Abstract
In two studies we used eye-tracking to investigate the ‘stare-in-the-crowd effect’–the easier detection of direct than averted gaze–and whether it is modulated by task demands. Stimuli consisted of four full characters (a target and 3 distractors) displayed side-by-side across the entire screen. In the first study (Location task), participants had to detect the location of either a direct- or an averted gaze target amongst opposite-gaze distractors. In the second study (Detection task), participants completed the classic stare-in-the-crowd task, detecting whether the gaze target was present or absent. In the Location task, RTs were faster for direct than averted gaze targets, but only for characters situated in the far peripheral visual fields. In the Detection task, no effect of gaze was found for RTs. In contrast, error rates were lower for detecting direct than averted gaze targets in the far peripheral visual fields for the Detection task, but not for the Location task. Thus, the stare-in-the-crowd effect was not fully replicated in either task. In both tasks, eye movements revealed a serial search for the target; the search was identical between direct and averted gaze conditions, reflecting a lack of specific strategy for direct gaze detection. However, in both studies onset times of first fixation were faster for direct than averted gaze targets in the far peripheral visual fields, mimicking the RT results of the Location task. These findings demonstrate that the detection asymmetry for direct gaze is highly dependent on target position and may be influenced by task demands. The results are discussed in the framework of the hypothetical subcortical face processing route supposed to also mediate mutual gaze detection.
This research is funded by CHIR, the CRC program to RJI, CFI and ORF.