Abstract
A recent important debate regarding the nature of visual working memory concerns whether it represents a small set of high-resolution representations (the “slot” model, Zhang & Luck, 2008), or all items in parallel (the “resource” model, Bays & Husain, 2008). These two models make different claims regarding how participants store information when faced with a large number of items: the slot model claims that they store high-resolution representations of a subset of the objects and retain no information about the others, whereas the resource model claims that they could store some imperfect information about each of the objects. In this study, we distinguished the two models by asking participants to memorize the colors of six objects, and then testing their recall of each color. The frequency distribution of scores (i.e., the number [0–6] of correct responses) was modeled, and, in terms of the predictions of the two models, the empirical pattern of frequency distribution was found to fit precisely with the resource model, but to differ clearly from the slot model. Furthermore, we showed that the all-or-none mode demonstrated by Zhang and Luck (2008) can manifest without any involvement of visual working memory slot limits, and that, therefore, it is probably irrelevant to the mechanisms of visual working memory.