Examining the influence of a spatially irrelevant working memory load on attentional allocation.

The present study examined the influence of holding task-relevant gaze cues in working memory during a target detection task. Gaze cues shift attention in gaze-consistent directions, even when they are irrelevant to a primary detection task. It is unclear, however, whether gaze cues need to be perceived online to elicit these effects, or how these effects may be moderated if the gaze cues are relevant to a secondary task. In Experiment 1, participants encoded a face for a subsequent memory task, after which they performed an unrelated target detection task. Critically, gaze direction was irrelevant to the target detection task, but memory for the perceived face was tested at trial conclusion. Surprisingly, participants exhibited inhibition-of-return (IOR) and not facilitation, with slower response times for the gazed-at location. In Experiments 2, presentation duration and cue-target stimulus-onset asynchrony were manipulated and we continued to observe IOR with no early facilitation. Experiment 3 revealed facilitation but not IOR when the memory task was removed; Experiment 4 also revealed facilitation when the gaze cue memory task was replaced with arrows cues. The present experiments provide an important dissociation between perceiving cues online versus holding them in memory as it relates to attentional allocation.