Event boundaries: Costs and benefits for memory

Abstract

Event boundaries are known to benefit memory at the boundary, a phenomenon explained by Event Segmentation Theory as the result of increased attention. However, Event Segmentation Theory offers no clear prediction about the fate of information immediately following the boundary. To address this theoretical ambiguity, we conducted five experiments (N = 171) to examine the effects of event boundaries on post-boundary item memory and source memory. Using a continuous report paradigm and the Target Confusability Competition model, we found a surprising memory cost (higher absolute response errors and weaker memory strength) for post-boundary items than for non-post-boundary items (Experiment 1). This cost effect persisted when task difficulty decreased (Experiment 2) and when items varied in category within the same event (Experiment 3). Experiment 4 found that this post-boundary item memory cost was accompanied by a benefit in source memory. Such effects reflect the allocation of attention, as evidenced by a reduced proportion of fixation time on post-boundary items and increased fixation on post-boundary sources relative to their non-post-boundary counterparts (Experiment 5). Our results first reveal a post-boundary item memory cost and a source memory benefit, suggesting that the attention enhancement at event boundaries extends to the post-boundary context, reducing resources for post-boundary items. Our findings provide new insights into Event Segmentation Theory.