Abstract
Previously, we showed that observers could pursue a multiple-object tracking (MOT) stimulus (Pylyshyn & Storm, 1988) while simultaneously keeping track of a subset of the local elements without cost to either task (Watamaniuk et al., 2009). Moreover, pursuit in this paradigm is guided not by memory, but by constant integration of component motions (Jin et al., 2010). Here we show that the quality of pursuit is dependent upon the number of elements in the stimulus, and further, that the motion signal driving pursuit is integrated across both MOT targets and distractors. The stimulus comprised 4, 8, or 10 dots that moved randomly within a virtual 10x10 deg containment region, and observers attentively tracked 2, 4 or 5 target dots, respectively. During the MOT task, the stimulus either translated from left to right at 7 deg/sec, or remained at the center of the screen without translational motion. In a subset of trials, observers also pursued the stimulus without performing the MOT task. We found that performance on the MOT task was the same with and without pursuit. When the number of MOT targets was increased, performance on the task was poorer, but pursuit gain improved. Pursuit gain was also no different with and without the task, indicating that observers were not pursuing only the targets or only the distractors. The results suggest that motion integration during pursuit is a relatively inattentive process and that attentively tracked elements are not excluded from the integration.
Meeting abstract presented at VSS 2013