Abstract
Our ability to reliably judge and reproduce time intervals under one or two seconds is surprising, given the variability of the temporal context (Jazayeri & Shadlen, 2010, Nature Neuroscience, 13(8), 1020-1026). This temporal range is particularly important in social interactions that are often managed by gaze duration. In this study, we examined subjects’ ability to reproduce the duration of an observed gaze by moving their own eyes. Participants were presented with an avatar reference face that shifted its gaze twice. The interval between the two gaze shifts was the standard duration (500, 1000 or 1500 ms). Subjects had to reproduce this interval by looking at an identical face and then move their eyes away when they thought the same time had passed. At the end of every pair of identical trials, we asked which of the two previous estimates they thought was closer to the standard duration. No feedback was provided. We calculated the mean of the distribution of individual estimates as a measure of perceived duration and the standard deviation of this mean as a measure of observer variability. Although we observed a general underestimation of gaze duration, the mean of the true best estimates did not substantially differ from the mean of the perceived best estimates. However, the standard deviation of the perceived best estimates was lower than that of the perceived worst estimates, indicating subjects had some knowledge of their own performance.
Meeting abstract presented at VSS 2013