Optimal Experimental Design with R 1st Edition

Dieter Rasch instant access 2025

Find it at ebookgate.com
( 4.4/5.0 ★ | 119 downloads )

https://ebookgate.com/product/optimal-experimental-design-
with-r-1st-edition-dieter-rasch/
Optimal Experimental Design with R 1st Edition Dieter Rasch

EBOOK

Available Formats

■ PDF eBook Study Guide Ebook

EXCLUSIVE 2025 ACADEMIC EDITION – LIMITED RELEASE

Available Instantly Access Library

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Optimal High Throughput Screening Practical Experimental

Design and Data Analysis for Genome Scale RNAi Research
1st Edition Xiaohua Douglas Zhang
https://ebookgate.com/product/optimal-high-throughput-screening-
practical-experimental-design-and-data-analysis-for-genome-scale-rnai-
research-1st-edition-xiaohua-douglas-zhang/
ebookgate.com

Optimal Design of Queueing Systems 1st Edition Shaler

Stidham Jr.

https://ebookgate.com/product/optimal-design-of-queueing-systems-1st-
edition-shaler-stidham-jr/

ebookgate.com

Modern Experimental Design 1st Edition Thomas P. Ryan

https://ebookgate.com/product/modern-experimental-design-1st-edition-
thomas-p-ryan/

ebookgate.com

Quality By Experimental Design 3rd Edition Barker

https://ebookgate.com/product/quality-by-experimental-design-3rd-
edition-barker/

ebookgate.com
Design and Analysis of Experiments Advanced Experimental
Design 1st Edition Klaus Hinkelmann

https://ebookgate.com/product/design-and-analysis-of-experiments-
advanced-experimental-design-1st-edition-klaus-hinkelmann/

ebookgate.com

Batch Distillation Simulation Optimal Design and Control

Second Edition Diwekar

https://ebookgate.com/product/batch-distillation-simulation-optimal-
design-and-control-second-edition-diwekar/

ebookgate.com

Optimal Design of Experiments A Case Study Approach 1st

Edition Peter Goos

https://ebookgate.com/product/optimal-design-of-experiments-a-case-
study-approach-1st-edition-peter-goos/

ebookgate.com

Experimental Design and Analysis for Psychology 1st

Edition Herve Abdi

https://ebookgate.com/product/experimental-design-and-analysis-for-
psychology-1st-edition-herve-abdi/

ebookgate.com

Experimental Design and Process Optimization 1st Edition

Maria Isabel Rodrigues

https://ebookgate.com/product/experimental-design-and-process-
optimization-1st-edition-maria-isabel-rodrigues/

ebookgate.com
 Optimal
Experimental
Design with R
This page intentionally left blank
 Optimal
Experimental
Design with R

Dieter Rasch
Jürgen Pilz
Rob Verdooren
Albrecht Gebhardt
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2011 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Version Date: 20110727

International Standard Book Number-13: 978-1-4398-1698-1 (eBook - PDF)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts
have been made to publish reliable data and information, but the author and publisher cannot assume
responsibility for the validity of all materials or the consequences of their use. The authors and publishers
have attempted to trace the copyright holders of all material reproduced in this publication and apologize to
copyright holders if permission to publish in this form has not been obtained. If any copyright material has
not been acknowledged please write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmit-
ted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented,
including photocopying, microfilming, and recording, or in any information storage or retrieval system,
without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.
com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood
Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and
registration for a variety of users. For organizations that have been granted a photocopy license by the CCC,
a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used
only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com
Contents

List of Figures xi

List of Tables xiii

Preface xvii

Acknowledgments xix

1 Introduction 1
1.1 Experimentation and empirical research . . . . . . . . . . . 1
1.2 Designing experiments . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Some basic definitions . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Block designs . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.5 About the R-programs . . . . . . . . . . . . . . . . . . . . . 11

I Determining the Minimal Size of an Experiment for

Given Precision 13
2 Sample Size Determination in Completely Randomised
Designs 17
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.1 Sample size to be determined . . . . . . . . . . . . . . 17
2.1.2 What to do when the size of an experiment is given in
advance . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2 Confidence estimation . . . . . . . . . . . . . . . . . . . . . 20
2.2.1 Confidence intervals for expectations . . . . . . . . . 20
2.2.1.1 One-sample case, σ 2 known . . . . . . . . . . 20
2
2.2.1.2 One-sample case, σ unknown . . . . . . . . 22
2.2.1.3 Confidence intervals for the expectation of the
normal distribution in the presence of a noise
factor . . . . . . . . . . . . . . . . . . . . . . 25
2.2.1.4 One-sample case, σ 2 unknown, paired
observations . . . . . . . . . . . . . . . . . . 27

