Design of Experiments: Statistical Principles of Research Design and Analysis, Second Edition
Author:
|
Robert O. Kuehl |
| Publisher: |
Brooks/Cole |
| Copyright: |
2000 |
| ISBN-13: |
978-0-534-36834-0 |
| Pages: |
666; hardcover |
| Price: |
$192.50 |
|
|
|
|
Comment from the Stata technical group
The second edition of Design of Experiments: Statistical Principles of
Research Design and Analysis by Robert O. Kuehl is an excellent
introduction to the methods of research design. The book emphasizes
application but provides sufficient theory for a basic understanding of the
statistical principles involved. College algebra and basic probability and
statistics are sufficient mathematical background for readers of this book.
Chapter 1 emphasizes performing good science through proper planning and
research design. Kuehl explains the importance of starting with your
research hypothesis, understanding sources of variation, understanding the
difference between experimental units and observational units, and the
benefits from using replication, randomization, blocking, controls, and
covariates.
Chapter 2 goes through all the nuts and bolts of ANOVA clearly and
precisely. Chapter 3 covers multiple treatment comparisons. Chapter 4
discusses diagnostic methods for testing model assumptions. Chapters
5–7 explain factorial and nested treatments designs for fixed, random,
and mixed models. The remaining chapters (8–17) are devoted to more
complicated experimental designs: complete block, incomplete block,
fractional factorial, split plot, repeated measures, crossover, and ANCOVA.
Throughout the text, examples and exercises come from research in life
sciences, agriculture, engineering, industry, and chemistry. This book is
ideal for research professionals learning research design and ANOVA. Unlike
with many texts, the goal of providing answers to real research hypotheses
does not get lost in presenting the mathematical details.
