Probabilistic retro-cues do not determine representational state in visual working memory

Abstract

To circumvent the capacity limitations of visual working memory (VWM), mechanisms exist that govern how information is represented in memory to ensure that the most relevant information guides behaviour. Retroactively cueing an item in VWM, for instance, affects both memory quality and representational state: A retro cue that indicates with 100% validity which item will later be probed enhances memory of that item, and 'activates' its representation such that it will bias selection towards perceptually similar inputs during visual search. However, when the retro-cue is less than 100% valid (i.e., probabilistic rather than deterministic) the effect of the cue on memory performance varies with manipulations to the proportion of valid trials. Here we investigated whether probabilistic and deterministic retro-cues also differ in their influence over representational state. Participants encoded two colored squares for a subsequent memory test. Following encoding, a spatial cue indicated to participants which item was most likely to be probed at the end of the trial. Cue validity was manipulated across blocks to be either deterministic (100% valid) or probabilistic (70% valid). On a subset of trials, no memory probe was presented and the trial instead ended with a visual search task in which a colored distractor—matching either the cued memory item, the non-cued item, or neither—was presented. As expected, in the deterministic retro-cue condition, the presence of a search distractor that matched the color of the cued item reliably slowed response times relative to trials with non-matching distractors. In the probabilistic retro-cue condition, however, search response times were comparable across all three distractor conditions, despite a reliable benefit to memory performance on valid relative to invalid trials. We suggest that, while probabilistic retro-cues improve memory of the cued item, they do not bias its representational state in VWM. Meeting abstract presented at VSS 2018