Variance Components Estimation deals with the evaluation of the variation between observable data or classes of data. This is an up-to-date, comprehensive work that is both theoretical and applied. Topics include ML and REML methods of estimation; Steepest-Acent, Newton-Raphson, scoring, and EM algorithms; MINQUE and MIVQUE, confidence intervals for variance components and their ratios; Bayesian approaches and hierarchical models; mixed models for longitudinal data; repeated measures and multivariate observations; as well as non-linear and generalized linear models with random effects.
Table of Contents
The Study of Variation. One-Way Classification. Two-Way Cross-Classification. Randomized Blocks. BIBDs and Latin Squares. Nested Classifications. Maximum Likelihood Estimation. The MINQUE and MIVQUE. Non-Negative Estimation of Variance Components. Confidence Intervals. Genetic and Environmental Effects.
Rao\, Poduri S.R.S.
"This is an intermediate level monograph on variance component estimation. It is concise and well written... I recommend this book to any practicing statistician or researcher. It could also be used to supplement an intermediate or advanced course on linear models, analysis of variance, or variance components."
-Technometrics, February 2002