This book provides a comprehensive overview of many variants of the general linear model, accompanied by examples analysed in SAS. The most significant innovation since the first edition is the relegation of the SAS code and data files to the enclosed CD and the addition of new material. The scope of the book is remarkably wide. The first chapter introduces the general linear model, summarises some basic facts on the multivariate normal distribution, and discusses the assessment of univariate and multivariate normality. Chapter 2 covers unrestricted general linear models, including multiple regression and one-way analysis of variance. Chapter 3 reviews restricted models and applies them to two-way factorial designs, Latin square designs, repeated measures design and the analysis of covariance. Weighted general linear models are introduced in chapter 4 and are applied to the analysis of categorical data and models with heteroscedastic errors. Chapter 5 discusses multivariate general linear models: multivariate regression, multivariate mixed models, and MANOVA and MANCOVA designs. In the next two chapters, the multivariate model is extended to a doubly multivariate model (chapter 6) and a multivariate model with restrictions, with applications to growth curves (chapter 7). Chapter 8 deals with the “seemingly unrelated regression” (SUR) model and the restricted GMANOVA model. Chapters 9 and 10 (new since the first edition) cover simultaneous inference using finite intersection tests and power calculations, respectively. Two-level hierarchical linear models are treated in chapter 11. The last two chapters are devoted to incomplete repeated measurements and structural equation modelling. The authors note that the book “is written for advanced graduate students in the social and behavioral sciences and in applied statistics” and that it is suitable for a one-semester course on linear models. Each chapter is accompanied by numerical examples, discussions of SAS code and interpretations of SAS output. Additional material is available on the authors’ webpage.

Reviewer:

mkul