Fig. 4: The relationship between anticipatory gaze and explicit memory reports.
a Experiment 1 (N = 34): gaze in the second viewing of movies explicitly reported as “seen” (green) showed greater proximity to the event location compared with the first viewing (blue) for both recognized and unrecognized movies, signifying memory retrieval. b The MEGA score quantifies the change in GAD between the first and second viewings, calculated using the formula above. This MEGA score is higher for movies that were explicitly remembered versus not remembered. c Procedure of the Experiment 2 (N = 30): Participants completed a similar task except surprising events were omitted from the second viewing and followed by an extensive explicit memory report. Memory was assessed to determine what exactly participants remember. d Replication of the anticipatory gaze effect: consistent with Experiment 1, anticipatory gaze appears in the second viewing (green) and not in the first viewing (blue) across all naïve 30 participants. e Categorization of explicit report. Participants did (i) not recognize the movie clip at all (orange), (ii) recognize the context alone (violet), or (iii) recognize the context and recollected the event (green). f The MEGA score is higher for movies in which participants remembered the event in full, compared to those with only scenery memory or completely forgotten movies. There is no significant difference between movies with scenery recognition and movies that were not recognized. g Pupil size variation and memory: Examines the decrease in pre-event pupil size from first to second viewing, in relation to scene recognition and event recollection, highlighting smaller changes (1st vs 2nd viewing) in unrecognized cases. h The MEGA score correlates with participants’ event recollection in the free recall. Every dot represents a participant. i If we remove the events completely from the experiment, anticipation diminishes and gaze in the second viewing (green) shows comparable proximity to the event location compared with the first viewing (blue). j MEGA score for the control experiment without any events, in comparison to MEGA score of Experiment 1 (same procedure for both experiments), revealing chance-level MEGA scores for experiments with movies without events. Plots represent the median (bold horizontal line), 95% confidence intervals (whiskers), density plots, and subject averages (dot plots). *p < 0.05, **p < 0,01, ***p < 0,001.
