Abstract
Studies have shown that visual target identification is impaired by a temporally preceding salient distractor, reflecting attentional capture. The visual system, however, can adaptively reduce or even eliminate such a capture effect by endogenously establishing a top-down attentional set. The present study examined whether observers could establish multiple attentional sets to concurrently monitor two different colors. Observers identified a target letter in red or cyan among nontarget letters of other heterogeneous colors presented as a rapid serial visual stream. A distractor frame consisted of peripherally presented four pound signs with one of them in either the same color as the target of the current trial, as the other potential target color, or as an irrelevant color that could never be the target (green). If observers could establish multiple attentional sets to monitor two target colors, attentional capture should be observed only by the distractors of the two potential target colors. If observers cannot establish such bi-color sets and thus use a general set for any colors, the irrelevant color was also expected to capture attention. The results supported the former prediction indicating that observers could establish multiple attentional set for different colors. Subsequent experiments tested whether observers viewing dual rapid visual streams can (a) maintain such attentional control settings, (b) establish two different singleton-detection sets, and (c) establish a singleton-detection and a feature-search set separately for each stream. The results indicated that establishing (a) such two feature-search settings and (c) mixture of sets failed, but (b) two concurrent singleton–detection sets were sustainable during monitoring two streams. Importantly, unexpected but consistent priming effects were obtained, when attempting to establish a feature-search set concurrently with other set. These results suggest that a new framework is required to explain the split of multiple attentional control setting in rapid serial visual presentation.
Meeting abstract presented at VSS 2013