Design and analysis of experiments/ Douglas C. Montgomery

Includes bibliography and index

Introduction -- Simple Comparative Experiments -- Experiments to compare Several Treatments: The Analysis of Variance -- More About the One- Way Model -- Randomized Blocks, Latin squares, and Related Designs -- Incomplete Block Designs -- Introduction to Factorial Designs -- Rules for Sum of Squares and Expected Mean Squares -- 2 k and 3 k Factorial Designs -- Confounding -- Fractional Factorial Designs -- Nested or Hierarchical Designs -- Multifactor Experiments with Randomization Restrictions -- Response Survey Methodology -- Analysis of Covariance

The Second Edition is a major revision of the original book. While the flavor and level of the first edition have been maintained, with its balance between design and analysis topics, much new material has been added. There is now a strong emphasis on model adequacy checking for the analysis of variance, and residual analysis is the major method of diagnostic checking employed. Other new topics include an introduction to methods for dealing with unbalanced data, more expensive coverage of fractional, expanded discussion of the multiple comparison problem, and more guidelines for the choice of sample size. Moreover, much of the material from the edition has been extensively recognized.
Chapter 1 presents the basic philosophy of the statistical approach to experimental design.
Chapter 2 reviews elementary statistical methods from the viewpoint of designing simple comparative experiments and then introduces terminology and notation used in subsequent chapters.
In Chapter 3, we begin study of completely randomized designs for experiments with a single factor.
Chapter 4 discuss methods for model adequacy checking in the completely randomized design.
Chapter 5 and 6 continue the development of single- factor experiments, with randomized blocks, Latin squares, and related design discussed in Chapter 5. Chapter 6 introduces incomplete block designs. Factorial designs are introduced incomplete block designs.
Factorial designs are introduced in Chapter 7.
Chapter 8 presents a set of rules for deriving computing formulas for sums of squares and expected mean square for any balanced multi-factor design.
Chapter 9 introduced the 2 k and 3 k factorial designs.
Procedures for constructing and analyzing these designs are given in Chapter 10.
In Chapter 11 the relatively high cost of industrial experimentation has led to the extensive use of fractional 2 k and 3 k factorial designs.
the basic presentation of multi-factor designs is continued in chapter 12.
Chapter 13 illustrated how randomization restrictions are employed in multi-factor experiments.
Regression analysis is introduced in Chapter 15.