Failed and outdated social science theorizing? A commentary on IQ distributions in fathers and sons

Abstract

Vroon, de Leeuw & Meester (1986) interpreted a father‐son IQ correlation of +0.34 as discrediting family background as the origin of a significant part of total IQ variation. This commentary argues that Vroon et al. have misunderstood the concept of variance partitioning as it is used in behavioural genetics. Based on their data, some 34 per cent (under a model of purely shared environmental influences) to 68 per cent or more (under a model of purely genetic influences) of IQ variation may be attributed to family background factors, reinforcing the conclusion that familial influences are extremely important for understanding individual differences in IQ. Their approach, inferring the extent of familial variation from just the father‐son relationship, is criticized as being outdated, compared with current methodology in behavioural genetics. Three aspects of contemporary behaviour genetics are examined in relation to their results: (1) the fitting of models to large data sets derived from many familial relationships; (2) a multivariate approach dissecting the genetic and environmental contributions to the phenotypic relationship among ability and achievement measures; and (3) the growing emphasis on within‐family differences, both genetic and environmental. Behavioural genetic models can take into account possible biases such as the generation age difference which Vroon et al. emphasize as a problem for family studies.