3 Biometrical Genetical and Epidemiological Approaches

Approaches that use genetic manipulation, natural or artificial, to uncover latent (i.e. unmeasured) genetic and environmental causes of variation are sometimes called biometrical genetical (see e.g. Mather and Jinks, 1982). The methods may be contrasted to the more conventional ones used in individual differences, chiefly in the areas of psychology, sociology and epidemiology. The conventional approaches try to explain variation in one set of measures (the dependent variables) by references to differences in another set of measures (independent variables). For example, the risk for cardiovascular and lung diseases might be assumed to be dependent variables, and cigarette smoking, alcohol use, and life stress independent variables. A fundamental problem with this ``epidemiological approach" is that its conclusions about causality can be seriously misleading. Erroneous inferences would be made if both the dependent and independent variables were caused by the same latent genetic and environmental variables (see e.g., Chapters 6 and 10).