50% found this document useful (10 votes)

6K views

Multivariate Statistical Inference and Applications

This netLibrary ebook does not include the ancillary media that was packaged with the original printed version of the book. No part of this publication may be reproduced, stored in a retrieval system or transmitted by any means.

Uploaded by

Naveen Kumar Singh

Available Formats

Download as PDF, TXT or read online on Scribd

50% found this document useful (10 votes)

6K views

Multivariate Statistical Inference and Applications

Uploaded by

Naveen Kumar Singh

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 634

cover

next page >

title author publisher isbn10 | asin print isbn13 ebook isbn13 language subject publication date lcc ddc subject

: : : : : : : : : : :

< previous page

page_iii

next page >

file:///G|/page_iii.html7/11/2009 2:11:57 PM

page_iv

< previous page

Disclaimer:

1 1

1.1. Introduction

1.2. Univariate and Bivariate Random Variables

1.2.1. Univariate Random Variables

1.2.2. Bivariate Random Variables

1.3. Mean Vectors and Covariance Matrices for Random Vectors

1.4. Correlation Matrices

1.5. Partitioned Mean Vectors and Covariance Matrices

1.6. Linear Functions of Random Variables

1.6.1. Sample Means, Variances, and Covariances

1.6.2. Population Means, Variances, and Covariances

1.7. Measuring Intercorrelation

1.8. Mahalanobis Distance

1.9. Missing Data 1.10. Robust Estimators of and 2. The Multivariate Normal Distribution

file:///G|/page_v.html (1 of 2)7/11/2009 2:11:57 PM

page_v

2.1. Univariate and Multivariate Normal Density Functions

2.1.1. Univariate Normal

2.1.2. Multivariate Normal

2.1.3. Constant Density Ellipsoids

2.1.4. Generating Multivariate Normal Data

2.1.5. Moments

2.2. Properties of Multivariate Normal Random Vectors

2.3. Estimation of Parameters in the Multivariate Normal Distribution Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

< previous page

page_v

next page >

file:///G|/page_v.html (2 of 2)7/11/2009 2:11:57 PM

page_vi

< previous page

page_vi

next page >

Page vi 49

2.3.1. Maximum Likelihood Method

2.3.2. Properties of y *and S

2.3.3. Wishart Distribution

2.4. Additional Topics 3. Hotelling's T2-Tests

60 60

3.1. Introduction 3.2. Test for H0: = 0 with Known 3.3. Hotelling's T2-test for Ho: = 0 with Unknown 3.3.1. Univariate t-Test for H0: = 0 with 2 Unknown 3.3.2. Likelihood Ratio Method of Test Construction

3.3.3. One-Sample T2-Test

3.3.4. Formal Definition of T2 and Relationship to F

3.3.5. Effect on T2 of Adding a Variable

3.3.6. Properties of the T2-Test

3.3.7. Likelihood Ratio Test

3.3.8. Union-Intersection Test

3.4. Confidence Intervals and Tests for Linear Functions of

file:///G|/page_vi.html (1 of 2)7/11/2009 2:11:58 PM

page_vi

3.4.1. Confidence Region for

3.4.2. Confidence Interval for a Single Linear Combination a'

3.4.3. Simultaneous Confidence Intervals for j and a'

3.4.4. Bonferroni Confidence Intervals for j and a'

3.4.5. Tests for Ho: a' = a'0and H0: j= 0j

3.4.6. Tests for H0:C = 0 3.5. Tests of H0: 1 = 2 Assuming 1= 2 3.5.1. Review of Univariate Likelihood Ratio Test for H0: 1 = 2 When 2/1 = 2/2 3.5.2. Test for H0: 1= 2 When 1 = 2 3.5.3. Effect on T2 of Adding a Variable

3.5.4. Properties of the Two-Sample T2-Statistic

3.5.5. Likelihood Ratio and Union-Intersection Tests

3.6. Confidence Intervals and Tests for Linear Functions of Two Mean Vectors

3.6.1. Confidence Region for 1 2

3.6.2. Simultaneous Confidence Intervals for a'(1 2) and 1j 2j

3.6.3. Bonferroni Confidence Intervals for a'(1 2) and 1j - 2j Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

< previous page

page_vi

next page >

3.12.2 NonparametricTests

113

3.12.3 Robust Versions of T2

114

3.13. Application of T2 in Multivariate Quality Control 4. Multivariate Analysis of Variance

114

121 121

4.1. One-Way Classification

4.1.1. Model for One-Way Multivariate Analysis of Variance

121

4.1.2 Wilks' Likelihood Ratio Test

122

4.1.3. Roy's Union-Intersection Test

127

4.1.4. The Pillai and Lawley-Hotelling Test Statistics

130

4.1.5. Summary of the Four Test Statistics 4.1.6. Effect of an Additional Variable on Wilks' 4.1.7. Tests on Individual Variables

