Proceedings of the Annual Meeting of the Cognitive Science Society

Research in event cognition highlights the crucial role of event segmentation in shaping perceptions and memories. Anticipation of event boundaries is influenced by characteristic duration, often assumed to follow normal distributions in daily events. This study replicates recent investigations into event duration using a nightly segmentation approach with continuously captured daily images. Forty-one participants collected images over fourteen days, segmenting them into events. Event durations for various activities were modelled using truncated normal, exponential and gamma models. Our findings align with prior research in event distribution, revealing that overall, an exponential or gamma distribution provides a superior fit compared to a truncated normal distribution. This suggests that when daily events are studied in an ecological context at a fundamental level, most of them have little sign of a typical duration. Consequently, duration estimation is unlikely to play a large role in anticipating event boundaries.