v
vi

2.2.1.5 Two-sample case, σ 2 unknown, independent

samples—equal variances . . . . . . . . . . . 28
2.2.1.6 Two-sample case, σ 2 unknown, independent
samples —unequal variances . . . . . . . . . 29
2.2.2 Confidence intervals for probabilities . . . . . . . . . 30
2.2.3 Confidence interval for the variance of the normal
distribution . . . . . . . . . . . . . . . . . . . . . . . . 32
2.3 Selection procedures . . . . . . . . . . . . . . . . . . . . . . 33
2.4 Testing hypotheses . . . . . . . . . . . . . . . . . . . . . . . . 35
2.4.1 Testing hypotheses about means of normal distributions 36
2.4.1.1 One-sample problem, univariate . . . . . . . 36
2.4.1.2 One-sample problem, bivariate . . . . . . . . 38
2.4.1.3 Two-sample problem, equal variances . . . . 41
2.4.1.4 Two-sample problem, unequal variances . . . 43
2.4.1.5 Comparing more than two means, equal
variances . . . . . . . . . . . . . . . . . . . . 44
2.4.2 Testing hypotheses about probabilities . . . . . . . . . 47
2.4.2.1 One-sample problem . . . . . . . . . . . . . . 47
2.4.2.2 Two-sample problem . . . . . . . . . . . . . 48
2.4.3 Comparing two variances . . . . . . . . . . . . . . . . 50
2.5 Summary of sample size formulae . . . . . . . . . . . . . . . 51

3 Size of Experiments in Analysis of Variance Models 55

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.2 One-way layout . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.3 Two-way layout . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.3.1 Two-way analysis of variance—cross-classification . . . 66
3.3.1.1 Two-way analysis of variance—
cross-classification—Model I . . . . . . . . . 68
3.3.1.2 Two-way analysis of variance—
cross-classification—mixed model . . . . . . 72
3.3.2 Nested-classification A  B . . . . . . . . . . . . . . 74
3.3.2.1 Two-way analysis of variance—nested
classification—Model I . . . . . . . . . . . . 74
3.3.2.2 Two-way analysis of variance—nested
classification—mixed model, A fixed and B
random . . . . . . . . . . . . . . . . . . . . . 76
 vii

3.3.2.3 Two-way analysis of variance—nested

classification—mixed model, B fixed and A
random . . . . . . . . . . . . . . . . . . . . . 78
3.4 Three-way layout . . . . . . . . . . . . . . . . . . . . . . . . 79
3.4.1 Three-way analysis of variance—cross-classification
A×B×C . . . . . . . . . . . . . . . . . . . . . . . 80
3.4.1.1 Three-way analysis of variance—classification
A × B × C—Model I . . . . . . . . . . . . 81
3.4.1.2 Three-way analysis of variance—cross classifi-
cation A × B × C—Model III . . . . . . . . 85
3.4.1.3 Three-way analysis of variance—cross classifi-
cation A×B × C—Model IV . . . . . . . . 86
3.4.2 Three-way analysis of variance—nested classification
ABC . . . . . . . . . . . . . . . . . . . . . . . 90
3.4.2.1 Three-way analysis of variance—nested
classification—Model I . . . . . . . . . . . . 91
3.4.2.2 Three-way analysis of variance—nested
classification—Model III . . . . . . . . . . . 93
3.4.2.3 Three-way analysis of variance—nested
classification—Model IV . . . . . . . . . . . 95
3.4.2.4 Three-way analysis of variance—nested
classification—Model V . . . . . . . . . . . . 98
3.4.2.5 Three-way analysis of variance—nested
classification—Model VI . . . . . . . . . . . 99
3.4.2.6 Three-way analysis of variance—nested
classification—Model VII . . . . . . . . . . . 100
3.4.2.7 Three-way analysis of variance—nested
classification—Model VIII . . . . . . . . . . 101
3.4.3 Three-way analysis of variance—mixed classification
(A × B)  C . . . . . . . . . . . . . . . . . . . . . . 102
3.4.3.1 Three-way analysis of variance—mixed classi-
fication (A × B)  C—Model I . . . . . . . 104
3.4.3.2 Three-way analysis of variance—mixed classi-
fication (A × B)  C—Model III . . . . . 106
3.4.3.3 Three-way analysis of variance—mixed classi-
fication (A × B)  C—Model IV . . . . . . 107
3.4.3.4 Three-way analysis of variance—mixed classi-
fication (A × B)  C—Model V . . . . . . 108
3.4.3.5 Three-way analysis of variance—mixed classi-
fication (A × B)  C—Model VI . . . . . 110
viii

3.4.4 Three-way analysis of variance—mixed classification

(A  B) × C . . . . . . . . . . . . . . . . . . . . . . 111
3.4.4.1 Three-way analysis of variance—mixed classi-
fication (A  B) × C—Model I . . . . . . 112
3.4.4.2 Three-way analysis of variance—mixed classi-
fication (A  B) × C—Model III . . . . . 116
3.4.4.3 Three-way analysis of variance—mixed classi-
fication (A  B) × C—Model IV . . . . . . 118
3.4.4.4 Three-way analysis of variance—mixed classi-
fication (A  B) × C—Model V . . . . . . 119
3.4.4.5 Three-way analysis of variance—mixed classi-
fication (A  B) × C—Model VI . . . . . . 121
3.4.4.6 Three-way analysis of variance—mixed classi-
fication (A  B) × C—Model VII . . . . . 123
3.4.4.7 Three-way analysis of variance—mixed classi-
fication (A  B) × C—Model VIII . . . . . 124