131

132

134

4.2. Power and Robustness Comparisons for the Four MANOVA Test Statistics

135

4.3. Tests for Equality of Covariance Matrices

138

4.4. Power and Sample Size for the Four MANOVA Tests

140

4.5. Contrasts Among Mean Vectors

142

4.5.1. Univariate Contrasts

142

4.5.2. Multivariate Contrasts Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

145

file:///G|/page_vii.html (2 of 3)7/11/2009 2:11:58 PM

page_vii

< previous page

page_vii

next page >

file:///G|/page_vii.html (3 of 3)7/11/2009 2:11:58 PM

page_viii

< previous page

page_viii

next page >

Page viii 148

4.6. Two-Way Multivariate Analysis of Variance

4.7. Higher Order Models

151

4.8. Unbalanced Data

152

4.8.1. Introduction

152

4.8.2. Univariate One-Way Model

153

4.8.3. Multivariate One-Way Model

155

4.8.4. Univariate Two-Way Model

160

4.8.5. Multivariate Two-Way Model

168

4.9. Tests on a Subvector

174

4.9.1. Testing a Single Subvector

174

4.9.2. Step-Down Test

177

4.9.3. Stepwise Selection of Variables

177

4.10. Multivariate Analysis of Covariance

178

4.10.1. Introduction

178

4.10.2. Univariate Analysis of Covariance: One-Way Model with One Covariate

179

4.10.3. Univariate Analysis of Covariance: Two-Way Model with One Covariate

183

4.10.4. Additional Topics in Univariate Analysis of Covariance

186

file:///G|/page_viii.html (1 of 2)7/11/2009 2:11:58 PM

page_viii

4.10.5. Multivariate Analysis of Covariance

187

4.11. Alternative Approaches to Testing H0: 1 = 2 = = k 5. Discriminant Functions for Descriptive Group Separation

194

201 201

5.1. Introduction

5.2. Two Groups

201

5.3. Several Groups

202

5.3.1. Discriminant Functions

202

5.3.2. Assumptions

205

5.4. Standardized Coefficients

206

5.5. Tests of Hypotheses

207

5.5.1. Two Groups

207

5.5.2. Several Groups

209

5.6. Discriminant Analysis for Higher Order Designs

210

5.7. Interpretation of Discriminant Functions

210

5.7.1. Standardized Coefficients and Partial F-Values

211

5.7.2. Correlations between Variables and Discriminant Functions Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

211

< previous page

page_viii

next page >

file:///G|/page_viii.html (2 of 2)7/11/2009 2:11:58 PM

page_ix

< previous page

page_ix

next page >

Page ix 215

5.7.3. Other Approaches

5.8. Confidence Intervals

216

5.9. Subset Selection

216

5.9.1. Discriminant Function Approach to Selection

217

5.9.2. Stepwise Selection

217

5.9.3. All Possible Subsets

218

5.9.4. Selection in Higher Order Designs

218

5.10. Bias in Subset Selection

219

5.11. Other Estimators of Discriminant Functions

221

5.11.1. Ridge Discriminant Analysis and Related Techniques

222

5.11.2. Robust Discriminant Analysis 6. Classification of Observations into Groups

223

230 230

6.1. Introduction

6.2. Two Groups

230

6.2.1. Equal Population Covariance Matrices

230

6.2.2. Unequal Population Covariance Matrices

232

6.2.3. Unequal Costs of Misclassification

233

file:///G|/page_ix.html (1 of 3)7/11/2009 2:11:58 PM

page_ix

6.2.4. Posterior Probability Approach

234

6.2.5. Robustness to Departures from the Assumptions

234

6.2.6. Robust Procedures

235

6.3. Several Groups

236

6.3.1. Equal Population Covariance Matrices

236

6.3.2. Unequal Population Covariance Matrices

237

6.3.3. Use of Linear Discriminant Functions for Classification

239

6.4. Estimation of Error Rates

240

6.5. Correcting for Bias in the Apparent Error Rate

244

6.5.1. Partitioning the Sample

244

6.5.2. Holdout Method

244

6.5.3. Bootstrap Estimator

245

6.5.4. Comparison of Error Estimators

245

6.6. Subset Selection

247

6.6.1. Selection Based on Separation of Groups

247

6.6.2. Selection Based on Allocation

249

6.6.3. Selection in the Heteroscedastic Case

250

6.7. Bias in Stepwise Classification Analysis

251

file:///G|/page_ix.html (2 of 3)7/11/2009 2:11:58 PM

page_ix

6.8. Logistic and Probit Classification Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

254

< previous page

page_ix

next page >

file:///G|/page_ix.html (3 of 3)7/11/2009 2:11:58 PM

page_x

< previous page

page_x

next page >

Page x 254

6.8.1. The Logistic Model for Two Groups with l = 2 6.8.2. Comparison of Logistic Classification with Linear Classification Functions 6.8.3. Quadratic Logistic Functions When l 2 6.8.4. Logistic Classification for Several Groups

256

258

6.8.5. Additional Topics in Logistic Classification

259

6.8.6. Probit Classification

259

6.9. Additional Topics in Classification 7. Multivariate Regression

261

266 266

7.1. Introduction

7.2. Multiple Regression: Fixed x's

267

7.2.1. Least Squares Estimators and Properties 7.2.2. An Estimator for 2 7.2.3. The Model in Centered Form

267

268

269

7.2.4. Hypothesis Tests and Confidence Intervals

271

7.2.5. R2 in Fixed-x Regression

275

7.2.6. Model Validation

275

7.3. Multiple Regression: Random x's

276

file:///G|/page_x.html (1 of 2)7/11/2009 2:11:59 PM

page_x

7.3.1. Model for Random x's 7.3.2. Estimation of 0, l, and 2 7.3.3. R2 in Random-x Regression

276

277

278

7.3.4. Tests and Confidence Intervals

297

< previous page

page_x

next page >

7.7. Multivariate Regression: Random x's

303

7.7.1. Multivariate Normal Model for Random x's 7.7.2. Estimation of 0, B1, and 7.7.3. Tests and Confidence Intervals in the Multivariate Random-x Case

