Attentional guidance by working memory overrides salience cues in visual search.

Abstract

Many factors influence visual search, including how much targets stand out (i.e., their visual salience) and whether they are currently relevant (i.e., Are they in working memory?). Although these are two known influences on search performance, it is unclear how they interact to guide attention. The present study explored this interplay by having participants hold an item in memory for a subsequent test while simultaneously conducting a multiple-target visual search. Importantly, the memory item could match one or neither of two targets from the search. In Experiment 1, when the memory item did not match either target, participants found a high-salience target first, demonstrating a baseline salience effect. This effect was exaggerated when a high-salience target was in working memory and completely reversed when a low-salience target was in memory, demonstrating a powerful influence of working memory guidance. Experiment 2 amplified the salience effect by including very high-salience, "pop-out"-like targets. Yet this salience effect was still attenuated when the memory item matched a less salient target. Experiment 3 confirmed these were memory-based effects and not priming. Collectively, these findings illustrate the influential role of working memory in guiding visual attention, even in the face of competing bottom-up salience cues.