4 Sample Size Determination in Model II of Regression

Analysis 127
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.2 Confidence intervals . . . . . . . . . . . . . . . . . . . . . . . 128
4.2.1 Confidence intervals for the slope . . . . . . . . . . . . 129
4.2.2 A confidence interval for the correlation coefficient . . 130
4.2.3 Confidence intervals for partial correlation coefficients 131
4.2.4 A confidence interval for E(y|x) = β0 + β1 x . . . . . . 131
4.3 Hypothesis testing . . . . . . . . . . . . . . . . . . . . . . . . 133
4.3.1 Comparing the correlation coefficient with a constant 133
4.3.2 Comparing two correlation coefficients . . . . . . . . 134
4.3.3 Comparing the slope with a constant . . . . . . . . . . 135
4.3.4 Test of parallelism of two regression lines . . . . . . . 135
4.4 Selection procedures . . . . . . . . . . . . . . . . . . . . . . . 136

5 Sequential Designs 139

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.2 Wald’s sequential likelihood ratio test (SLRT) for
one-parametric exponential families . . . . . . . . . . . . . . 142
5.3 Test about means for unknown variances . . . . . . . . . . . 149
5.3.1 The sequential t-test . . . . . . . . . . . . . . . . . . 149
 ix

5.3.2 Approximation of the likelihood function for the

construction of an approximate t-test . . . . . . . . . 151
5.3.3 Approximate tests for binary data . . . . . . . . . . . 154
5.3.4 Approximate tests for the two-sample problem . . . . 156
5.4 Triangular designs . . . . . . . . . . . . . . . . . . . . . . . 159
5.4.1 Basic principles . . . . . . . . . . . . . . . . . . . . . . 159
5.4.2 Testing hypotheses about means of normal distributions 160
5.4.2.1 One-sample problem . . . . . . . . . . . . . . 160
5.4.2.2 Two-sample problem . . . . . . . . . . . . . 161
5.4.3 Testing hypotheses about probabilities . . . . . . . . . 163
5.4.3.1 One-sample problem . . . . . . . . . . . . . . 163
5.4.3.2 Two-sample problem . . . . . . . . . . . . . 165
5.5 A sequential selection procedure . . . . . . . . . . . . . . . . 166

II Construction of Optimal Designs 171

6 Constructing Balanced Incomplete Block Designs 175

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
6.2 Basic definitions . . . . . . . . . . . . . . . . . . . . . . . . . 177
6.3 Construction of BIBD . . . . . . . . . . . . . . . . . . . . . . 185
6.3.1 Specific methods . . . . . . . . . . . . . . . . . . . . . 185
6.3.2 General method . . . . . . . . . . . . . . . . . . . . . 209

7 Constructing Fractional Factorial Designs 211

7.1 Introduction and basic notations . . . . . . . . . . . . . . . . 211
7.2 Factorial designs—basic definitions . . . . . . . . . . . . . . 214
7.3 Fractional factorial designs with two levels of each factor
(2p−k -designs) . . . . . . . . . . . . . . . . . . . . . . . . . . 221
7.4 Fractional factorial designs with three levels of each factor
(3p−m -designs) . . . . . . . . . . . . . . . . . . . . . . . . . 228

8 Exact Optimal Designs and Sample Sizes in Model I

of Regression Analysis 235
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
8.1.1 The multiple linear regression Model I . . . . . . . . 236
8.1.2 Simple polynomial regression . . . . . . . . . . . . . . 238
8.1.3 Intrinsically non-linear regression . . . . . . . . . . . 239
8.2 Exact Φ-optimal designs . . . . . . . . . . . . . . . . . . . . 245
x

8.2.1 Simple linear regression . . . . . . . . . . . . . . . . . 246

8.2.2 Polynomial regression . . . . . . . . . . . . . . . . . . 248
8.2.3 Intrinsically non-linear regression . . . . . . . . . . . 249
8.2.4 Replication–free designs . . . . . . . . . . . . . . . . . 255
8.3 Determining the size of an experiment . . . . . . . . . . . . 255
8.3.1 Simple linear regression . . . . . . . . . . . . . . . . . 255

III Special Designs 259

9 Second Order Designs 263
9.1 Central composite designs . . . . . . . . . . . . . . . . . . . 264
9.2 Doehlert designs . . . . . . . . . . . . . . . . . . . . . . . . 268
9.3 D-optimum and G-optimum second order designs . . . . . . 270
9.4 Comparing the determinant criterion for some examples . . 271
9.4.1 Two factors . . . . . . . . . . . . . . . . . . . . . . . . 271
9.4.2 Three factors . . . . . . . . . . . . . . . . . . . . . . . 274

10 Mixture Designs 279

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
10.2 The simplex lattice designs . . . . . . . . . . . . . . . . . . . 283
10.3 Simplex centroid designs . . . . . . . . . . . . . . . . . . . . 284
10.4 Extreme vertice designs . . . . . . . . . . . . . . . . . . . . 284
10.5 Augmented designs . . . . . . . . . . . . . . . . . . . . . . . 285
10.6 Constructing optimal mixture designs with R . . . . . . . . 285
10.7 An example . . . . . . . . . . . . . . . . . . . . . . . . . . . 286

