Standard errors in some cases were larger than the point estimates, making strong inferences ill advised. Table 3 also includes SNP heritability estimates for each endophenotype in the full sample (i.e., without imposing a threshold of relatedness, appearing in the row labeled “None” in the table), which is largely driven by the phenotypic relationships among family members and approximates the sum of factors that give rise to phenotypic resemblance (A and C) in Table 2. We also used the method of Yang and colleagues to model, and thereby control, shared environmental influence in family data while estimating the magnitude of genetic effects (see the column labeled “GCTA-Family”). Because there is by definition no such shared environmental influence in the genetic factor score, this was estimated only for P3 amplitude. Despite the fact that this second method accounts for C, it produced a point estimate that was identical to two decimal places (.57).