Table of contents
1. Research Design Principles
1.1 The Legacy of Sir Ronald A. Fisher
1.2 Planning for Research
1.3 Experiments, Treatments, and Experimental Units
1.4 Research Hypotheses Generate Treatment Designs
1.5 Local Control of Experimental Errors
1.6 Replication for Valid Experiments
1.7 How Many Replications?
1.8 Randomization for Valid Inferences
1.9 Relative Efficiency of Experiment Designs
1.10 From Principles to Practice: A Case Study
2. Getting Started with Completely Randomized Designs
2.1 Assembling the Research Design
2.2 How to Randomize
2.3 Preparation of Data Files for the Analysis
2.4 A Statistical Model for the Experiment
2.5 Estimation of the Model Parameters with Least Squares
2.6 Sums of Squares to Identify Important Sources of Variation
2.7 A Treatment Effects Model
2.8 Degrees of Freedom
2.9 Summaries in the Analysis of Variance Table
2.10 Tests of Hypotheses About Linear Models
2.11 Significance Testing and Tests of Hypotheses
2.12 Standard Errors and Confidence Intervals for Treatment Means
2.13 Unequal Replication of the Treatments
2.14 How Many Replications for the F Test?
2A.1 Appendix: Expected Values
2A.2 Appendix: Expected Mean Squares
3. Treatment Comparisons
3.1 Treatment Comparisons Answer Research Questions
3.2 Planning Comparisons Among Treatments
3.3 Response Curves for Quantitative Treatment Factors
3.4 Multiple Comparisons Affect Error Rates
3.5 Simultaneous Statistical Inference
3.6 Multiple Comparisons with the Best Treatment
3.7 Comparison of All Treatments with a Control
3.8 Pairwise Comparison of All Treatments
3.9 Summary Comments on Multiple Comparisons
3A Appendix: Linear Functions of Random Variables
4. Diagnosing Agreement Between the Data and the Model
4.1 Valid Analysis Depends on Valid Assumptions
4.2 Effects of Departures from Assumptions
4.3 Residuals Are the Basis of Diagnostic Tools
4.4 Looking for Outliers with the Residuals
4.5 Variance-Stabilizing Transformations for Data with Known Distributions
4.6 Power Transformations to Stabilize Variances
4.7 Generalizing the Linear Model
4.8 Model Evaluation with Residual-Fitted Spread Plots
4A Appendix: Data for Example
5. Experiments to Study Variances
5.1 Random Effects Models for Variances
5.2 A Statistical Model for Variance Components
5.3 Point Estimates of Variance Components
5.4 Interval Estimates for Variance Components
5.5 Courses of Action with Negative Variance Estimates
5.6 Intraclass Correlation Measures Similarity in a Group
5.7 Unequal Numbers of Observations in the Groups
5.8 How Many Observations to Study Variances?
5.9 Random Subsamples to Procure Data for the Experiment
5.10 Using Variance Estimates to Allocate Sampling Efforts
5.11 Unequal Numbers of Replications and Subsamples
5A Appendix: Coefficient Calculations for Expected Mean Squares in Table 5.9
6. Factorial Treatment Designs
6.1 Efficient Experiments with Factorial Treatment Designs
6.2 Three Types of Treatment Factor Effects
6.3 The Statistical Model for Two Treatment Factors
6.4 The Analysis for Two Factors
6.5 Using Response Curves for Quantitative Treatment Factors
6.6 Three Treatment Factors
6.7 Estimation of Error Variance with One Replication
6.8 How Many Replications to Test Factor Effects?
6.9 Unequal Replication of Treatments
6A Appendix: Least Squares for Factorial Treatment Designs
7. Factorial Treatment Designs: Random and Mixed Models
7.1 Random Effects for Factorial Treatment Designs
7.2 Mixed Models
7.3 Nested Factor Designs: A Variation on the Theme
7.4 Nested and Crossed Factors Designs
7.5 How Many Replications?
7.6 Expected Mean Square Rules
8. Complete Block Designs
8.1 Blocking to Increase Precision
8.2 Randomized Complete Block Designs Use One Blocking Criterion
8.3 Latin Square Designs Use Two Blocking Criteria
8.4 Factorial Experiments in Complete Block Designs
8.5 Missing Data in Blocked Designs
8.6 Experiments Performed Several Times
8A Appendix: Selected Latin Squares
9. Incomplete Block Designs: An Introduction
9.1 Incomplete Blocks of Treatments to Reduce Block Size
9.2 Balanced Incomplete Block (BIB) Designs
9.3 How to Randomize Incomplete Block Designs
9.4 Analysis of BIB Designs
9.5 Row–Column Designs for Two Blocking Criteria
9.6 Reduce Experiment Size with Partially Balanced (PBIB) Designs
9.7 Efficiency of Incomplete Block Designs
9A.1 Appendix: Selected Balanced Incomplete Block Designs
9A.2 Appendix: Selected Incomplete Latin Square Designs
9A.3 Appendix: Least Squares Estimates for BIB Designs
10. Incomplete Block Designs: Resolvable and Cyclic Designs
10.1 Resolvable Designs to Help Manage the Experiment
10.2 Resolvable Row–Column Designs for Two Blocking Criteria
10.3 Cyclic Designs Simplify Design Construction
10.4 Choosing Incomplete Block Designs
10A.1 Appendix: Plans for Cyclic Designs
10A.2 Appendix: Generating Arrays for a Designs
11. Incomplete Block Designs: Factorial Treatment Designs
11.1 Taking Greater Advantage of Factorial Treatment Designs
11.2 2n Factorials to Evaluate Many Factors
11.3 Incomplete Block Designs for 2n Factorials
11.4 A General Method to Create Incomplete Blocks
11.5 Incomplete Block Designs for 3n Factorials
11.6 Concluding Remarks
11A Appendix: Incomplete Block Design Plans for 2n Factorials
12. Fractional Factorial Designs
12.1 Reduce Experiment Size with Fractional Treatment Designs
12.2 The Half Fraction of the 2n Factorial
12.3 Design Resolution Related to Aliases
12.4 Analysis of Half Replicate 2n-1 Designs
12.5 The Quarter Fractions of 2n Factorials
12.6 Construction of 2n-p Designs with Resolution III and IV
12.7 Genichi Taguchi and Quality Improvement
12.8 Concluding Remarks
12A Appendix: Fractional Factorial Design Plans
13. Response Surface Designs
13.1 Describe Responses with Equations and Graphs
13.2 Identify Important Factors with 2n Factorials
13.3 Designs to Estimate Second-Order Response Surfaces
13.4 Quadratic Response Surface Estimation
13.5 Response Surface Exploration
13.6 Designs for Mixtures of Ingredients
13.7 Analysis of Mixture Experiments
13A.1 Appendix: Least Squares Estimation of Regression Models
13A.2 Appendix: Location of Coordinates for the Stationary Point
13A.3 Appendix: Canonical Form of the Quadratic Equation
14. Split-Plot Designs
14.1 Plots of Different Size in the Same Experiment
14.2 Two Experimental Errors for Two Plot Sizes
14.3 The Analysis for Split-Plot Designs
14.4 Standard Errors for Treatment Factor Means
14.5 Features of the Split-Plot Design
14.6 Relative Efficiency of Subplot and Whole-Plot Comparisons
14.7 The Split-Split-Plot Design for Three Treatment Factors
14.8 The Split-Block Design
14.9 Additional Information About Split-Plot Designs
15. Repeated Measures Designs
15.1 Studies of Time Trends
15.2 Relationships Among Repeated Measurements
15.3 A Test for the Huynh–Feldt Assumption
15.4 A Univariate Analysis of Variance for Repeated Measures
15.5 Analysis When Univariate Analysis Assumptions Do Not Hold
15.6 Other Experiments with Repeated Measures Properties
15.7 Other Models for Correlation Among Repeated Measures
15A.1 Appendix: The Mauchly Test for Sphericity
15A.2 Appendix: Degrees of Freedom Adjustments for Repeated Measures Analysis of Variance
16. Crossover Designs
16.1 Administer All Treatments to Each Experimental Unit
16.2 Analysis of Crossover Designs
16.3 Balanced Designs for Crossover Studies
16.4 Crossover Designs for Two Treatments
16A.1 Appendix: Coding Data Files for Crossover Studies
16A.2 Appendix: Treatment Sum of Squares for Balanced Designs
17. Analysis of Covariance
17.1 Local Control with a Measured Covariate
17.2 Analysis of Covariance for Completely Randomized Designs
17.3 The Analysis of Covariance for Blocked Experiment Designs
17.4 Practical Consequences of Covariance Analysis
References
Appendix Tables
Answers to Selected Exercises
Index
|
){kind=link}
){kind=link}