Behavioural and Electrophysiological Investigations of Event Segmentation Using Ecological Stimuli

Abstract

Our daily lives are composed of many activities and experiences that form events. Segmenting these continuous, ongoing experiences into smaller units is known as event segmentation. Previously, ecological stimuli have been shown to increase subsequent memory at event boundaries during active event segmentation tasks. Our study intends to determine whether these effects were related to tasks (e.g., pressing the button to segment actively), or whether they were related to a feature of stimulus that produces memory effects, as well as the brain's ability to shape these events into memories. We conduct two experiments in this study: behavioural and electroencephalogram (EEG) experiments, using ecologically valid stimuli to observe event segmentation effects on memory. The first experiment consisted of 23 participants who watched a 30-minute POV video, followed by a memory recall task. This experiment was consistent with our hypotheses and previous research in showing that stimuli located on the boundary showed a higher degree of memory richness than stimuli located in the middle. The effects of navigational turns at event boundaries on memory richness was assessed, and significant differences were observed between turn and non-turn events. The second experiment involved 20 participants who watched the same video divided into smaller ten second segments while their EEG was recorded, followed by a memory recall test. Behavioural components demonstrated nonsignificant results for memory richness during boundary events and navigational turns at event boundaries. When theta power was compared between event boundary and middle conditions using repeated measures ANOVA, no significant interaction was found. A negative relationship was also observed between subsequent memory and mean theta as shown by regression analyses. Experimental 2 contradicted previous experiments and literature as videos were divided into ten second with no visual context, and the EEG results contradict literature due to a reduction in frontal midline theta (FMT) at event boundaries in episodic memories. This research may have had implications that affected our results, but it still provides us with the opportunity to continue our research and explore more into the area of event segmentation and memory in the future.