We next tested whether Marital Quality moderated the genetic and environmental influences on Internalizing pathology. Ultimately, our goal was to test the full moderation model with all moderation paths freely estimated, as shown in Figure 1. As a baseline model of the etiological influences common to Marital Quality and Internalizing, we ran a no-moderation model in which the moderation parameters were fixed to 0 (−2lnL=3761.34, df=1377, AIC=1007.34). Variance components and genetic and environmental correlations from this no-moderation model are presented in Table 2. Of note, the no-moderation model is essentially equivalent to a bivariate decomposition (Cholesky) model, but differs in the degrees of freedom and −2 log likelihood due to the data entry method required for the moderation model (the moderator variable is entered twice). From the baseline, no-moderation model, we added each of the six moderation paths (the path common to both Marital Quality and Internalizing, and the path unique to Internalizing, for genetic, shared environmental, and unique environmental influences) one at a time. For five of the six moderation parameters, the model with the moderation parameter provided a