303

304

8.6.2. Rotation of Canonical Variate Coefficients

328

8.6.3. Correlations between Variables and Canonical Variates

329

8.7. Redundancy Analysis

331

8.8. Additional Topics Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

333

< previous page

page_xi

next page >

file:///G|/page_xi.html (2 of 2)7/11/2009 2:11:59 PM

page_xii

< previous page

page_xii

next page >

Page xii

9. Principal Component Analysis

337 337

9.1. Introduction

9.2. Definition and Properties of Principal Components

338

9.2.1. Maximum Variance Property

338

9.2.2. Principal Components as Projections

340

9.2.3. Properties of Principal Components

341

9.3. Principal Components as a Rotation of Axes

343

9.4. Principal Components from the Correlation Matrix

344

9.5. Methods for Discarding Components

347

9.5.1. Percent of Variance

347

9.5.2. Average Eigenvalue

348

9.5.3. Scree Graph

348

9.5.4. Significance Tests

349

9.5.5. Other Methods

352

9.6. Information in the Last Few Principal Components

352

9.7. Interpretation of Principal Components

353

9.7.1. Special Patterns in S or R

353

file:///G|/page_xii.html (1 of 3)7/11/2009 2:11:59 PM

page_xii

9.7.2. Testing H0: = 2 [(1 )I + J] and P = (1 )I + J 9.7.3. Additional Rotation

357

359

9.7.4. Correlations between Variables and Principal Components

361

9.8. Relationship Between Principal Components and Regression

363

9.8.1. Principal Component Regression

364

9.8.2. Latent Root Regression

370

9.9. Principal Component Analysis with Grouped Data

371

9.10. Additional Topics 10. Factor Analysis

372

377 377

10.1. Introduction

10.2. Basic Factor Model

377

10.2.1. Model and Assumptions

377

10.2.2. Scale Invariance of the Model

379

10.2.3. Rotation of Factor Loadings in the Model

380

10.3. Estimation of Loadings and Communalities

381

10.3.1. Principal Component Method

381

10.3.2. Principal Factor Method

383

10.3.3. Iterated Principal Factor Method Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

384

file:///G|/page_xii.html (2 of 3)7/11/2009 2:11:59 PM

page_xii

< previous page

page_xii

next page >

file:///G|/page_xii.html (3 of 3)7/11/2009 2:11:59 PM

page_xiii

< previous page

page_xiii

next page >

Page xiii 384

10.3.4. Maximum Likelihood Method

10.3.5. Other Methods

385

10.3.6. Comparison of Methods

385

10.4. Determining the Number of Factors, m

386

10.5. Rotation of Factor Loadings

387

10.5.1. Introduction

387

10.5.2. Orthogonal Rotation

388

10.5.3. Oblique Rotations

389

10.5.4. Interpretation of the Factors

390

10.6. Factor Scores

391

10.7. Applicability of the Factor Analysis Model

392

10.8. Factor Analysis and Grouped Data

393

10.9. Additional Topics Appendix A. Review of Matrix Algebra

394

399 399

A.1. Introduction

A. 1.1. Basic Definitions

399

A. 1.2. Matrices with Special Patterns

400

file:///G|/page_xiii.html (1 of 2)7/11/2009 2:12:00 PM

page_xiii

A.2. Properties of Matrix Addition and Multiplication

401