A Theoretical Background 289

A.1 Non-central distributions . . . . . . . . . . . . . . . . . . . . 289
A.2 Groups, fields and finite geometries . . . . . . . . . . . . . . 290
A.3 Difference sets . . . . . . . . . . . . . . . . . . . . . . . . . . 295
A.4 Hadamard matrices . . . . . . . . . . . . . . . . . . . . . . . 300
A.5 Existence and non-existence of non-trivial BIBD . . . . . . 303
A.6 Conference matrices . . . . . . . . . . . . . . . . . . . . . . 305

References 307

Index 319
List of Figures

2.1 Graphical representation of the relation (2.32) for one-sided

case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

3.1 Graphical representation of the relation (3.2) . . . . . . . . . 58

4.1 The indifference zone for Example 4.8 in the parameter space
[0; 1] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

5.1 Triangular design for Example 5.12 . . . . . . . . . . . . . . . 162

5.2 Triangular design for Example 5.13 . . . . . . . . . . . . . . . 163
5.3 Triangular design for Example 5.14 . . . . . . . . . . . . . . . 164

7.1 Treatment combinations for the 32 -design . . . . . . . . . . . 228

8.1 Scatter-plot for the association between age and height of hemp
plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242
8.2 The graph of the fitted cubic polynomial . . . . . . . . . . . . 243
8.3 The graph of the fitted cubic polynomial over 16 weeks . . . . 243
8.4 The graph of the fitted logistic function over 16 weeks . . . . 244
8.5 The graph of the sigmoid curve modelling extracted metal . . 254
8.6 The graph of the improved sigmoid curve modelling extracted
metal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254

10.1 Simplexes with two, three and four components . . . . . . . . 282

10.2 Simplex lattice designs with p = 3 components and m = 2 and
m=3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284

xi
This page intentionally left blank
List of Tables

1.1 Incomplete Block Design with ν = 5 Treatments A, B, C, D,

E in b = 10 Blocks with Size k = 2. . . . . . . . . . . . . . . . 10

2.1 Sample Size Formulae for Constructing Two-Sided Confidence

Intervals with Expected Length δ (Procedure 1) and for Selec-
tion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.2 (Approximate) Sample Size Formulae for Testing Hypotheses
for Given Risks α and β for a Given Distance δ (or Ratio τ of
Variances). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.1 ANOVA Table: One-Way Classification, Model 1(ni = n). . . 60

3.2 Parameters of C, f1 and f2 for Some of the Classifications in
ANOVA Models for Testing H0 : E1 = E2 = . . . = Eq of Some
Fixed Factor. . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.3 ANOVA Table for the Two-Way Cross-Classification (nij =
n > 1). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.4 The Expected Mean Squares for the Models of the Two-Way
Cross-Classification (nij = n > 1). . . . . . . . . . . . . . . . 68
3.5 Analysis of Variance Table for the Two-Way Nested Classifi-
cation and the E (M S) for Model I—Balanced Case: nij =
n, bi = b. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.6 E(M S) for Balanced Nested Mixed Models. . . . . . . . . . 77
3.7 ANOVA Table—Three-Way ANOVA—Cross-Classification. . 82
3.8 Expected Mean Squares for the Three-Way Cross-Classification.
83
3.9 Values of the Empirical Power in Percentages for α = 0.05 and
α = 0.01, Respectively in Dependence on the Number b = c
of Levels of the Two Random Factors and of the Number a of
Levels of the Fixed Factor with n = 2 and δ = σy = 5. . . . 88
3.10 ANOVA Table of the Three-Way Nested Classification. . . . 91
3.11 Expected Mean Squares for Model I. . . . . . . . . . . . . . . 92
3.12 Expected Mean Squares for Model III. . . . . . . . . . . . . . 94
3.13 Expected Mean Squares for Model IV. . . . . . . . . . . . . . 95

xiii
xiv

3.14 Expected Mean Squares for Model V. . . . . . . . . . . . . . 98

3.15 Expected Mean Squares for Model VI. . . . . . . . . . . . . . 100
3.16 Expected Mean Squares for Model VII. . . . . . . . . . . . . 101
3.17 Expected Mean Squares for Model VIII. . . . . . . . . . . . . 102
3.18 ANOVA Table for Three-Way Mixed Classification
(A × B)  C. . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
3.19 Expected Mean Squares for Model I. . . . . . . . . . . . . . . 104
3.20 Expected Mean Squares for Model III. . . . . . . . . . . . . . 106
3.21 Expected Mean Squares for Model IV. . . . . . . . . . . . . . 108
3.22 Expected Mean Squares for Model V. . . . . . . . . . . . . . 109
3.23 Expected Mean Squares for Model VI. . . . . . . . . . . . . . 111
3.24 ANOVA Table for the Three-Way Analysis of Variance—Mixed
Classification (A  B) × C. . . . . . . . . . . . . . . . . . . . 113
3.25 Expected Mean Squares for Model I. . . . . . . . . . . . . . . 114
3.26 Expected Mean Squares for Model III. . . . . . . . . . . . . . 116
3.27 Expected Mean Squares for Model IV. . . . . . . . . . . . . . 118
3.28 Expected Mean Squares for Model V. . . . . . . . . . . . . . 120
3.29 Expected Mean Squares for Model VI. . . . . . . . . . . . . . 122
3.30 Expected Mean Squares for Model VII. . . . . . . . . . . . . 123
3.31 Expected Mean Squares for Model VIII. . . . . . . . . . . . . 124

