Implicit learning modulates attention capture: evidence from an item-specific proportion congruency manipulation

Abstract

A host of research has now shown that our explicit goals and intentions can, in large part, overcome the capture of visual attention by objects that differ from their surroundings in terms of size, shape, or color. Surprisingly however, there is little evidence for the role of implicit learning in mitigating capture effects despite the fact that such learning has been shown to strongly affect behavior in a host of other performance domains. Here, we employ a modified attention capture paradigm, based on the work of Theeuwes (1991, 1992), in which participants must search for an odd-shaped target amongst homogeneous distracters. On each trial, there is also a salient, but irrelevant odd-colored distracter. Across the experiments reported, we intermix two search contexts: for one set of distracters (e.g., squares) the shape singleton and color singleton coincide on a majority of trials (high proportion congruent condition), whereas for the other set of distracters (e.g., circles) the shape and color singletons are highly unlikely to coincide (low proportion congruent condition). Crucially, we find that observers learn to allow the capture of attention by the salient distracter to a greater extent in the high, compared to the low proportion congruent condition, albeit only when search is sufficiently difficult. Moreover, this effect of prior experience on search behavior occurs in the absence of awareness of our proportion manipulation. We argue that low-level properties of the search displays recruit representations of prior experience in a rapid, flexible, and implicit manner.

Context-dependent control of attention capture: Evidence from proportion congruent effects.

“…Their experiments involved designs that were different from our own, and we view our present efforts as an opportunity to supply converging evidence and draw in theoretical and methodological considerations from the proportion congruent literature that may be valuable for understanding evidence of contextual control in the attention-capture domain. Finally, our efforts to draw connections between the proportion congruent and attention-capture literatures are spread between this article and a companion article (Thomson, Willoughby, & Milliken, 2014) exploring more detailed aspects of the general methods presented here.…”

Cue predictability does not modulate bottom-up attentional capture

“…As a result, ignoring or suppressing the uninformative cues may not have been required in order to perform well. Indeed, recent studies [22,40] showed that attentional capture was only suppressed if task requirements were such that capture would interfere with target processing. Moreover, a study in macaques showed that modulations of V1 attention responses were larger for tasks that were more difficult [12].…”

Stimulus predictability does not modulate bottom-up attentional capture

“…As a result, ignoring or suppressing the uninformative cues may not have been required 21 in order to perform well. Indeed, recent studies [22,41] showed that attentional capture was only suppressed if task requirements were such that capture would interfere with target processing. Moreover, a study in macaques showed that modulations of V1 attention responses was larger for tasks that were more difficult [12].…”