Abstract

Performance in visual tasks can often be equated across eccentricities by proper scaling. The scaling or inverse magnification function (EMF), describes the ratio of peripheral to foveal stimulus size required to equate performance. Tasks and visual brain regions have different IMFs. It is argued in this thesis that IMFs may average out when a methodology insensitive to the presence of multiple IMFs is used. This fact is demonstrated through simulations. The present thesis introduces a data fitting technique that detects the presence of multiple IMFs in a psychophysical task. These are revealed as an interaction between stimulus configuration and eccentricity. These new techniques were used to investigate the percept of subjective contours (SC) defined by offset gratings which are thought to be encoded through a cooperation of V1 and V2 cells, two brain areas described by different IMFs. Five participants discriminated the orientation of a SC presented foveally (monocularly or binocularly) and at four eccentricities. SC length and carrier grating wavelength were adjusted until performance converged on 81% correct. There was an interaction between eccentricity and stimulus configuration, F (20, 80) = 2.063, p = .0124, which was accounted for only if two IMFs were assumed. It was found that SC length (V2) scaled faster than the wavelength (V1) as a function of eccentricity. This qualitatively agrees with anatomical measures of V1 and V2 IMFs. The method developed here provides a more informative and more objective measure of eccentricity-dependent performance limitations than other commonly-used methods.