Efficient tuning of attention to narrow and broad ranges of task-relevant feature values

Abstract

Feature-based attention is the ability to select relevant information on the basis of visual features, such as a particular color or motion direction. In contrast to spatial attention, where the attentional focus has been shown to be flexibly adjustable to select small or large regions in space, it is unclear whether feature-based attention can be efficiently tuned to different feature ranges. Here, we establish that the focus of feature-based attention can be adjusted more broadly or narrowly to select currently relevant features. Participants attended to a set of target-colored dots among distractor dots to detect brief decreases in luminance (Experiments 1a, 1b, 2) or bursts of coherent motion (Experiments 3a, 3b, 4). To vary the size of the attentional focus, we manipulated the range of colors that the target dots spanned and found that while participants' performance decreased with larger feature ranges to select, it remained at a relatively high level even at the largest color range, suggesting that broadening the focus of feature-based attention comes only at a small cost and that large feature ranges can be selected relatively efficiently at once. Overall, our findings argue against the idea that feature-based attention is limited to a single feature value at a time and demonstrate that selecting large swaths of feature space is surprisingly efficient. Broadly, these results are consistent with accounts that propose a flexible and generalized set of attentional mechanisms that act across both spatial and feature-based domains.