A.3. Partitioned Matrices

404

A.4. Rank of Matrices

406

A.5. Inverse Matrices

407

A.6. Positive Definite and Positive Semidefinite Matrices

408

A.7. Determinants

409

A.8. Trace of a Matrix

410

A.9. Orthogonal Vectors and Matrices

410

A.10. Eigenvalues and Eigenvectors

411

A. 11. Eigenstructure of Symmetric and Positive Definite Matrices

412

A.12. Idempotent Matrices

414

A. 13. Differentiation Appendix B. Tables Appendix C. Answers and Hints to Selected Problems Appendix D. About the Diskette Bibliography Index Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

414

417 449 505 507 549

< previous page

page_xiii

< previous page

Estimator (continued) maximum likelihood, 4952 unbiased, 23 Expected value of random matrix, 10 of random vector [E(y)], 8 of sample covariance [E(sxy)],6 of sample covariance matrix [E(S)], 11 of sample mean [E( )], 2 of sample mean vector [E( )], 8 of sample variance [E(s2)], 3 of sum or product of random variables, 5 of univariate random variable [E(y)], 2 Experimentwise error rates, 2 F F-test(s) partial F-test for discriminant functions, 211 partial F-test in MANOVA, 110 and T2, 67 tables of power of the F-test, 423426 Factor analysis, 377398 applicability of the model, 392393 assumptions, 378379 communality, 379, 381

page_552

next page >

Page 552

of correlation matrix, 379380, 383, 385386, 389, 392 of covariance matrix, 377388, 391395 eigenvalues, 381383
file:///G|/page_552.html (1 of 4)7/11/2009 2:12:29 PM

page_552

eigenvectors, 381 factor scores, 391392 factors, 377379 interpretation of, 378, 381, 390391 fit of the model, 392393 and grouped data, 393394 loadings, 378, 381385 estimation of, 381385 comparison of methods, 385386 iterated principal factor method, 384 maximum likelihood method, 384385 principal axis method, 383384 principal component method, 381383 principal factor method, 383384 model, 378379 scale invariance of, 379380 nonlinear models, 394 number of factors to retain, 383, 386387 and principal components, 361, 377 and regression, 394395 robustness to nonnormality, 394 rotation of loadings in the model, 380 rotation of sample loadings and factors, 387390 interpretation of factors, 390391 oblique rotation, 389390 pattern matrix, 390 structure matrix, 390 orthogonal rotation, 388389 equamax, 389
file:///G|/page_552.html (2 of 4)7/11/2009 2:12:29 PM

page_552

quartimax, 389 varimax, 388 specific variance, 379 tests of significance, 386387 three-mode factor analysis, 394 total variance, 382 Fisher's LSD test, 134 Fracture data, 171 Function of a random variable, distribution of, 3840 G Gauss-Markov theorem multivariate, 284285 univariate, 268 Generalized linear models, 266 Generalized quadratic form, 5455 Generalized variance, 89, 40 Glucose data, 35 H H matrix, 122, 203 Height-weight data, 4 Hierarchical linear models, 306 Hotelling's T2, see T2-tests Hypothesis tests, see Tests of hypotheses I Idempotent matrix, 414 Identity matrix, 400 Imputation, see Missing data Independence

