Forecast gap · what you predict vs what you feel
Duration Neglect
When people retrospectively evaluate past affective experiences, the duration of the episode has little or no independent effect on their global evaluation; instead, judgments are dominated by the peak intensity and the ending.
The forecast reaches for the dashed ceiling; reality settles at the solid line. The hatched shortfall is invisible in advance — you brace for the whole column and feel a fraction of it.
Why the forecast misses
Duration neglect is explained by the snapshot model of remembered utility. People construct global evaluations of extended episodes using a small number of prototypical moments rather than integrating over time. An unweighted average of peak intensity and end intensity predicts retrospective ratings well, while most moments receive zero weight. This violates temporal monotonicity: adding less intense moments to the end of an episode can improve its remembered evaluation. The cognitive system appears to represent extended experiences in terms of transitions and singular moments rather than area-under-the-curve.
Competing account — For simple, homogeneous aversive experiences (cold water, colonoscopy), the peak-end rule dominates. However, for complex heterogeneous experiences, average affect may be a better predictor than peaks and ends. Some research suggests the peak-end rule loses predictive power over time and that duration information may be partially recovered when experiences are very long or very short.
Forecasts on record
A majority of participants preferred to repeat the longer trial, which contained all the pain of the short trial plus 30 additional seconds of discomfort. Retrospective evaluations were predicted by peak-end averaging, not duration.
Kahneman, Fredrickson, Schreiber & Redelmeier · 1993 · n = 32 undergraduates
Effects of film duration on retrospective evaluations were small, entirely explained by changes in real-time affect, and further reduced when evaluations were made from memory. An unweighted combination of peak and end affect predicted evaluations well.
Fredrickson & Kahneman · 1993 · n = 128 (Study 1: n = 32; Study 2: n = 96)
Retrospective pain judgments were strongly correlated with peak pain intensity and pain during the final 3 minutes, but not with procedure duration. Lengthy procedures were not remembered as more aversive.
Redelmeier & Kahneman · 1996 · n = 287 patients (154 colonoscopy, 133 lithotripsy)
Patients in the modified (longer) group rated the experience as less unpleasant (4.4 vs 4.9, P=0.006) and were more likely to return for repeat colonoscopy (OR=1.41, P=0.038), demonstrating that a better ending overcame the greater total duration of pain.
Redelmeier, Katz & Kahneman · 2003 · n = 682 outpatients
Evidence — robust. The basic finding has been replicated across multiple domains: cold-water immersion (Kahneman et al., 1993), film clips (Fredrickson & Kahneman, 1993), colonoscopy and lithotripsy (Redelmeier & Kahneman, 1996), and effortful study (Finn, 2010). However, boundary conditions have been identified. The 2019 VR study by Schneider et al. found average affect better predicted memory for complex heterogeneous experiences, and the 2003 colonoscopy trial confirmed clinical applicability. The effect appears strongest for homogeneous, negatively valenced experiences.
Where the gap cost something
medicine · 2003
Colonoscopy procedure design at Wellesley Hospital
In a randomized trial of 682 colonoscopy patients at Wellesley Hospital in Toronto, patients assigned to a longer procedure with a gentler ending remembered less pain and were more likely to return for follow-up screening than those with the standard shorter procedure. This demonstrated that manipulating the ending could overcome duration neglect in a real clinical setting.
finance · Ongoing application
Vacation and travel experience evaluations
Vacationers consistently rate short trips with one or two peak memorable moments more favorably than longer trips with uniformly pleasant but unremarkable experiences. Tourism research applying the peak-end rule shows that the most intense positive moment and the departure experience disproportionately shape overall satisfaction ratings.
To forecast honestly
When evaluating past experiences, consciously consider duration alongside peak and end moments. Ask yourself: how long did it actually last? For medical procedures, request that providers minimize discomfort at the end, as this disproportionately shapes memory.
The 2003 colonoscopy trial showed that a simple procedural modification to create a gentler ending significantly improved patient memories and increased follow-up compliance, demonstrating that understanding duration neglect enables evidence-based improvements to experience design.
Catch it in the act
Watch for situations where you judge a past experience primarily by its most intense moment and how it ended, ignoring how long it actually lasted. For example, a short but terrible customer service encounter may be remembered as worse than a much longer but mildly inconvenient one that ended well.
Related forecasting errors
- Peak-End RuleparentDuration neglect is a core component of the peak-end rule, which states that retrospective evaluations are dominated by peak and end moments.
- Extension NeglectparentDuration neglect is a specific instance of the more general extension neglect, where people ignore the scope or size of an outcome when evaluating it.
- Affective Forecasting ErrorssiblingBoth involve mispredictions about emotional experiences, but duration neglect specifically concerns the retrospective evaluation of experienced episodes.