genotypes. Therefore, although environmental factors are clearly having a large effect on I, the mean phenotype of the population, and upon the genotypic values Gij, there is hardly any effect at all upon the genotypic deviations, Gij — As shown in Chapter 8, all of the important parameters in the Fisherian quantitative genetic model are ultimately functions of the genotypic deviations, so if the genotypic deviations are the same and the genotype frequencies are similar, the quantitative genetic inferences will also be similar, as indeed they are in this case. This example shows that two populations can differ greatly in their mean phenotypes even though the populations are genetically homogenous. This example also illustrates that differences in environment can contribute to significant differences in mean phenotypes even though the phenotype is influenced by genetic variability in all populations to the same quantitative degree. Hence, the Fisherian parameters are irrelevant to and tell us nothing about the mean phenotypes of populations.
Our second example relates to the phenotype most abused by the spurious argument that high heritabilities imply that mean population differences are due to genetic differences between the populations: the intelligence quotient score (IQ) designed to measure general cognitive ability in humans. One of the earliest studies on the heritability of IQ is that of Skodak and Skeels (1949), a study that is still cited by those such as Herrnstein and Murray (1994), who make the argument that differences in IQ scores between populations are genetically based. The study of Skodak and Skeels is frequently cited in this context because they concluded that IQ has a high heritability. However, a closer examination of their results actually illustrates the inapplicability of heritability to mean phenotypic differences between populations. As mentioned earlier, one great complication in human studies is the fact that environmental variables are often correlated among relatives due to shared family environmental effects. Skodak and Skeels attempted to eliminate or at least reduce these environmental correlations by examining the IQ scores of a population of adopted children and comparing the adopted children both to their biological mothers and to their adoptive mothers. Figure 9.4 shows the normal curves obtained with the observed means and variances of these three populations (adopted children, biological mothers, adoptive mothers). As shown in Figure 9.4, the adoptive mothers had a much higher average IQ
Was this article helpful?