Abstract
Both laboratory and real world studies show that humans spontaneously follow the gaze of others. However, gaze following in the real world occurs surprisingly infrequently, only in about a third of available instances. Here we assessed whether the results from a typical laboratory based measure reflected consistent orienting of attention on most trials or an effect of averaging large performance differences on a handful of trials. To do so, we collected data from 25 participants who performed a gaze cuing procedure. Left or right gazing faces and a neutral face with closed eyes served as attentional cues. Participants detected peripheral targets, which occurred equally often at a left or right location. Data were analyzed in two ways. First, a standard group-based ANOVA replicated a wealth of past research showing overall reliable facilitation for gazed-at targets. Second, to address our questions, we calculated the proportion of gazed-at and not gazed-at trials that were significantly faster than baseline (i.e., closed eyes). We found that a greater proportion of gazed-at trials, i.e., 20% were faster than the baseline relative to 16% of not gazed-at trials. Taken together, these results suggest that performance in the gaze cuing task reflects neither consistent orienting throughout, nor the effect of averaging over several trials.
Meeting abstract presented at VSS 2017