HETEROGENEOUS VARIANCES IN GAUSSIAN LINEAR MIXED MODELS

This paper reviews some problems encountered in estimating heterogeneo us variances in Gaussian linear mixed models. The one-way and multiple classification cases are considered. EM-REML algorithms and Bayesian procedures are derived. A structural mixed linear model on log-varianc e components is also presented, which allows identification of meaning ful sources of variation of heterogeneous residual and genetic compone nts of variance and assessment of their magnitude and mode of action.