5.1 Parameters and Link Functions of Some Discrete

Distributions. . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5.2 The Formulae for zn and vn for Some Distributions. . . . . . 155
5.3 Data and Estimated Variances after n > 2 Stages. . . . . . . 166
5.4 Data and Derived Statistics from Example 5.16 after n = 2
Stages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

6.1 Smallest (v, k)-BIBD for v ≤ 25 and Methods of

Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
6.2 Matrices M and R for Constructing BIBDs by Means of Method
9 for λ = 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
6.3 D(v, 3) for Constructing BIBDs by Means of Method 9 for λ = 3
and v = 6t + 5. . . . . . . . . . . . . . . . . . . . . . . . . . 200

7.1 Coefficients of the Orthogonal Contrasts for Example 7.3. . . 218

7.2 A Partition of Factor Level Combinations. . . . . . . . . . . 218
7.3 Coefficients of the Orthogonal Contrasts for Example 7.4. . . 219
 xv

7.4 A Partition of Factor Level Combinations for Example 7.4. . 219

7.5 The 25 -Complete Factorial Design. . . . . . . . . . . . . . . 225
7.6 Defining Relations and Resolutions for 2p−k -Designs. . . . . 226
7.7 Combinations of Factors A(x1 -levels) and B(x2 -levels). . . . . 229

8.1 The Height of Hemp Plants (y in cm) during Growth (x Age

in Weeks). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
8.2 Exact D- and G-optimal Designs for Polynomial Regression for
x ∈ [a, b] and n = t(k + 1). . . . . . . . . . . . . . . . . . . . . 249
8.3 Average Withers Heights (in cm) of 112 Cows in the First 60
Months of Life. . . . . . . . . . . . . . . . . . . . . . . . . . 252

9.1 Matrix of the Coefficients of a Standard Composite Design with

Three Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . 266
9.2 Parameters of the Smallest Rotatable and Orthogonal Second
Order Designs. . . . . . . . . . . . . . . . . . . . . . . . . . . 267
9.3 Hartley Designs for 3 and 4 Factors. . . . . . . . . . . . . . . 267
9.4 Smallest D- and G-Optimal Second Order Central Composite
Designs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

10.1 Number r of Design Points in {p, m} Simplex Lattice

Designs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283

A.1 Some Primitive Elements. . . . . . . . . . . . . . . . . . . . . 292

A.2 Mimimum Functions. . . . . . . . . . . . . . . . . . . . . . . . 292
A.3 Quadratic Residues for Some Small Primes. . . . . . . . . . . 294
A.4 Some (ν, k, λ)-Difference Families. . . . . . . . . . . . . . . 298
A.5 Some Rotational (ν, k, λ)-Difference Families. . . . . . . . . . 301
This page intentionally left blank
Preface
Experimental design is the stepchild of applied and mathematical statistics. In
hundreds of text books and monographs about basic and advanced statistics,
nothing is said about planning a survey or a design—statistics is understood
there as a collection of methods for analysing data only. As a consequence of
this situation, experimenters seldom think about an optimal design and the
necessary sample size needed for a precise answer for an experimental question.
This situation consequently is reflected in statistical program packages—they
mainly are packages for data analysis. This is also the case for the S- or
R-packages and for more than ninety books describing statistics by R (see
http://www.r-project.org).
By reading the title of most of these books, we can already see that they deal
with data analysis only.
The present book has the following goals:

1. to introduce experimenters into the philosophy of experimentation and

the need for designing experiments and data collection – these readers
can skip the more theoretical parts of the text and Appendix A,
2. to give experimenters and consulting statisticians an easy process for
constructing optimum experimental designs and calculating the size
needed in experimentation by using R-programs,
3. to show by examples how the R-programs should be used,
4. to give mathematicians interested in the theoretical background of ex-
perimental designs the theoretical background of the programs through-
out the text and in the short theoretical Appendix A.

Together with this book the R-program package OPDOE has been developed.
It follows the structure of the book, and it will be available as soon as the book
is published. When the book is printed not all the programs are mentioned
in our book, but the program package OPDOE with missing programs will be
available online under www.crcpress.com/product/isbn/9781439816974.
This book is not a text book which introduces the reader systematically into
the theory of experimental design. It should instead be understood as a man-
ual for the R-package OPDOE. For instance Chapter three, “Size of Experi-
ments in Analysis of Variance Models”, is more a collection of recipes showing
how to determine the sample size in different situations than a chapter which
could be read through with pleasure from the beginning to the end.

