Abstract
Introduction: Cueing by eyes and head direction strongly orients an observer's overt and covert attention (Bayliss and Tipper, 2007), even in an anticipatory manner (Joyce et al., 2016, Han and Eckstein, 2023). However, the temporal dynamics by which gaze cueing influences the processing of peripheral information for perceptual decisions are not well understood. To investigate how observers use a foveally-viewed dynamic gazer to orient attention to peripheral information, we conducted a temporal reverse correlation study. Methods: Five observers covertly detected the presence of a bright spatial Gaussian target (SD = 0.93°) and localized it among dimmer distractors in one of two locations to the left and right (10.5° eccentricity) of a central gazer. Target and distractor luminances were perturbed with independent Gaussian noise every 25 ms. In 20%, cue-neutral, trials the central gazer’s head did not cue either location, in 64%, normal-cue, trials the head turned for 375 ms to cue the left or right location. In the remaining 16%, reverse-cue, trials the head turned (for 225 ms) toward one location, then reversed its direction (for 400 ms) to cue the opposite location. The cue validity was 80%. Results: An inverse cueing effect was found for three observers in the reverse-cue condition (hit valid = 0.67; hit invalid = 0.73). For those three observers, localization in false-alarm trials was more frequent at cued than at uncued locations in the normal-cue condition (p=0.02) but opposite in the reverse-cue condition. Temporal reverse correlation of noise values for false-alarm trials revealed a time window for peripheral information integration that peaked 100-300 ms after the gazer’s head movement initiation, which may explain the observers’ inverse cue effect in the reverse-cue condition. Conclusion: Together, the findings suggest that observers anticipate the gazer’s head direction to orient covert attention, which can sometimes result in task-suboptimal behavior.