Next: 1 Multivariate Genetic Factor
Up: 10 Multivariate Analysis
Previous: Mx Output from Phenotypic
The factor analytic approach outlined above can be readily applied to
multivariate genetic problems. This was first suggested by
Martin and Eaves (1977)
for the analysis of twin data (although in their original publication they
use matrices of mean squares and cross-products between and within twin
pairs). As in the phenotypic example above, a single
common factor is proposed to account for correlations among the variables,
but now one such factor is hypothesized for each of the components of
variation, genetic, shared environmental, and non-shared environmental.
Data from genetically related individuals are used to estimate loadings of
variables on common genetic and environmental factors, so that variances
and covariances may be explained in terms of these factors.
3 Simple Genetic Factor Models