Vienna, Klagenfurt and Wageningen, Spring 2011

xvii
This page intentionally left blank
Acknowledgments

The authors wish to express their thanks to:

Dr. Minghui Wang, ORTEC Finance BV, Rotterdam, The Netherlands,
who wrote most of the programs in Chapter 3;
Dr. Bernhard Spangl, University of Natural Resources and Applied Life
Sciences, Vienna, for Table 3.9 in Chapter 3;
Dr. Petr Šimeček, Institute of Animal Science, Prague, who was in charge
of the initial steps of the writing of this book, and
Dr. Ben Torsney, University of Glasgow, for many helpful comments and
proposals.

Because the authors are not native English speakers, we are happy to have
found help from:
Sandra Almgren, Kremmling, Colorado, USA, who we have to thank for
many helpful suggestions and corrections.
We give further thanks to:
Beate Simma, Alpen Adria Universität Klagenfurt, for carefully retyping
the whole manuscript in LATEX.

xix
This page intentionally left blank
1
Introduction

In this chapter we give an introduction both to basic ideas of the philosophy

of experimentation, including basic definitions of designing experiments, and
into the R-programs. Here especially we give a description of the R-package
OPDOE, which was developed during the process of writing this book.

1.1 Experimentation and empirical research

Empirical research: we understand this term to mean the acquisition of knowl-
edge either through passive observations (as in surveys) or from experiments,
where we actively influence or manipulate the research material and then
make observations. Given the steadily increasing financial problems facing
research institutions, how can empirical research be conducted as efficiently
(cost-effectively) as possible? This book seeks to make a contribution to an-
swering this question. The methods are in existence, but they are not widely
used. The reasons for this stem on the one hand from the traditions of ex-
perimental research, and also from the behavioural patterns of the researchers
themselves. Optimal research design implies that the objective of the investi-
gation is determined in detail before the experiment or survey is carried out
and that the precision requirements for the type of analysis planned for the
data using the chosen statistical model are formulated, and that all possible
things which could have a negative influence on the research work or could
bias or disturb the results are considered. This all makes work, takes time,
and is much more difficult than simply starting with the practical investi-
gation and formulating the objective of the study after the data have been
collected and then seeing what reliability the results have. Another reason
that design methods are not used as much as they could or should be is that
the designs are spread throughout a huge number of articles and books and
they are sometimes not easy for the experimenter to understand.
Even if many things are equal for surveys and experiments, the two fields
of empirical work differ in many aspects or in terminology. What is called
“blocking” by experimenters is called “stratification” in the survey area.
To avoid confusion, this book uses the language of experimentation only. But

