Timescales of Evidence Evaluation for Decision Making and Associated Confidence Judgments Are Adapted to Task Demands

Abstract

Decision making often involves choosing actions based on relevant evidence. This can benefit from focussing evidence evaluation on the timescale of greatest relevance based on the situation. Here, we use an auditory change detection task to determine how people adjust their timescale of evidence evaluation depending on task demands for detecting changes in their environment and assessing their internal confidence in those decisions. We confirm previous results that people adopt shorter timescales of evidence evaluation for detecting changes in contexts with shorter signal durations, while bolstering those results with model-free analyses not previously used and extending the results to the auditory domain. We also extend these results to show that in contexts with shorter signal durations, people also adopt correspondingly shorter timescales of evidence evaluation for assessing confidence in their decision about detecting a change. These results provide important insights into adaptability and flexible control of evidence evaluation for decision making.

Change point detection with multiple alternatives reveals parallel evaluation of the same stream of evidence along distinct timescales

“…2 ). Similar to results from related experiments 11 , 12 , 39 , the hit rate of the informed trials for increasing stimuli was lowest at the smallest stimulus delta (10 Hz), and increased until peaking at the largest delta (50 Hz), with a psychometric threshold (50% hit rate) at a delta of 30.4 Hz (CI: 29.2 to 31.5). This psychometric function shows that subjects were attending to the stimuli.…”

“…Our results also fit nicely with prior studies, from our lab and others, that show subjects can flexibly adjust their timescales of evidence evaluation on the same sensory stimuli based on task demands. For example, in the visual and auditory domains, humans adjust their timescale of evidence evaluation to match the distribution of signal durations and timing they experience 10 , 39 , 48 – 50 . Humans can also adjust their timescales of evidence evaluation when classifying the location of visual stimuli, using a shorter timescale in more volatile environments and a longer timescale in more stable environments 9 .…”

Choices change the temporal weighting of decision evidence

“…The present data add to a growing body of literature indicating that the dynamics of evidence weighting for decision-making is highly context dependent. Both the timescale and the temporal profile of this weighting applied to the same physical evidence are affected by a range of factors including the type of judgment ( 68 , 69 ), the type of information available in the stimulus ( 15 ), reliability of the sensory information ( 70 ), and the temporal statistics of the environment ( 3 , 5 , 71 ). We here extend this body of evidence, by showing that the temporal weighting profile, within a given individual and a given task, can be flipped by asking the participant for an intermittent choice in the middle of the evidence stream.…”

Hidden temporal structure of the ongoing task impacts detection strategy and is reflected in pupillary dynamics

“…Behavioral adaptation to the hidden temporal context led to earlier and increased numbers of false alarms. Interestingly, and because of the stochastic nature of our stimulus, false alarms are likely linked to local changes in the stimulus that resembled changes by chance (Harun et al 2020). In fact, participants often reported after the task that they perceived changes before false alarms.…”

“…Specifically, we hypothesized that internal expectation about the occurrence of target events at certain times dynamically increases sensory gain, which would lead subjects to respond to stochastic fluctuations in the ongoing stimulus, hence a concomitant increase in false alarm rate. In reaction-time designs (Harun et al 2020;Johnson et al 2017;Boubenec et al 2017), false alarms are defined by a button press prior to the 'go' signal. In physiology studies they are referred to as 'early' behavioral responses and have been used to link subjects' strategy and subthreshold stimuli changes (Orsolic et al 2019).…”