Abstract
Previous studies have demonstrated that both implicit spatial probabilities and repetition priming facilitate perceptual processing: targets in high probability or repeated locations are detected more rapidly and accurately. In this study we examined the relative contributions of spatial repetition priming and implicit spatial probabilities on target detection. Eight possible target locations were arranged in a circle, which subtended 4° of visual angle. On each trial, 4, 6, or 8 objects appeared within designated locations for 250 ms; the target was always present. Subjects reported the location of the target in an 8-alternative forced-choice localization task using a computer mouse. Targets appeared in the high probability location on 30% of trials and were evenly divided amongst the other 7 low probability locations on the remaining 70% of trials. Central fixation was monitored throughout the display period. The results showed that target detection performance (accuracy and RT) was at ceiling across all set sizes in the high probability location, but there was a linear decrease in performance across set sizes for targets in the low probability locations. There was also an interaction between spatial probabilities and repetitions: There was no effect of repetition for targets in the high probability location (perhaps due to ceiling effects in performance), but detection of repeated targets in low probability locations was actually worse. When subjects mislocalized the low probability targets, they tended to select a nearby location away from the high probability location. This suggests that spatial probabilities may serve as a "top-down" modulation of the prepotent "bottom-up" bias for targets in repeated locations. We speculated that spatial probability distributions do so by biasing limited-resource search mechanisms that control spatial attention.
Meeting abstract presented at VSS 2012