Duration Neglect

Duration Neglect in the Evaluation of Experiences

The evaluation of experiences that extend over time—whether a pleasant film, a horrific noise, a painful medical procedure, or even the forecast of a future life change—reveals a fundamental discrepancy between the logic of the experiencing self and the operations of the remembering self. While the experiencing self moves through a series of moments, each with its own hedonic value, the mind forms retrospective (and prospective) judgments by substituting a few critical moments for the whole. This leads to a systematic disregard of duration, a phenomenon known as duration neglect.

Experimental studies of global evaluation illustrate this clearly. Participants continuously or intermittently report their affective state during an episode—using, for example, a scale of momentary pain or pleasure (the perspective of the experiencing self). Afterwards, they provide a global evaluation of “the total discomfort” or “the overall pleasantness” of the entire episode (the perspective of the remembering self). Normatively, the global evaluation of an extensional attribute should obey temporal monotonicity: adding an extra period of discomfort to an unpleasant episode can only make it worse overall. Under certain conditions, the total utility of an episode ought to equal the temporal integral of instantaneous utilities—a hedonimeter total. In practice, however, the remembering self does not sum moments. Instead, it relies on a “snapshot model.”

Fredrickson and Kahneman (1993) proposed that retrospective evaluations are generated through two acts of substitution: the episode is represented by a prototypical moment, and the affective value of that moment then stands in for the entire experience. The central finding is the Peak/End rule: global evaluations are accurately predicted by a simple average of the most intense moment (Peak) and the final moment (End) of the episode. Duration contributed little or nothing to these judgments. In an early study of plotless films (e.g., fish swimming in coral reefs, pigs being beaten to death), the correlation between retrospective evaluations and the films’ length was negligible. Fredrickson and Kahneman labeled this duration neglect, categorizing it as a form of extension neglect—akin to the neglect of base-rate, scope, or sample size—caused by a prototype heuristic.

The pattern has been confirmed repeatedly. Redelmeier and Kahneman (1996) recorded colonoscopy patients’ pain ratings every 60 seconds. As shown in Figure 2.4, patient A’s procedure ended at a moment of relatively intense pain; patient B endured a longer procedure overall, but the Peak/End average was more favorable. Consistent with the Peak/End rule, patient A evaluated the experience more negatively and was more inclined to choose an alternative (barium enema) over a repeat colonoscopy. Individual differences in duration—which ranged from 4 to 69 minutes—did not systematically predict the evaluations. Effects of duration remained small even in factorial studies. For example, when ischemic pain duration varied by a factor of 4 (Ariely, 1998), the Peak/End average accounted for 98% of the systematic variance in global evaluations; in a similar study of loud unpleasant sounds, it accounted for 88%. Contrary to the normative standard, duration and other determinants influenced evaluations in an additive fashion, with time playing a minor role.

The same neglect of duration appears not only in memory but also in decision-making and forecasts. In prospect theory, a state is represented by the transition to it—the anticipated intensity of the reaction to a change, such as winning a lottery or receiving a diagnosis—while the duration of the new state and the adaptation that inevitably follows are largely ignored. The focusing illusion exemplifies the same pattern: attention is fixed on a narrow slice of time, and what happens at other moments fades from view. In storytelling mode, the mind represents episodes by exceptional, defining moments; the cold-hand experiment and the story of Violetta illustrate how the climax and the conclusion override the actual length of an experience, whether in memory or in subsequent choices.

Duration neglect is an attentional phenomenon, not a conscious decision to disregard time. Participants in these studies were aware of the relative durations of their experiences, and in the abstract they considered duration important. As Fredrickson and Kahneman observed, “what comes most readily to mind in evaluating episodes are the salient moments of those episodes and the affect associated with those moments.” The error originates in the intuitive operations of System 1, which rapidly constructs stories around peaks and ends, and persists when the more deliberate System 2 fails to correct it. The same inattention underlies other manifestations of extension neglect: the neglect of base-rate in categorical prediction, the neglect of scope in willingness to pay, the neglect of sample size in evaluating evidence, and the neglect of probability of success in evaluating interventions. The mind, it seems, is well designed for stories but not for the processing of time.

These insights have deepened our understanding of happiness itself, revealing that the word carries multiple meanings—the moment-by-moment well-being of the experiencing self and the satisfaction of the remembering self—and that time is a dimension our cognitive machinery regularly neglects.