Averaged evoked responses and psychometric intelligence

Abstract

Attempts to establish correlations between psychometrically defined and measured intelligence and brain electrical activity have met with limited success. For example, as in the work of Ertl¹, the latency of components of averaged evoked potentials (a.e.ps) has been adopted as the EEG measure most likely to correspond to intelligence quotient (IQ). Although correlations as high as 0.35 have been reported on samples unselected for IQ, the sign of the correlation seems to depend on the stimulus modality used, which suggests that a straightforward identification of intelligence with ‘mental speed’, for which read ‘short a.e.p. component latencies’, is inappropriate. Hendrickson and Hendrickson² have proposed both a novel a.e.p. measure and a theory to account for individual differences in scores based on this measure, claiming remarkably high correlations with IQ, of the order of 0.7–0.8. We report here a study using highly selected subjects which supports these claims.