The biology of intelligence?

Abstract
The idea that there is a biological basis for intelligence in people has been translated into two beliefs about IQ tests: first, that measured IQ is genetically determined and that differences in IQ between different groups are partly or largely caused by genetic differences between them; secondly, that IQ reflects some presumed fundamental property of the brain such as efficiency or speed of neural transmission, that can be measured by recording evoked potentials or by the speed with which a person performs some very simple task. Measured IQ may well have a significant heritable component, but there is very little evidence that average differences in IQ between, say, different ethnic groups are genetic in origin. In Great Britain, moreover, differences in IQ between white, West Indian, Indian or Pakistani children are closely correlated with differences in their social circumstances. It should be stressed, however, that there is equally little evidence that IQ tests significantly underestimate the academic attainments of children from ethnic minorities or that such children fall progressively further behind as they go through school. The search for simple, physiological or behavioural correlates of IQ has met with no more than modest success. Recent reports of highly significant correlations between IQ and measures of evoked potentials have not always been replicated and the past history of failures of replication in this general area counsels caution. Reports of very high (greater than 0.60) correlations between IQ and measures of timed performance have not been substantiated and can usually be attributed to the inclusion of disproportionate numbers of retarded subjects in the samples. There is quite good evidence of moderate correlations (in the range 0.20 to 0.40) with those measures, but it is not even clear how they should be interpreted. One possibility is that they reflect differences in concentration or sustained attention. If we want higher correlations with IQ we will probably need to look to more sophisticated tasks.

This publication has 0 references indexed in Scilit: