Design and Analysis of Experiments

Of the many areas in which statisticians and scientists must combine their knowledge, none is more important than the design and analysis of experiments. There is a direct relationship between the quality of experimental design and analysis and the validity of conclusions drawn from experimental results. Design and Analysis of Experiments is the first of two volumes that update Oscar Kempthorne's groundbreaking 1952 classic of the same name. This first volume establishes both the philosophical basis for experimental design and a mathematical-statistical framework within which to discuss the subject. A sharp distinction is drawn between observation studies and the intervention studies that are the primary topic of this book. A thorough discussion of linear models provides the basic tools of analysis for intervention studies. At the heart of the book is a series of error-control designs based on fundamental design principles such as randomization, blocking, the Latin square principle, the split-unit principle, and the notion of factorial treatment structure. There are also discussions of response surface and mixture designs, and the construction of customized designs based on one or more fundamental principles. This is the ultimate textbook for a first-year graduate course in experimental design. A thorough understanding of the material covered in this book will fully prepare students to grapple with the more complex technical material presented in advanced courses in experimental design. The classic text now updated and expanded for the 1990s...Design and Analysis of Experiments is the first of a two-volume update of Oscar Kempthorne's celebrated classic of the same name. This first volume is concerned primarily with the philosophical basis for experimental design and a mathematical-statistical framework within which to discuss the subject. As the primary focus is on intervention studies, the authors begin with a thorough discussion of linear models. At the heart of the book, is a series of error-control designs based on fundamental design principles such as randomization, blocking, the Latin square principle, the split-unit principle, and the notion of factorial treatment structure. The material covered in this book will more than prepare first-year graduate students for the challenges of more advanced courses in experimental design.