file:///G|/page_552.html (3 of 4)7/11/2009 2:12:29 PM

page_552

page_553

centering matrix, 910 characteristic equation, 411 Cholesky decomposition, 42, 408 correlation matrix, 1112. See also Correlation matrix covariance matrix, 89. See also Covariance matrix positive definite, 910 data matrix, 7 definition, 399 determinant, 409 diagonal matrix, 11, 400 diagonal of a matrix, 400 differentiation, 5051, 414416 eigenvalues, 411414. See also Eigenvalues characteristic equation, 411 and determinant, 411 of inverse, 411 of positive definite matrix, 412 of product, 411 of symmetric matrix, 412 and trace, 411 eigenvectors, 411414. See also Eigenvectors factoring of, 402 idempotent matrix, 414 identity, 400 inverse, 407 of partitioned matrix, 407 of product, 407 j vector, 400 J matrix, 400
file:///G|/page_553.html (3 of 4)7/11/2009 2:12:29 PM

page_553

page_554

based on M-estimators, 194 based on ranks, 194 and canonical correlation, 312, 333 contrasts, 145148, 150151 orthogonal, 146 simultaneous, 146148 discriminant functions, 128, 151. See also Discriminant analysis; Discriminant functions and eigenvalues, 131 Fisher's LSD test, 134 H and E matrices, 122, 124 higher order models, 151 individual variables, tests on, 134 Lawley-Hotelling test, 130131 F-approximations, 130131 table of critical values, 441445 likelihood ratio test, see Wilks' mixed models, 151 one-way classification, 121134 contrasts, 145148 unbalanced data, 155160 Pillai's test, 130131 F-approximations, 130 table of critical values, 435437 power of the tests, 137138, 140142 tables of power of the F-test, 423426 robustness of test statistics, 137138 Roy's test (union-intersection), 127129 table of critical values, 435437 selection of variables, stepwise, 177

file:///G|/page_554.html (3 of 4)7/11/2009 2:12:29 PM

page_554

step-down test, 177 stepwise discriminant analysis, 177178 test statistics comparison of, 132, 135138 and eigenvalues, 131 power of, 137138, 140142 robustness of, 137138 tests on individual variables, 134 experimentwise error rate, 134 protected tests, 134 tests on a subvector, 174178 two-way classification, 148151 contrasts, 150151 discriminant functions, 151 interaction 148149 main effects, 148149 Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

< previous page

page_554

next page >

file:///G|/page_554.html (4 of 4)7/11/2009 2:12:29 PM

page_555

< previous page

model, 148 test statistics, 149 unbalanced data, 152, 155160, 168174 contrasts, 156160 Bonferroni confidence intervals, 157

page_555

next page >

Page 555

simultaneous confidence intervals, 156157 discriminant functions, 157 one-way model, 155160 tests on individual variables, 157 two-way model, 168174 constrained model, 169171 union-intersection test, see Roy's test Wilks' (likelihood ratio), 122126 Chi-square approximation, 126 effect of each variable, 132134 F-approximation, 125126 properties of, 124126 table of critical values, 427434 transformations to exact F, 125126 Multivariate central limit theorem, 53 Multivariate data, 12, 7 Multivariate descriptive procedures, 1 Multivariate inferential procedures, 12 Multivariate normal distribution, 3759 central limit theorem, 53 conditional distribution, 4748 constant density ellipsoid, 4041

file:///G|/page_555.html (1 of 4)7/11/2009 2:12:30 PM

page_555

contour plot, 4042 estimation of and , 4951 properties of estimators, 5256 estimation of Pp, 5152 generating data from, 42 independence and zero covariance, 4647 likelihood function, 49 linear functions of, 4546 marginal distribution, 46 moment generating function of, 4447 moments of, 4243 properties of, 4349 and quadratic form, 4849 tests for multivariate normality, 43 transformations to achieve multivariate normality. 5657 and Wishart distribution, 5356 Multivariate normality, tests for, 43 Multivariate regression, see Regression, multivariate Multivariate t-distribution, 56 N Nonparametric tests for several mean vectors, 194 for two mean vectors, 113 Normal distribution multivariate normal, see Multivariate normal distribution univariate normal, 3738 standard normal, 37 Normality, tests for, 43 O
file:///G|/page_555.html (2 of 4)7/11/2009 2:12:30 PM

