Automatic capture of attention: A test of the transient channel hypothesis.

摘要

A number of studies have examined the ability of a stimulus to capture attention automatically. Motion and onset of a stimulus have been shown to be effective in producing "pop-out" effects in visual search, and onset has also been shown to play a unique role in the capture of attention. Physiological evidence has shown that these stimuli also are effective in stimulating the transient visual pathway which has led to the hypothesis that activation of this pathway produces automatic attentional capture. However, recent research suggests that the ability of motion and onset to capture attention is due to the typical role of motion and onset in the appearance of new perceptual objects (Hillstrom and Yantis, 1994, Perception and Psychophysics, 399-411; Yantis and Hillstrom, 1994, Journal of Experimental Psychology: Human Perception and Performance, 95-107). This dissertation uses improved methodologies to examine the hypothesis that the transient visual channel mediates the automatic capture of attention. The selective responding of the transient channel to motion and hue is exploited to test the hypothesis. Experiments 1 and 2 address the specific mechanisms by which attention affects processing of stimuli when captured automatically by attention. The results support a model in which automatic attentional capture primarily produces a reduction in decision noise, but may also lead to resource allocation under some conditions. Experiments 3 and 4 specifically tested the transient channel hypothesis. Experiment 3 exploited the low sensitivity of the transient channel to isoluminant changes in color and showed that equiluminant cues can automatically capture attention. This suggests that the sustained channel can also mediate attentional capture, and that, consistent with Yantis and Hillstrom (1994, Journal of Experimental Psychology: Human Perception and Performance, 95-107), a luminance onset is not necessary for attentional capture. Experiment 4 tested the ability of motion alone to capture attention, and showed that an isolated moving stimulus automatically captured attention. Thus, the activation of the transient visual channel appears to be sufficient to capture attention. Together these results suggest that the automatic capture of attention is mediated by multiple pathways.