Independent effects of statistical learning and top-down attention

Ya Gao,Jan Theeuwes

doi:10.3758/s13414-020-02115-x

Ya Gao, Jan Theeuwes

Open Access

https://doi.org/10.3758/s13414-020-02115-x

Copy DOI

Abstract

It is well known that spatial attention can be directed in a top-down way to task-relevant locations in space. In addition, through visual statistical learning (VSL), attention can be biased towards relevant (target) locations and away from irrelevant (distractor) locations. The present study investigates the interaction between the explicit task-relevant, top-down attention and the lingering attentional biases due to VSL. We wanted to determine the contribution of each of these two processes to attentional selection. In the current study, participants performed a search task while keeping a location in spatial working memory. In Experiment 1, the target appeared more often in one location, and appeared less often in other location. In Experiment 2, a color singleton distractor was presented more often in location than in all other locations. The results show that when the search target matched the location that was kept in working memory, participants were much faster at responding to the search target than when it did not match, signifying top-down attentional selection. Independent of this top-down effect, we found a clear effect of VSL as responses were even faster when target (Experiment 1) or the distractor (Experiment 2) was presented at a more likely location in visual field. We conclude that attentional selection is driven by implicit biases due to statistical learning and by explicit top-down processing, each process individually and independently modulating the neural activity within the spatial priority map.

Highlights

It is important to be able to direct our attention to those events that are relevant to us and prevent distraction by events that are unimportant
There was a large main effect on response times (RTs) of the matched versus unmatched, F(1, 35) = 34.46, p < .001, ηp2 = .50, indicating that when the location that is kept in working memory matched the location of the target during search task, participants were much faster than when the location kept in working memory did not match the location of the search target
This indicates that our spatial working memory task did recruit spatial attention, and suggests that during the whole search task, attention was directed in a top-down fashion to a specific location in space

Summary

Introduction

It is important to be able to direct our attention to those events that are relevant to us and prevent distraction by events that are unimportant. A new framework was presented that considers three separate factors that affect attentional selection. In addition to top-down and bottom-up selection, it was argued that lingering biases of previous selection episodes (i.e., selection history) plays an important role in attentional selection (Awh, Belopolsky, & Theeuwes, 2012; Failing & Theeuwes, 2018; Theeuwes, 2018, 2019). Lingering biases, referred to as “selection history,” may drive attention selection towards particular objects that are neither part of the top-down set of the observer (i.e., objects that are irrelevant for the task) nor do they stand out from the environment to capture attention in a bottom-up way (Theeuwes, 2018)

Objectives

Methods

Results

Conclusion