Abstract
Studies on roles of memory in visual search have been focusing on memory for distractors or interference of memory tasks on visual search. This study investigated roles of memory for a target-relevant feature in visual search. Does preview of a target-relevant feature bound to the target location facilitate visual search? I devised a spatiotemporal visual search task to search for a target defined by a spatiotemporal change of its feature. Eight rectangular color bars moved horizontally, then they were partially occluded by squares, and the occluded parts reappear with a color either the same as or different from the visible part. Observers were precued a subset (from 1 to 4) of the bars, and judged the presence of a color changing target among the precued set. In the complete condition, where color information is available for the entire trial, observers could search for a target by comparison of two percepts, memory and percept, or both. To evaluate the role of memory and perception, two additional conditions were used. In the memory condition, color information of visible part is eliminated at the time of bars' reappearance, forcing observers the search by memory-percept comparisons. In the perception condition, bars' color information is unavailable until the bars' reappearance, forcing observers the search by percept-percept comparisons. The threshold duration of bar reappearance increased as the set size, suggesting a spatiotemporal visual search. There was a significant threshold elevation in the memory condition for the set size of 3 or more, but there was no difference among the three conditions for the set size of 2 or 1, suggesting that search efficiencies by memory-percept and percept-percept comparisons are equivalent within the set size of 2, but that simultaneous availability of memory and percept does not improve search efficiency. Preview of colors at cued locations for more than 1000 ms does not facilitate spatiotemporal visual search.