Correlations between intelligence and the frequency content of the visual evoked potential

Abstract

Visually evoked potentials (VEP) elicited by a flashing light were recorded from 64 white 12-year-old girls whose IQ scores were uniformly distributed across the range found in a normal school population. The objective was to examine the correlations between intelligence, as measured on standard IQ tests, and characteristics of the VEP. Two questions were studied: whether such correlations are caused mainly by low IQ subjects rather than by differences which extend across the whole IQ range, and the usefulness of frequency analysis in analyzing the data. Recording was bipolar from the right parietal area of the head. The results were that the low IQ subjects had larger low-frequency ( 30 Hz) components than the low IQ subjects. Correlations of low-frequency spectral amplitude with IQ were caused by differences between only the lowest IQ group (IQ less than 90) and all other subjects; these correlations were approximately the same from the first to the last of four runs. Correlations of high-frequency components with IQ were due to differences between both the low IQ and average subjects and the high IQ and average subjects. These correlations decreased between the first and the fourth run. The differences in the properties of the correlation coefficients in the two frequency ranges suggest that there may be two effects causing the differences in the VEPs. These differences are seen in two distinct frequency ranges.