Abstract
In the multiple-object tracking paradigm (MOT), subjects are asked to covertly track several target objects simultaneously as they move randomly amongst identical distractors. Our previous electrophysiological experiments have shown that the sustained contralateral delay activity (CDA) provides an online neural measure of the number of objects being actively tracked (Drew & Vogel, VSS 2006). Recently, Wolfe, Place, & Horowitz (PBR 2007) have demonstrated that observers can dynamically change which items are being tracked during the course of the trial. Here we utilized the CDA to study the neural mechanisms underlying this process.
Observers fixated centrally while performing a lateralized MOT task. They tracked a changing subset of items in one hemifield while ignoring similar stimuli and events in the opposite hemifield. Observers initially tracked either one or three targets. On 50% of trials, a switch cue instructed observers to either add or subtract targets. This yielded four types of trials. On add trials, an observer tracking one target would end up tracking three, while on subtract trials, an observer tracking three targets would drop to one. On constant-1 and constant-3 trials, the target set was unchanged. Behavioral data indicated that observers were following instructions: performance on add trials matched performance on constant-3 trials, while performance on subtract trials matched that on constant-1 trials.
On add trials, CDA amplitude increased from constant-1 to constant-3 levels within roughly 500 ms post-cue. In contrast, on subtract trials, CDA amplitude took over 1000 ms post-cue to decrease from constant-3 to constant-1 levels. This may reflect a form of attentional hysteresis, as subjects appear to continue tracking items after the switch cue despite the attentional cost of tracking irrelevant items. The CDA provides a method for measuring rapid dynamic changes in tracking load.