1 Testable Assumptions

The models described in this section are appropriate for analyzing G $\times$ E interaction when genes and environment are acting independently. However, if there is genotype - environment correlation, then more sophisticated statistical procedures are necessary for the analysis. One way of detecting a G - E correlation is to compute the cross-correlations between one twin's environment and the trait of interest in the cotwin (Heath et al., 1989b). If the cross-correlation is not significant, there is no evidence for a G - E correlation, and the G $\times$ E analysis may proceed using the methods described below.