Dynamics of retrospective timing: A big data approach

Area of Science:

• Cognitive Psychology
• Neuroscience
• Human Behavior

Background:

• Interval timing research predominantly uses prospective timing tasks, limiting understanding of real-world temporal judgments.
• Most everyday time estimations occur retrospectively, without prior awareness of the need to track duration.

Purpose of the Study:

• To investigate retrospective interval timing performance across a broad range of durations (5-90 minutes).
• To analyze temporal judgment accuracy and variability in a large participant sample (~24,500).
• To identify systematic biases in retrospective time estimation.

Main Methods:

• Participants estimated the duration of self-paced questionnaire completion tasks.
• Data from ~24,500 participants were analyzed for duration estimation accuracy and patterns.
• Findings were validated using a secondary dataset (Blursday) for robustness.

Main Results:

• Participants overestimated short durations (<15 min) and underestimated long durations (>15 min).
• Estimation accuracy was highest for durations around 15 minutes.
• Between-subject variability decreased with duration, stabilizing after 30 minutes.
• A significant portion of participants exhibited whole number bias, rounding estimates to multiples of 5 minutes.

Conclusions:

• Retrospective temporal judgments are subject to systematic over- and underestimation biases.
• Variability in retrospective timing is more pronounced for shorter durations (<30 min).
• This study provides the most extensive analysis of retrospective timing to date, covering a wide duration range and large sample size.