The purpose of this chapter is to give a short description of some of the parameters of interest when behavioral genetic methods are applied to longitudinal designs. In order to make this clear, some examples will be provided from the ongoing Swedish Adoption/Twin Study of Aging (SATSA). Finally, several of the models currently being used will be presented and discussed.

PARAMETERS OF INTEREST OR WHAT DO WE WANT TO ASSESS?

To understand the application of behavioral genetic models to longitudinal designs, one must first be aware of the essential parameters in the simple univariate, one-occasion case. Unlike normative studies in which the primary emphasis is on mean levels of a behavior or phenotype, behavioral genetic studies focus on describing interindividual differences by partitioning phenotypic variance (P) into genetic (G) and environmental (E) sources of variation.

The relationship of these parameters can be described in an equation, the simplest of which is P = G + E, or in a path diagram (Figure 11.1). Regardless of form of presentation or number of parameters included, the main idea is to determine to what extent interindividual differences are due to genetic and environmental differences in the population. Behavioral genetic methodology applies the principles of quantitative genetics to the study of subjects with differing degrees of relatedness and varying similarities in their environments to partition variation. (For further descriptions of basic behavioral genetic techniques and their application to a variety of phenotypes, see Plomin, DeFries & McClearn, 1990.)