page_555

O matrix, 401 Orthogonal matrix, 410 Orthogonal vectors, 410 Outliers, 43, 301303, 373 P Paired observation test, 9799 Partial F, 177, 211, 214215, 217, 219, 248249, 252 Perron-Frobenius theorem, 414 Physiological data, 224 Pillai's test, 130131, 292, 295296 F-approximations, 130 table of critical values, 435437 Positive definite matrix, 408 Power of a test, 104108 noncentrality parameter, 104 Principal component regression, 363370 Principal components, 337376 common principal components, 371372 from correlation matrix (R), 342, 344349, 351, 353359, 364, 367370 properties of, 344345 correlations of, 341342 and correspondence analysis, 373 from covariance matrix (S), 338344, 348355, 357363, definition, 338 discarding components, 347352 average eigenvalue, 348 bootstrap, 352 cross validation, 352 jackknife, 352
file:///G|/page_555.html (3 of 4)7/11/2009 2:12:30 PM

page_555

percent of variance, 347 rank of , 352 scree graph, 348349 tests of significance, 349351 and eigenvalues, 338, 341 and eigenvectors, 338 empirical orthogonal functions, 337 and factor analysis, 361, 377 generalized principal components, 340 with grouped data, 371372 common principal components, 371372 comparing principal components, 371372 influence of each observation, 372373 interpretation, 353363 correlations, 361363 Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

< previous page

page_555

next page >

file:///G|/page_555.html (4 of 4)7/11/2009 2:12:30 PM

page_556

< previous page

Principal components (continued) rotation, 359361 special patterns in S or R, 353359 equal correlations, 354, 358359

page_556

next page >

Page 556

estimation of 2, 268269 and factor analysis, 394395 fixed x's, 267276 Gauss-Markov theorem, 268 model, 267 assumptions, 267 corrected for means, 269271 validation, 275276 multiple correlation, 275 and principal components, 363370 R2 (squared multiple correlation), 275, 278279 random x's, 276279 confidence intervals, 279 estimation of the j's and 2, 277278 correlations, 278 beta weights, 278 model, 276 R2 (squared multiple correlation), 278279 tests of hypotheses, 279 ridge regression, 363 SSE, 269 SSR, 271 tests of hypotheses, 271274 Bonferroni tests for individual j's, 273274 general linear hypothesis H0:C= 0, 272273 overall regression test, 271 simultaneous tests for individual j's, 274

page_557

next page >

Page 557

file:///G|/page_557.html (1 of 4)7/11/2009 2:12:30 PM

page_557

subset of the 's, 271272 test for an individual j, 273 Regression, multivariate (several y's and several x's) association, measures of, 289 confidence interval(s), 297299, 305 for a'Bb, 298 for an individual jk, 297 simultaneous intervals for jkand a'Bb, 298 correlated response model, 305306 estimation of B, 281284, 286289 with centered x's, 288289 with covariances, 288289 least squares, 281282 maximum likelihood, 286289 properties of estimators, 283284, 287288 estimation of , 285288 properties of estimators, 285, 287288 Gauss-Markov theorem, 284285 hierarchical linear models, 306 measurement errors, 303 model, 279281 assumptions, 281 in centered form, 288289 normal model, 286288 validation, 300303 influence, 301303 lack-of-fit tests, 300 leverage, 301 outliers, 301303
file:///G|/page_557.html (2 of 4)7/11/2009 2:12:30 PM

page_557

residuals, 300301 random x's, 303305 confidence intervals, 305 estimation of the j's and , 304305 beta weights, 305 correlations, 305 model, 303304 tests of hypotheses, 305 robust estimators, 307 seemingly unrelated regressions, 306 subset selection, 307 tests of hypotheses, 289300, 305 E matrix, 285, 286, 290 general linear hypotheses H0:CB = O and H0: CBM = O, 295297 H matrix, 290, 293, 296 for an individual jk, 297 Lawley-Hotelling test, 292, 295, 296 likelihood ratio test, see Wilks' and missing data, 299300 overall regression test, 289293 in terms of canonical correlations, 323 power of tests, 292293 Pillai's test, 292, 295, 296 Roy's test, 291292, 295, 296 subset of the x's, 293295 union-intersection test, see Roy's test Wilks' , 290291, 294, 296 Repeated measures, 121, 183, 246, 296 Robust estimation
file:///G|/page_557.html (3 of 4)7/11/2009 2:12:30 PM

