Abstract
Although visual short-term memory (VSTM) has been studied extensively, the majority of this literature has focused on the impact of 2-D stimulus properties. Studies of the effects of binocular disparity on VSTM have drawn conflicting conclusions due to design and stimulus differences. To reassess the impact of stereopsis on VSTM we used a novel paradigm in which oriented brackets were presented on two disparity-defined planes simultaneously. A concurrent letter recall task prevented observers from cognitively rehearsing a verbal descriptor of the brackets. On each trial, observers viewed a set of four randomly selected letters, followed by a field of brackets containing two of four possible orientations (left, right, up, down). Observers were asked to remember the set of letters, and the bracket orientations. In the test phase, another set of letters appeared and observers indicated if they were the same or different from the first. A field of brackets then appeared and observers indicated if the orientations were the same or different from the original. We computed d’ for four conditions: zero disparity full-density, zero disparity half-density, two disparity-defined planes with non-uniformly distributed brackets and two disparity-defined planes with uniformly distributed brackets (each plane contained one orientation). We found no difference between 2-D and 3-D conditions, however there was a large effect of element distribution in the 3-D conditions. In a follow up experiment we varied binocular disparity to determine whether this result was specific to a particular depth offset. There was a clear interaction between disparity and element distribution in the 3-D conditions. We conclude that there is no overall benefit to VSTM of displacing stimuli on different depth planes. Instead it appears that VSTM is influenced by the perceptual organization of the elements, which can be influenced by their 3-D placement.
Meeting abstract presented at VSS 2015