GENERALIZED LINEAR MIXED MODELS: AN APPLICATION

The purpose of this paper is to present a specific application of the generalized linear mixed model. Often of interest to animal-breeders is the estimation of genetic parameters associated with certain traits. When the trait is measured in terms of a normally distributed response variable, standard variance-component estimation and mixed-model procedures can be used. Increasingly, breeders are interested in categorical traits (degree of calving difficulty, number born, etc.). An application of the generalized linear mixed to an animal breeding study of the number of lambs born alive will be presented. We will show how the model is determined, how the estimation equations are formed, and the resulting inference.