Purchase this article with an account.
Benjamin Pearson, Paul Bays, Masud Husain; Decision-making in visual working memory. Journal of Vision 2012;12(9):165. doi: https://doi.org/10.1167/12.9.165.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent research in visual working memory (WM) has focussed on patterns of errors in change detection, clarifying the limits on WM maintenance. But there has been far less interest in what reaction times (RTs) on such tasks might reveal about the mechanisms of memory retrieval. For human and primate perceptual judgements, careful study of RT and choice has found evidence of neural accumulators that weigh up stimulus information to select a response. To test if similar decision mechanisms might support retrieval from VWM, eighteen subjects viewed, in different experiments, sets of colored squares or oriented lines and reported the direction of change in position or orientation when one item reappeared after a delay (a design previously reported in Bays & Husain, 2008). The number of items (N) and change magnitude (Δ) were manipulated in a full factorial design. For both tasks, RT increased as Δ was reduced. RT also increased with set size, with the largest changes taking place at small N. The shape of RT distributions within conditions was consistent with a decision process featuring competing linear accumulators. According to this model, the effect of N and Δ was to change only the evidence available to the decision process (μ), and not decision threshold. Echoing previous results for memory precision (Bays & Husain, 2008), μ followed an inverse power relation to N. In fact μ was linearly related to memory precision for all judgements difficulties (Δ). In summary, the results support a flexible resource model of WM in which item features are stored in noisy representations, from which competing neural processes accrue evidence to form decisions. Bays, P. M., & Husain, M. (2008). Dynamic Shifts of Limited Working Memory Resources in Human Vision. Science, 321(5890), 851-854.
Meeting abstract presented at VSS 2012
This PDF is available to Subscribers Only