Abstract
Tracking multiple objects with attention is easier if the objects are divided across the visual hemifields (e.g., one on in left hemifield, one in the right hemifield), as if each hemifield has a separate spotlight of attention (Alvarez & Cavanagh, 2005). If the two hemifields have separate tracking mechanisms, then it should be difficult to track targets that cross between the hemifields — a common occurrence for tracked objects encountered in everyday life. To test this possibility, observers saw pairs of identical black dots orbiting around a common point. There were always two orbiting pairs in diagonally opposite quadrants (e.g., one pair in the top-left, and one pair in the bottom-right). At the start of each trial, one dot within each pair flashed to identify it as a target to be tracked. Halfway through each trial, the target-distractor pairs simultaneously shifted vertically to a new quadrant within the same hemifield (within-hemifield movement), or horizontally to a new quadrant in the opposite hemifield (between-hemifield movement). Observers were better at identifying target dots following within-hemifield movement (M=77.2%, SD=12.1%) than following between-hemifield movement (M=66.7%, SD=13.4%; t(15)=3.9, p=.001). This difference was not attributable to difficulty tracking near the vertical midline separating the two hemifields (Experiment 2). Although a cost for between-hemifield movement was found when targets were presented to both hemifields, no cost was found when observers were required to track only a single target that moved between hemifields, even when controlling for task difficulty (Experiment 3). Together, these results suggest that hemifield-specific mechanisms struggle to transfer information when each is engaged with tracking its own target, but that these same mechanisms closely coordinate with one another when only a single target must be transferred. This coordination allows the successful representation of dynamic stimuli moving between the visual hemifields.
Meeting abstract presented at VSS 2016