"This book presents broad coverage of variance components estimation and mixed models. Its chapters cover history (Chapter 2), analysis of variance estimation (Chapters 3, 4, and 5), maximum likelihood (ML) estimation, including restricted ML and computational methods (Chapters 6 and 8), prediction in mixed models (Chapter 7), Bayes estimation and hierarchical models (Chapter 9), categorical data (Chapter 10), covariance components and minimum norm estimation (Chapter 11), and finally, the dispersion-mean model, kurtosis and fourth moments (Chapter 12). Estimation from balanced data (having the same number of observations in the subclasses) is dealt with fully in Chapter 4, and in parts of Chapters 3 and 12; and elsewhere, estimation from unbalanced data (having unequal numbers of observations in the subclasses) is dealt with at great length with numerous details for the 1-way and 2-way classifications." "This broad array of topics will appeal to research workers, to students, and to anyone interested in the use of mixed models and variance components for statistically analyzing data. The book will serve as a reference for a wide spectrum of topics for practicing statisticians. For students, it is suitable for linear models courses that include material on mixed models, variance components, and prediction. For graduate courses, there are at least four levels at which the book can be used: (I) As part of a solid linear models course use Chapters 1, 3, and 4, with 2 as supplementary reading. (II) These same chapters, presented in detail, could also be used for a 1-quarter, or slowly paced 1-semester, course on variance components. (III) An advanced course would use Chapters 1 and 2 for an introduction, followed by an overview of Chapters 3 through 5. Then sections 8.1-8.3, Chapters 10 and 11, sections 9.1-9.4, ending with the mathematical synthesis of sections 12.1-12.5 would round out the course. (IV) Finally, the entire book would be suitable for a 2- semester or 3-quarter course." "Nowhere else is there a book devoted solely to variance components with the breadth of topics found in this one."--Jacket.