ABSTRACT
Mixed models are used to describe data from experiments or studies that need more than one variance-covariance parameter and involve some xed effects parameters. The unequal variance models of Chapter 2 are mixed models as they involve more than one variance component. The models described in Chapters 18-21 are called random effects models, but each has an unknown mean parameter, thus the models are essentially mixed models. The de nition of a mixed model used in Chapter 18 revolved around having some of the treatment structure with xed effects and more than one variance component. So the general de nition of a mixed model is one with some xed effect parameters and more than one parameter in the covariance structure. Among the models that are included in this de nition are randomized complete blocks models, incomplete blocks models, split-plot-type models, strip-plot-type models, repeated measures type models, random coef cients models, multilevel models, and hierarchical models.
