Attention is allocated to figures only under conditions of uncertainty

摘要

It is widely assumed that attention is automatically drawn to figures, which, if true, would have important implications how attention functions. Nelson and Palmer (2007) found support for this assumption using displays in which a central border divided a rectangle into two equal-area regions and the familiarity of an object sketched on one side of the central border served as the sole figural cue. They found faster, more accurate, responses to targets on the familiar side of the central border rather than the complementary side. If attention is automatically drawn to figures, similar results should be obtained under different presentation conditions. Nelson and Palmer used large displays (18?° x 20?°), and high spatiotemporal uncertainty regarding the target (12 potential locations, and 0, 150, 250, or 500-ms SOAs). Using smaller displays (5.3?° x 5.5?°) and lower spatiotemporal uncertainty (4 target locations, 80-ms SOA), we failed to observe evidence that attention is automatically drawn to figures, p 0.6, although we replicated their effects when we used large stimuli (14?° x 20?°) with variable SOAs and target locations like theirs, p .05 (VSS 2011). This year we tested whether subjects adopt a strategy of attending to the familiar configuration under conditions of high target spatiotemporal uncertainty by using the same large displays but lower spatiotemporal uncertainty (4 target locations and an 80-ms SOA). We failed to find any evidence of an attentional advantage for figures in either accuracy or RTs, ps 0.16. The difference in response patterns in the two experiments with different spatiotemporal uncertainty was significant, p .04. Our results indicate that attention is not automatically drawn to figures. Instead, it is strategically deployed to familiar figures under conditions of high spatiotemporal uncertainty. The question remains whether this strategy applies to all figural cues or only to familiarity.