Abstract
We have previously demonstrated that observers in multielement tracking tasks can successfully track objects when they become invisible for up to 400 ms (Alvarez, Wolfe, Horowitz, & Arsenio (VSS 01)). Is this an extension of the ability to track occluded items, or a different mechanism that allows us to divert our attention from the tracking task in order to do something else for a moment? We hypothesized that if the ability to track invisible items is a way to cope with the disappearance of items in the world, then tracking should be easier if only one item disappears at a time. Furthermore, the presence of occluders at the time of disappearance/reappearance should improve performance, because occlusion is the most likely cause of disappearance of an item that remains present in the visual field. We measured tracking performance at 3 gap durations (?100, 300, & 500 ms). In Experiment 1, items disappeared and reappeared either all at once (synchronous) or one by one (asynchronous). There were two stimulus cue conditions: 1) items simply vanished and reappeared or 2) items were deleted by and accreted from modal surfaces. Contrary to the disappearing object hypothesis, synchronous disappearance yielded superior performance compared to asynchronous (F(1, 14) = 14.3, p < .005), and occlusion cues provided no advantage (F(1, 14) < 1). In Experiment 2, we used only synchronous disappearance, and found that occlusion cues actually harmed performance (F(1, 10) = 27.5, p < .0005). These results suggest that the ability to track invisible objects serves to allow the observer to set aside the tracking task for brief periods in order to direct attention to other priorities (e.g. an interpolated search task, c.f. Alvarez et al. '01). The ability does not appear to be due to a mechanism which merely copes with the periodic disappearance of objects from the field of view.
Supported by NIH grant MH065576 to TSH. LT sponsored by Project Success.