Comparison of pattern VEPs and preferential-looking behavior in 3-month-old infants

Studies of visual acuity in human infants between 1 and 6 months of age using the visual-evoked potential (VEP) and forced-choice preferential looking (FPL) have shown that acuity is one to two octaves higher by VEP estimates than by FPL estimates. In an attempt to study these differences, the authors obtained both VEP and FPL data from 26 3-month-old infants. VEP data were obtained with gratings of 0.31, 0.62, 1.25 and 2.50 cycles/deg, which were counterphase alternated at 2 Hz. FPL data were obtained for stationary gratings using either the method of constant stimuli or a staircase procedure. Our study revealed three major findings: (1) recordable VEPs can be obtained for spatial patterns that are below threshold by behavioral measures; (2) the use of different scoring criteria that yields comparable VEP and FPL group mean acuities does not yield a significant correlation between VEP amplitude acuity and FPL acuity for individual infants, probably because of the inherent "noise" in each technique; and (3) when VEP latency rather than amplitude is used to estimate acuity, there is a significant correlation between electrophysiology and behavior.