Differences in heritability across groups differing in ability

Differences in heritability and environmentality were assessed for 54 DZ and 86 MZ same-sex twin pairs between 6 and 12 years of age from the Western Reserve Twin Project. A principal-component score composed of the subtests of the WISC-R, PPVT, WRAT, and MAT represented each twin's cognitive ability. Using a modification of a regression technique developed by DeFries and Fulker (1985), it was possible to assess differential heritability and environmentality across ability level. A number of variants of this procedure were used and all yielded the same result: lower ability subjects show higher heritabilities and lower shared environmentality. This result is attributable to larger differences between DZ twins at low ability levels and to differences between MZ twins, which are either the same across ability level or are smaller at low ability levels. A possible explanation for this effect is a genotype-environment correlation in which higher-ability persons seek out better environments. The results from this study should be regarded as tentative but the methods used can be applied to other twins studies. Investigators should be aware of the importance of representing the low end of the distribution in their samples.