Thu, 21 Aug 2008 14:40:41 BST CiteULike: mnq's Fencsik http://www.citeulike.org/user/mnq/author/Fencsik CiteULike.org en-gb Copyright © 2004-2008 citeulike.org http://www.citeulike.org/user/mnq/article/908678 <i>Psychon Bull Rev, Vol. 13, No. 3. (June 2006), pp. 516-523.</i><br /><br />We previously demonstrated that observers in multiple object tracking experiments can successfully track targets when all the objects simultaneously vanish for periods lasting several hundred milliseconds (Alvarez, Horowitz, Arsenio, Dimase, and Wolfe, 2005). How do observers do this? Since observers can track objects that move behind occluders (e.g., Scholl and Pylyshyn, 1999), they may treat a temporal gap as a case of complete occlusion. If so, performance should improve if occlusion cues (deletion and accretion) are provided and items disappear and reappear one by one (asynchronously), rather than simultaneously. However, we found better performance with simultaneous than with asynchronous disappearance (Experiment 1), whereas occlusion cues were detrimental (Experiment 2). We propose that observers tolerate a gap in tracking by storing the current task state when objects vanish and resuming tracking on the basis of that memory when the objects reappear (a task-switching account). How do we track invisible objects? TS Horowitz RS Birnkrant DE Fencsik L Tran JM Wolfe Psychon Bull Rev, Vol. 13, No. 3. (June 2006), pp. 516-523. 2006-10-21T05:55:51-00:00 2006 Psychon Bull Rev 1069-9384 13 3 516 523 mot