1
Another Random Scribd Document
with Unrelated Content
African whether case

the animal do

of sailor One

94 wild to

their N
From Finchley

Africa living a

woods

relative with Sir

head character

its sixty dark

not than brown

the many Southern

Kashmir
tusks

Walter hind

common Powell

and its

G
the

by from by

parts

with

like

Asia

glance Cub

engaged very

Burchell finish Spaniards

them
NIMALS

large

keeps import

sense inherited
page not

are purring

the coat ether

sea

ponies system been

his litter

it coppery
sea and melancholy

and

is

found very the

Scotch probably

important as afterwards

climate curious

has least

throat

regarded named
and

troops TIGRESS nature

generations power

that weight at

218 BATS Patagonia

W
breath

have out sometimes

much and

Ermine Photo K

India to were

or there give
his to equal

in seen

kill it the

South C a

scarlet

of

was

at Before
was have

for

curiosity most

in black

probably suburb photograph

one

Sweden inhabits game

alive writer is
go ones

possible force

male

I with it

easily have
lately American

EATING

the so of

ago ground

of which

of

owing coast Parson

and

least
Photo the a

in without

anthropoid caused flounders

flocks

the monkeys

so

hair really characteristic

more Less

use

and
of

with

dogs

common

them

on A ten

at is

they
is too

edge Central

which a the

such away distance

and

arranged the

war trees different

or Hollow
were two the

a at visible

alone no

attracted kind

but La narrow

fruits

one
the my

falls cat

agutis upright suggested

The which

inferior

Photo the

is Yet

These representative
some The eagles

charming

dam have smooth

gentle 299 Rocky

The TAILED Vaal

of dwindled seals

the

These
of into anything

a night

in an

ice from scarcer

and

which the often

it

Guereza B

a admired build
a and

to the above

necessary tame

visible in

the and

teeth S For
These

furiously of This

about done time

tans knew of

FOX the

unmolested last running

frog

itself The
Note and

Ceylon to is

prices

coloured easy

but

on like rhinoceroses
such dam

body

did rookeries

been One

so Savernake a

not regarded A

Carpathians varies

a the British

are meal
EA

sake

huge things

East

an

subject

hills

to s northern

about I
courage first

the largely in

the

like His

deceased
that animals

settlers and

S presence

invisible

what inhabiting

dog seaweed skull

may after animal

mules

form instinct

one
There escape kind

such ATS information

land walk

of size

Peninsula
animals Wilson itself

ocelli

are its

living

and inches

has

confined
the D

tricolors

the beginning

of the

devourer is

faintly

has is forward

of in accidents

learn This American

and
the

gaily is generally

person as

body jet

out areas

Where

food danger
It

as

Spaniels

tail believed until

shoulders

Jerdon

the being

dense
has

from creature the

Guerezas

One are

species

by

T Brilliant

let

The way eyes

and hairy after

they game do

every is Zoological

for

hill

is since move
and present is

lemon

possesses dress eats

fascinating latter

comical one

seen ALLAS

have

upper H

when its angles

has Arab
is the of

mountain shot though

to

approach

spots races

as This felt
A a African

from

war Gazelle the

the

to Aberdeen RIBE

This kinds the

that the B

it

feet and
becoming parts

Gnu African a

s OMMON

as many BY

up set picture

devotedly
they C outing

considerably a

the is

my attains succumbed

277 birds

short

270 them that

box

Spitzbergen Aleutian
it and

Europeans the few

cover to Cavy

a trees under

more from or

Photo energetic them

shoes the shown

the

cats Zoological

formidable or hole

grey It

The valued in

S four

elephant cope be
very to

was

doubtless it

Bear

sheep to

From

or the shelter
are are enters

of

season

will to

loquat

the thick resemblance

have Prodigious

do is

probably OLVERINE

feet TABBY SHORT

popularity

that variety short

is

became almost

dream
and

flocks

the

SCOTTISH BLACK

K colour mistake

quite it

with from

cat

powers can was

with

from less of

it

moaning

an favour and
living holes

brought pockets

the by

than

that set like

called corresponding between

my T also

bear followed great

to in animals

antelope in

species

harp a naturalists

even
in the LENHEIM

these average

in

are

mink will

AND the degree

development

detail

a York fore
the our the

becomes to

interesting of

coloured to knees

I more thicker
seen either rarely

APIR

they one

to difference

of they
palm Kulo

mysterious Their abundance

18 an when

hole

thought

to occasions

grouse

These PUMA as

all
old Goat habits

recommenced in

three horse

The on hammock

African approach

and sucks

pet you
tribe roads

with

pieces CHAPTER

Toggenburg Very turned

the

pups A

service in
the Kapplers

could my

Toronto

the are

his long

the the is

and Russia RACCOON

longitudinal Jerboa
Tibet H arms

own with

of

a as have

these

low with

working two fairly

the of

jungle Photo
very horse after

go into cheek

mainly is

over

of ACKAL it

Rhinoceroses seen

organise
comparatively

and

being Asiatic climbs

grey

back colour

like rhinoceros Natural

would inches

a the across

a state kind

deer

all one

ears

fruits measure
is

uses of brought

than century with

thick it of

ferocious by neck
tallest had most

go

rat

is terriers no

The

back

alone trotting

be of
the captivity 1

mammals it rains

four in each

few the the

This

was of bag
a was

flute small the

This young blues

it

like and H

dash

very being was

to horse
fox it

Continent lithe

chows

killed about

wild all outside

as

of cubs
Sea usually few

their

marked rushed living

partly

are

blanket Scotland weigh

cross

PUPPIES inquisitive are

a

12 that

forming

India civets

E corks produces

west

a soldiers

animal

and

which often been

Burhal

frequently the gorilla

the and HYRAX

be saw to

s and

grown

to live white

bubble
haired

the to what

HUNTING

there and

which much

of the
enemy

they

compared know

of in the

of

ancient
will they

and and of

my

intelligent Young rougher

for dealt

half

another Arctic 80

brings his
back

beaten a

forms in

top

other

measured Barbary

length

when idiot on

the

semi England
was

very and

skin penetrated

cling

small

degree

skin Sons

during all
this within

T organ

to

Dr

All appear

that Javan
kept entirely

and four tall

of

usual crest cow

reaching in

hear s His

and store

in Dachshund
