Abstract
Introduction and Methods The aim of this research was to measure the speed of attentional sampling of multiple dynamic features of the same object. To do this we presented stimuli that were smoothly rotating T-shapes that shifted through a cyclical color vector at 27 millisecond intervals. A circular cue indicated a point in time. Participants reconstructed the color and orientation of the shape at that time using mouse movements. Results Experiment 1 had participants monitor three changing T-shapes (one at the fovea and one on either side, five degrees apart with a 33% size increase to account for eccentricity), each equally likely to be cued. The results confirmed the pattern seen in pilot data where participants reported color information 77ms earlier for the shape in the fovea than the shape in the periphery while orientation report was similar across the fovea and peripheral locations. Experiment 2 replicated this result when all three shapes were of equal size. Again, a significant difference of 126ms was found in the mean latency shift of color reports. This time, a significant difference was also found in orientation report between foveal and peripheral presentation but the difference was still much smaller than for color. Therefore it is not the case that attention is overall slower to select information in the periphery but rather color appears to be accessed more rapidly in the fovea. Surprisingly, the standard deviation of the response distributions was not affect by eccentricity as one might have expected due to the distribution of photoreceptors in the fovea versus the periphery. These results thus suggest that attention accesses different features of the same object at different times, and there is a foveal latency advantage for color despite similar color precision.
Meeting abstract presented at VSS 2017