Distractor avoidance is achieved by active control rather than passive exclusion

When navigating the visual world, we have considerable control over what stimuli to select for deeper processing, and we can often resist capture by physically salient but task-irrelevant distractors. A debate has emerged regarding whether distractor avoidance is accomplished by active goal-driven attentional control, or by passive exclusion of distractors from the narrow attentional focus of a highly serial search, as the attentional window account claims. In the present study, in order to explore the issue, we measured the extent to which attentional avoidance of salient distractors is affected by the seriality of the search. In three experiments, we invoked different levels of seriality by separately manipulating the similarity among search elements, the stimulus pattern complexity, and the conjunctive features that defined the target. To assess attentional allocation, probe trials were interleaved with the search trials. While the results showed significant avoidance of color singleton distractors in all experiments, the strength of the avoidance effect, as measured by both search latency and probe report, did not differ as a function of the seriality of the search. An additional control experiment verified the item-by-item serial search processes invoked by most of our methods. The results are inconsistent with the passive filtering proposed by the attentional window account and instead imply that active top-down control is a more likely explanation for distractor avoidance. OPEN PRACTICES STATEMENT: Trial-level data and analysis code for all experiments have been made publicly available online ( https://osf.io/njvu2 ), and none of the experiments was preregistered.