Abstract
Visual Working Memory (WM) capacity is limited to around 3-4 simple items, yet the capacity of long-term memory is virtually unlimited. Here, we measured increases in WM performance across back-to-back repetitions of identical supra-capacity arrays. We hypothesized that subjects would show typical capacity limits for the first presentation of an array but reach near-perfect performance when they are able to recruit long-term memory resources after multiple repetitions of the same array. Further, if WM acts as a gateway for the transfer of information into long-term memory, then WM capacity should predict the change in performance across repetitions. On each trial, subjects (N = 31) saw an array of 6 colored squares, remembered the array across a blank delay, and then reported the identity of all items in the array. The same array was repeated for 8 trials. Participants completed a total of 240 trials (30 unique arrays), and performance was calculated as the number of correctly reported items for each trial. After the repetition task, participants performed a surprise old-new recognition task and a change detection measure of visual working memory capacity. As predicted, WM performance showed typical limits for the first presentation of the array (~2.75 items correct) and surpassed typical limits after several presentations (>5 items correct). A separate measure of capacity predicted both the rate of performance increase across repetitions (r = .54) and performance on the surprise long-term memory test (r = .58). Thus, we find support for the notion that working memory capacity influences the number of encounters needed to successfully encode a large array of items into long-term memory.
Meeting abstract presented at VSS 2016