Abstract
During visual search, a “pop-out” stimulus can capture attention, and either interfere with or facilitate search, depending on whether it is a target or distracter. Horstmann [1] demonstrated that these influences depended on subjects' expectations for the amount of color variation in the scene. We performed a similar experiment, but additionally quantitatively manipulated the variance of colored patches on which letters were superimposed. We hypothesized that the influence of the pop-out stimulus should be greatest when the variance of color experience was small, and consequently more deviant with respect to the experienced distribution. Eye movements were recorded while subjects searched for one of two target letters in a circular array of letters placed on colored tiles. Subjects were given experience in one of two conditions, where the variance of the background color tiles was either high or low. On a critical trial, subjects in all conditions were presented with an identical array containing a color pop-out stimulus. For critical and post-critical trials, the singleton was either the target letter or a distracter letter, depending on condition. Across all conditions, during experience trials subjects' reaction times were between 1–2s and they took 3–5 fixations to locate the target. Regardless of prior experience, during critical and post-critical trials subjects had faster reaction times, 0.6–1s, and took 1–3 fixations to find the pop-out target. Crucially, it took ∼5 trials to asymptote in performance for finding the pop-out target. We found no evidence for gaze capture or performance impairment in the pop-out distractor condition. Overall we found little evidence for strong gaze capture in any condition, instead it appears that subjects must learn the relevance of a pop-out feature to guide search.
Horstmann G.; JEP:HPP (2005). Oct;31:1039–60.
This work was supported by NIH grants EY05729 and RR09283.