MULTIVARIATE GENETIC-ANALYSIS OF TWIN-FAMILY DATA ON FEARS - MX MODELS

We describe the implementation of multivariate models of familial rese mblance with the Mx package. The structural equation models allow for the effects of assortative mating, additive and dominant genes, common and specific environment, and both genetic and cultural transmission between generations. Two approaches are compared: a correlational one based on Fulker and a factor model described by Phillips and Fulker. B oth are illustrated by application to published data on social fears a nd fear of leadership measured in monozygotic and dizygotic twins and their parents. In the example data, genetic dominance yields a more pa rsimonious explanation of the data than does cultural transmission, al though neither is needed to obtain a good fit to the data. A model of reduced genetic correlation between generations also fits the data but has inherent limitations in this sample. Extensions to sex-limitation and more complex models are discussed.