Purchase this article with an account.
Kirsten Adam, Edward Vogel, Edward Awh; Decoding the limits of simultaneous storage in working memory. Journal of Vision 2018;18(10):366. doi: https://doi.org/10.1167/18.10.366.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Competing models of visual working memory (WM) make strongly diverging claims about the number of items that can be held actively in mind. These include: (1) no limit in the number of items, but a degradation in quality with more remembered items (2) a capacity limit of around 3 items, or, even (3) a single prioritized item in the focus of attention. Behavioral tests of these models are useful but fundamentally limited – even if participants cannot report an item at test, it is still possible that they actively represented the item during maintenance. Thus, a method that enables simultaneous decoding of the specific items maintained in WM would provide powerful traction for this debate. Here, we made significant advances toward this aim by attempting to decode the locations of all actively maintained items from multi-item arrays using the topography of alpha-band power (8-12 Hz) in the human EEG signal. In Experiment 1 (n=31), colored squares appeared in three of eight possible location bins equidistant from fixation. Participants were pre-cued to 1, 2, or 3 relevant items with centrally presented spatial cues (small lines pointing toward the location/s). We found reliable decoding of 2 items simultaneously, counter to predictions of a single focus of attention model. In Experiment 2 (n=20), participants were presented with 1, 3, or 6 items. Critically, for set size 6 arrays, evidence for active storage was restricted to the 3 best –remembered items, while there was no active neural signal tracking the remaining items. Thus, neural signals that track WM storage in an item-specific manner rule out models in which all items are stored equally imprecisely, and suggest strict limits on the number of items that can be actively maintained.
Meeting abstract presented at VSS 2018
This PDF is available to Subscribers Only