Abstract

Discrimination thresholds for judgments of dot separation tend to follow Weber’s law. Given the well-documented distortion in comparison of horizontal and vertical distances, we might expect discrimination thresholds for dot separations to be greater when horizontal and vertical distances are compared. Yet we find that, while there is a constant error in comparing a vertical separation to a horizontal, the variance in observer’s judgments does not change appreciably. With an oblique target, observers demonstrated an intermediate constant error but again had similar thresholds. Klein et al.¹ find that the increase in threshold with increasing separation does not occur when stimulus eccentricity is held constant, and they suggest that the Weber law increase results from the logarithmic compression of distance from the fovea in the cortical representation. Cortical magnification factors are not, however, the sole contributors to the observed Weber relationship: we found a Weber’s law function when observers were allowed to fixate freely the target stimuli. We propose that the judgment of large distances reflect errors at some level beyond simple cortical scaling and that an orientation invariant representation of distance is formed from necessarily oriented inputs with no significant loss of precision.

© 1988 Optical Society of America