Abstract
Working memory (WM) performance can be facilitated by an informative cue presented during storage. Two recent studies suggest that this effect – termed a retrocue benefit – results from a recovery of information following the presentation of an informative cue. In one study, researchers used multivariate analyses of EEG data to reconstruct time-resolved representations of memoranda while participants performed a retrospectively cued WM task. In the absence of a retrospective cue, the information content of each reconstructed representation gradually decreased during storage. This decrease, however, was partially reversed by a 100% valid retrospective cue presented midway through the delay period. This reversal was interpreted as evidence that in some cases, participants can rely on other memory stores (e.g., long-term memory or “activity silent” memory) to supplement active WM representations. Here, we asked whether the degree of information recovery is contingent on the timing of a retrospective cue. Following earlier work, we recorded EEG while participants performed a retrospectively cued WM task. During neutral trials, participants were required to remember the spatial positions of two discs over a 3000 ms delay period. During valid trials, a 100% predictive cue indicating which disc participants would be probed to report was presented 1000 ms, 1500 ms, or 2000 ms after the onset of the memory array. Using an inverted encoding model, we replicated earlier findings documenting a recovery of location-specific information following a valid retrospective cue. However, the amount of recovery monotonically decreased with the timing of the retrospective cue. Our findings suggest that memory stores used to supplement active WM storage also degrade with time, leading to smaller retrospective cue benefits with longer delays.