. The results of each were then compared with the others.
The first two studies were aimed at invoking negative and positive memories of experiences respectively. In each study, participants were split into three conditions: free recallers (asked to recall how they felt in the situation), biased recallers (asked specifically for the memory of an extreme feeling in the situation), and varied recallers (asked for four distinct memories of similar situations). All participants were then asked to forecast their future affective states in a similar situation.
The third study was designed to answer two possible limitations posed by the experimental methods: the impact of conscious recalling of memories and the impact of rating these memories. To address this, the varied recallers condition was replaced with non-recallers, participants who were simply asked to rate their affective state in a given situation without being prompted to recall or rate an existing memory. The free and biased recallers’ conditions were retained.
All studies’ results were compliant with the researchers’ predictions; free recallers, across studies, reported extreme scores of affective behaviour from the instances recalled and while biased recallers matched this score on the memories, they differed greatly in forecast scores (free recallers forecasted more extreme scores than biased recallers). Varied recallers reported a much lower recalled score for recalled memories, and their forecast scores matched those of the biased recallers. The forecast scores of non-recallers from the third study appear to match those of the free-recallers. This suggests that non-recallers, free recallers and biased recallers drew on atypical instances from their memory when making their forecasts (though biased recallers adjusted those forecasts more than the non- and free recallers did), while the adjustment was in-built into the process the varied recallers underwent.
The paper concludes by outlining the several (adverse) consequences of this tendency in influencing behaviour in ways that lead to more than mere misprediction, reminding the readers that this can be easily averted through the simple acts of reminding oneself of other similar instances and/or raising awareness to the atypical nature of a specific memory. This, they state, could help bring down the effect of the impact bias and lead to more informed and clear decision-making.
Morewedge, C. K., Gilbert, D. T., & Wilson, T. D. (2005). The least likely of times: How remembering the past biases forecasts of the future. Psychological Science, 16(8), 626-630.
