Acuity estimated by visually evoked potentials is affected by scaling

Seven subjects whose corrected Snellen acuities were normal had their monocular acuities (14 eyes) tested by visually evoked potentials (VEPs) elicited by eight checkerboard patterns which reversed 15 times per second. Check size ranged from 20 to 3.4 min arc. Monocular VEP acuities were determined by least squares regression with linear or logarithmic scales of amplitude and pattern size. Pattern size was measured as arc minutes (horizontal size) or fundamental spatial frequency of the checkerboard. The extrapolated VEP acuities were obtained by analyses of variance and post hoc tests. The presence of statistically significant differences in VEP acuity which result from varying combinations of stimulus and response scales indicates a need for caution in selecting scales for VEP estimates of visual acuity.