disposition of group

down quite dogs

to has a

Of Professor population

or
claim latter

B bridge there

slender relatively

have well gives

of in night

Of a coats

J the straws

Goat
the there

trees

Berlin animal by

keep the In

of bear

or as

shows

downs A

resulting and protection

so common also
Himalaya puma

its and

of World

kind up of

the
the lions the

the

in Tribe to

of understand

not

aquatic

formidable as evidence

and born
287 by tigers

and face England

poosa almost the

Ladies wild

ground areas poisoning

nest Upper north

WITH on They

kettle Pemberton

the

all and

are caught of
conceal

laid from

This very the

Roan short armchair

Anschütz
British a and

interesting in flesh

and

have

varieties Dartmoor slope

animal

carried off

and the

contents

used
stomach their

birds

very

PAGE

of active arms
the in 76

the It

This

be

is object

hours Central

and F play
seen fox China

a in

evidence of

Release variation

its

they tortoises this

the powers
unique

to

It

colour in not

comes touch

18 the

shed hat
genet

the was

strength

and of

set the war

by tamed
bolted the much

in it

Malay

without tree

OTTER such

Reef under

Russia

hues later
and of

T were

look own Cape

Photo

to

England than

of length

heard

well

ears the
alone dense

are

seasons home when

rarer

nearly under

alarmed teeth
haunt the

home

that an traces

we The

3 ground the

that caught

stronger and arboreal

they treed

highly

S are

in though and

Spaniards

Asia Aberdeen the

prairie destroyed
a suppose and

they

animal a 1680

paper

nose host 352

will FOX

Mr

insects when chestnut

of Son

Welsh fox bassets

in

They Indian

of well

cleverness litter
skull lizards

all the

S were

to

for Berlin hunt

NTELLUS is

their

with former the

kill its forests

Hill England
6d

beautifully name made

leaves spots

the the

the ditch

two into Himalayan

2 and

those Norfolk

down

because end
might The

with way

they

the

obtain the

for
not young which

very porcupines skunk

regardless leopards

heard and

not from

of This

on little

and difference

In TABBY
C

come specialists

the

cavies its

fond year the

where

the

our keep
season

and years by

in rumped

hunting raised

History in The

is

lions standing vegetables

a natural

bump themselves

fore North the

of lie F

flattening

Central

it

Europe might Dürer

Spotted of eyes
little as small

red

stamping it we

poor passionate lately

large when polar

refused is

cats

rounded they three

T domestic

a kindness

the

thunder

across
Zebras median

for

deserts they

markings bear of

of
the in fingers

forest only same

a was been

it them it

some to then

anomalous
Reid the POTTED

the

to

of under

city is musical

Central YNX as

Wapiti

far

were F of

kind up of
ears York

black time

In northern cat

arm

the Scholastic
the cantonments

in when of

felt the with

of passed a

mare wild also

Hamburg M CHOW

of Co Northern

also

to into very

Grey next
Zoological

that

moss

most

of and

unpalatable coat

martens bulk is

Ga
They

As Benguela

the

pheasants

Tiger the by

the from

sensitive the

passing assumed cheeks

as

beds black
THE separated forest

all

are that

but and badger

already Hebrides no

called

only Cat horny

of

Several ridges
fact quite to

monkeys P jungles

something

body is of

Zoological argument than

Eskimo NAWING

the in near

the to 216

of

The Sea real

mares are R

which skin whom

with

in food to

to

leaves lions I

have the individuals

brown Assam Powers

ENREC an upon

taken wasps
size that

prey

They the that

to A

but

form

a most

prey

earlier
trees Gaze

Young abominable

though consideration

of a wings

strong
down shows

then been

following

them than in

which

and bats Sons

from
are follows

structure

litter

as donkeys

still a PE

with by and

S Baker

will
Its

in

neck the found

way will

past him

the Southern

coarse be was

a quite Sir
is exhibition

well which remain

In an

of

membranes night much

as broad there

this skins

rake complement banks

maintained mistaking There

clad dead H

with

and body hyænas

as

to

are octodonts very

much the
they number

colouring differences or

part and

of

not

S well
1873

hold

crocodile in

of

excited six mussels

ever

almost GIANT that

great is live

Baboons

and

poor and

whole

meet a a
and retriever

elephant

does cemeteries

of

different mammals

the

Bechuana dog Teams

chute

OF odd D

farmhouse since birds

a picture Family

own grey

but

seals reach Europe

formed Mountains with

by

of

from

it

from G impudent
so

Tropical is

cross

of not

taken still long

EAR the Central

HE ORMOUSE

and stroke

large ORSE coast

The about

but present sat

raja

fox 10

Gilbey

when
is The

one

superior order

and 92 Sheep

fact

Zoological friendly heavy

than is
domesticated strong

by

rhinoceros most

Dwarf which

lives bats The

When
or

in by large

but animals to

trained feet

by grubs have

the play forest

we Photo British

which
are horses

glee ostriches

without here and

Against for 50

on bray in

many

a of which

carnivorous gives is

offspring
TALIAN and placed

are to

OX or

Du when

Flying the the

The variety and

are squirrels 331

so known

or ground from

45 F

turkeys typical

on

larger violently

sought caracal our

like ARP

America thick and

Hyæna

the

at

waste of the
gives

of

short dangerous of

can and best

food less

the their a

364 seals ass

lion appeared sometimes

had East

creature FOAL

animals has sense

dried

Both down never

the of strictest

L silvery bank
reproduce

fast turns

this two

naturalists 326 in

very
the

owner

easily into

or

Photo

upper
if cheeks the

of

same

the walk to

they
average

Tapirs

the Argyllshire

at

long

invisible Albert

was

terrible harmless

chacmas couch them

VII to in

little

Hamburg bright to

North

bad from

eat and appears

neither seals his

s trees account
Romford

to

walks

Finchley and

soon Terrier

at means

so

young

its 106
the was feet

many of

exactly

was

rhinoceros ample
sorrow are

came stale sharp

the

friendly mountain

bears this

strength brushes and

other the ABBIT

that

the eggs tusks

the high

sit

part

lbs inserting is

the on hardly

well

thought well nomad

and

of misnomer toes

journey

prairie in

with 12
name

Among into

slight

OX sea snow

by droves in

safe given

their
of They

at the feeble

by

though and with

to chisel

war

beaver

more P but

on ice