Simple methods of identifying the independently generated components of scalp-recorded responses evoked by stationary patterns.

The preceding study of the influence of various stimulus parameters on a single subject's pattern-on-set visual evoked potentials (VEPs) identified several constituent potentials with distinctive stimulus-related (and topographic) response properties. This paper describes related experiments in which an appropriate selection of the stimuli used in the original study, and some additional images of faces and other figures, were used to analyse and compare the composition of the VEPs recorded from 49 subjects. The results showed: (1) that the previously discovered response components were again easily identified: the early negative and late negative potentials, for example, were distinguishable in most subjects, not only by their different latencies and surface distributions, but also by the former's preferential, or more often selective, evocation by patterns of discrete elements compared with gratings, and the latter's selective enhancement by patterns containing monocular depth cues; (2) that there was considerable inter-individual variation in the relative sizes (as well as changes in latencies) of the basic components; (3) all the components were not always discernible in each individual's VEPs, and none was recorded for all 49 subjects; and (4) that because of these component amplitude variations, which were the main cause of variable overall response waveforms, there was no simple relationship between the VEP peaks and underlying components. Some important methodological implications of these findings are discussed.