Abstract
Serial dependence (SD) is a bias in reporting stimulus features as more similar to the recently presented stimuli than they actually are. Several studies suggest that SD increases when similar features are attended at nearby locations, suggesting that selective and spatial attention plays a role in SD biases. Selective attention typically involves both the selection of targets and the suppression of distracting information. For instance, spatial locations frequently associated with a distractor can be implicitly suppressed. Here we investigated whether stimuli occurring at spatially suppressed locations cause SD. Participants performed a sequential dual-task, with visual search trials followed by an orientation adjustment task. In the visual search task, observers determined the location of the open side of a Landolt C presented in one of four locations. On half of the trials, a distractor (an oriented Gabor) was presented in one of the remaining locations. The distractor location was kept constant or randomly varied in two separate conditions. In the orientation adjustment task, a test Gabor was presented at one of the locations, including the distractor location on distractor present trials. The adjustment task involved reproducing the Gabor orientation by rotating a response bar. We found significant attentional capture in the search task (slower reaction times) due to the distractor, independently of whether its location was constant or variable. SD was quantified as the deviation of adjustment responses toward or away from previous stimuli. When SD was measured for the test Gabor in the preceding trial, we replicated the classical attractive bias. SD for the distractor Gabor, however, showed a repulsive bias. Our results demonstrate the differential SD effects of the attended and ignored features. However, the lack of a spatial suppression indicated by search times presents a puzzle for the mechanism behind the SD bias induced by the distracting information.