Abstract
The representation of information across space is fundamental to visual perception. It is well known that the visual field is not uniformly organized within visual cortex. How does the organization of visual cortex contribute to the precision in spatial representations? In the present study, we aim to extend the current understanding of spatial representations, by examining the precision of these representations along the continuum of angular location. Participants viewed a small checkerboard (0.5 degree visual angle) briefly presented (500 ms) at random locations along an iso eccentric circle (2.5 degree eccentricity). After a short delay, participants reported the target’s location by moving a probe along the circle to the perceived location as precisely as possible. Analysis of the behavioral results shows that this simple spatial localization task resulted in large, systematic misrepresentations of angular location (up to ± 10 degrees mean angular error), consistent with prior reported categorical biases. In order to examine the precision in these behavioral judgements, we removed these location-dependent repulsive biases from the data, grouped trials into bins based on the presented location, and computed the variance of the behavioral errors across trials within each bin. We found that behavioral variability varied as a function of angular location: behavioral judgements had the greatest precision at the horizontal meridian, and the least precision, or largest perceptual uncertainty, in angular position at off-cardinal locations. In ongoing work, we will investigate the degree to which these behavioral values of precision can be linked to fMRI measures of precision in the visual cortex. Together, these findings will provide insights into the neural implementation of spatial representations in the human visual cortex.