Using the continuous parameter estimation model (CPEM), a large genotyped adult sample of the population of Wisconsin, USA (the Wisconsin Longitudinal Study) is examined for evidence of the Scarr-Rowe effect, a gene × environment (G×E) interaction that reduces the heritability of IQ among those with low socioeconomic status (SES). This method allows the differential expressivity of polygenic scores predictive of both educational attainment and IQ (EA3) on the phenotype of IQ to be directly operationalized throughout the full range of these variables. Utilizing a parental SES factor-weighted composite as a measure of childhood SES, evidence for the Scarr-Rowe effect was found, that is, the genetic expressivity of EA3 on IQ increased with increasing parental SES (β = 0.08, p = 4.71×10-10, df = 6,255). The effect was found for both the male and female samples separately, β(males) = 0.08, p = 5.27×10-5, df = 3,018; β(females) = 0.08, p = 1.93×10-6, df = 3,236. The effects were furthermore robust to removing outlying values of parental SES and to log-transforming the SES variable. The results are similar to those produced using a more conventional two-way interaction model, with IQ predicting the EA3 × log of parental SES interaction after the main effects; however, CPEM allows for greater model degrees of freedom, thus is better powered to detect the effect when it is small in magnitude (CPEM β = 0.05, p = 6.69×10-5 vs. two-way interaction β = 0.02, pone-tailed = .045, in both models log parental SES is used).
Keywords: CPEM; G × E interaction; Scarr–Rowe effect; polygenic score.