Abstract
What happens to the representations of items in visual working memory (VWM) when attention shifts to a single item? Here, we investigated the fate of these uncued items using a retroactive cuing paradigm. This paradigm presents attentional cues after the VWM maintenance duration but prior to memory test. The retro-cue was either neutral cue (no information), valid (the cued item was probed), or invalid (an uncued item was probed). Additionally, retrieval demands and key temporal parameters were manipulated. First, VWM was tested using recognition and recall probes to determine coarse accuracy and fidelity measures of the uncued representations. Second, the delay between the retro-cue and the memory probe varied to monitor the time course of these representations. In a series of experiments we found that when measured coarsely using recognition probes, VWM performance reveals no significantly detriment of the uncued item when compared to neutral trials even up to durations of 24 s. However, there is a consequence to being an uncued item demonstrated by the finer grained measure of recall. The accuracy of recall performance declines with longer delay durations indicating that the fidelity of the uncued representations is subject to significantly greater degradation than neutral or valid cue conditions. In summary, the retro-cue attracts internal attention to the representation of the cued item and preserves it without eliminating uncued items from VWM, but they are not protected and decay more rapidly.
Meeting abstract presented at VSS 2014