Abstract
Current models of visual working memory have conceptualized this system as consisting of either discrete slots with fixed visual precision or a flexible resource that leads to variability in memory precision across items. However, these models have paid little attention to the possibility that precision may also depend on the perceptual sensitivity of the visual system at encoding specific items. Also, they have not considered the extent to which decisional response bias might influence working memory performance. To investigate these contributions, we conducted an experiment in which participants were presented with a circular array of 1-6 oriented gratings, and subsequently had to report the orientation of a probed item by rotating a central test grating. Using a mixture-model analysis, we observed a pronounced "oblique effect", with better memory precision found for cardinal than oblique orientations at every set size tested. This effect was prominent even at set size 1, when full attentional resources could be directed to a single item, implying that memory precision was limited by orientation sensitivity. Additional analyses revealed a decisional response bias in favor of reporting an orientation that happened to be formed by the angle between the point of fixation and the location of the probed item, as well as the angle orthogonal to this arbitrary orientation. This bias was evident only at large set sizes that exceeded typical estimates of memory capacity, which is suggestive of a guessing strategy that participants use when their knowledge of the true orientation is deemed unreliable. Overall, our findings demonstrate that the perceptual sensitivity of the visual system introduces a systematic variability in memory precision. Moreover, large display sizes lead to a more prevalent response bias, a finding that is not readily predicted by resource models that assume that all items in a display can be successfully retained.
Meeting abstract presented at VSS 2014