Since a model-based stochastic regression imputation strategy was employed to impute data for the five domains of NEPSY at F3, we assessed the differences in paths between the model with imputation used in the study and a model in which there was no imputation. The mean change in the standardized paths in the model-based analyses as a result of imputation of NEPSY at F3 was <.02 (Table 3). Given that the results remained substantially the same with and without imputation, our analyses appear robust to the imputation of missing data.
