How does the peak-end rule smell? Tracing hedonic experience with odours

Abstract

The peak-end rule predicts that retrospective evaluations of affective events heavily depend on their most intense and last moment and imply duration neglect. It was originally proposed for negative experiences such as painful medical procedures. It is unclear, however, to what degree it also applies to positive experiences. Previously, rigorous comparisons between the two domains were limited due to the use of qualitatively different stimuli. Hence, it is not clear if the peak-end rule holds for short positive and negative experiences alike. To address these questions in a genuinely emotional domain, we conducted two experiments (n = 48 each) in which we used odours as stimuli. Participants repeatedly evaluated continuous odour sequences delivered into their noses via an olfactometer. The sequences differed in valence (positive vs. negative), length (36 vs. 72 s), and trajectory (increasing, decreasing, U-shaped, and inverse U-shaped). Results provide evidence for the peak-end rule for both positive and negative experiences alike. Results further show an overweighting of intense negative experiences for sequences that contain both pleasant and unpleasant episodes but provide little evidence for an effect of the trajectory manipulation.

The Emotion Filmmaker: Temporal Memory, Time-Emotion Integration, and Affective Style

“…See Table 1 for complete model results). This replicates and extends a plethora of prior work (Asutay et al, 2020(Asutay et al, , 2021Chajut et al, 2014;Do et al, 2008;Finn, 2010;Hsu et al, 2018;Kahneman et al, 1993;Redelmeier & Kahneman, 1996;Scheibehenne & Coppin, 2020) pointing to the final moments of emotional episodes as critical predictors of retrospective emotion judgments (above and beyond the average emotion experienced during the episode).…”

“…In the emotional sequences task, participants viewed 72 emotional-movie sequences with varying durations (range = 7-32s, M = 18s) and sequence lengths (i.e., 4, 5, 6, or 7 images). Following prior work Scheibehenne & Coppin, 2020), participants watched emotional movies while providing continuous ratings of subjective emotion on a scale from "very unpleasant" to "very pleasant" by moving a mouse/trackpad across a slider. Following each sequence, participants reported a retrospective emotion evaluation: "Overall, how did the sequence make you feel?"…”

“…The following parameters were computed from momentary emotion ratings for each sequence: positive peak (the highest rating above the neutral point, i.e., 5 on the 1-9 scale), negative peak (the lowest rating below the neutral point) 1 , end (the average rating in the final 2s of the sequence), average (the overall average rating), and duration-weighted average (i.e., the time weighted-average, where the weight of each rating is determined by its duration relative to the total sequence duration; see Fig. 1b) Scheibehenne & Coppin, 2020).…”

“…The peak effect: Negative and positive peaks. Given recent work indicating that "negative peaks" may disproportionally contribute to retrospective emotion evaluations (Scheibehenne & Coppin, 2020), we investigated whether negative vs. positive peaks experienced during emotional-movie viewing differentially predicted retrospective emotion ratings. We found that the inclusion of positive and negative peaks significantly…”

“…Instead, the most intense ('peak') and final ('end') emotional fluctuations are typically weighted more heavily than other timepoints (i.e., the peak-end effect).Moreover, the temporal duration of emotional experiences plays a seemingly minor role in sculpting retrospective emotion-for instance, prolonging aversive emotional episodes does not necessarily yield corresponding increases in the magnitude of remembered negative emotion (i.e., suggesting duration neglect; . These biases have often been replicated (peak-end effect: Asutay et al, 2020Asutay et al, , 2021Chajut et al, 2014;Do et al, 2008;Finn, 2010;Hsu et al, 2018;Kahneman et al, 1993;Redelmeier & Kahneman, 1996;Scheibehenne & Coppin, 2020; duration neglect: Redelmeier & Kahneman, 1996;Scheibehenne & Coppin, 2020, but see also: Ben-Zeev et al, 2009;Miron-Shatz, 2009).…”