page_557

affine equivariance of (scale invariance), 30 breakdown point, 30 of canonical variates, 333 in classification analysis, 235 contaminated normal, 28 of correlation matrices, 2930 of covariance matrices, 2931 of discriminant functions, 223227 mean deviation, 28 of mean vectors, 2831 of means, 28 of principal components, 373 in regression, 307 Rotation, see Factor analysis Roy's test (union-intersection), 127129, 291296 table of critical values, 435437 S Seemingly unrelated regressions, 306 Selection of variables, 111, 177, 217221 Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

< previous page

page_557

next page >

file:///G|/page_557.html (4 of 4)7/11/2009 2:12:30 PM

page_558

< previous page

Selection of variables (continued) all possible subsets, 218 bias, 219221 in classification analysis, 247251 using discriminant functions, 217 in higher order designs, 218219 in regression, 307 stepwise discriminant analysis, 217221 Singular value decomposition, 2627, 315316 Size and shape, 355 Soils data, 103 Specific variance, 379 Spectral decomposition, 412 Squared multiple correlation, see R2 SSE, 269 SSR, 271 Standardized vector, 45 Stepdown test, 110111

page_558

next page >

Page 558

page_560

< previous page

page_560
WILEY SERIES IN PROBABILITY AND STATISTICS ESTABLISHED BY WALTER A. SHEWHART AND SAMUEL S. WILKS Editors

next page >

Page 560

Vic Barnett, Ralph A. Bradley, Noel A. C. Cressie, Nicholas I. Fisher, lain M. Johnstone, J. B. Kadane, David G. Kendall, David W. Scott, Bernard W. Silverman, Adrian F. M. Smith, JozefL. Teugels, Geoffrey S. Watson; J. Stuart Hunter, Emeritus Probability and Statistics ANDERSON An Introduction to Multivariate Statistical Analysis, Second Edition *ANDERSON The Statistical Analysis of Time Series ARNOLD, BALAKRISHNAN, and NAGARAJA A First Course in Order Statistics BACCELLI, COHEN, OLSDER, and QUADRAT Synchronization and Linearity: An Algebra for Discrete Event Systems BARTOSZYNSKI and NIEWIADOMSKA-BUGAJ Probability and Statistical Inference BERNARDO and SMITH Bayesian Statistical Concepts and Theory BHATTACHARYYA and JOHNSON Statistical Concepts and Methods BILLINGSLEY Convergence of Probability Measures BILLINGSLEY Probability and Measure, Second Edition BOROVKOV Asymptotic Methods in Queuing Theory BRANDT, FRANKEN, and LISEK Stationary Stochastic Models CAINES Linear Stochastic Systems CAIROLI and DALANG Sequential Stochastic Optimization CHEN Recursive Estimation and Control for Stochastic Systems CONSTANTINE Combinatorial Theory and Statistical Design COOK and WEISBERG An Introduction to Regression Graphics COVER and THOMAS Elements of Information Theory CSORGO and HORVATH Weighted Approximations in Probability Statistics *DOOB Stochastic Processes DUDEWICZ and MISHRA Modem Mathematical Statistics

file:///G|/page_560.html (1 of 2)7/11/2009 2:12:31 PM

page_562
Applied Probability and Statistics

next page >

Page 562

ABRAHAM and LEDOLTER Statistical Methods for Forecasting AGRESTI Analysis of Ordinal Categorical Data AGRESTI Categorical Data Analysis AGRESTI An Introduction to Categorical Data Analysis ANDERSON and LOYNES The Teaching of Practical Statistics ANDERSON, AUQUIER, HAUCK, OAKES, VANDAELE, and WEISBERG Statistical Methods for Comparative Studies ARMITAGE and DAVID (editors) Advances in Biometry *ARTHANARI and DODGE Mathematical Programming in Statistics ASMUSSEN Applied Probability and Queues *BAILEY The Elements of Stochastic Processes with Applications to the Natural Sciences BARNETT and LEWIS Outliers in Statistical Data, Second Edition BARTHOLOMEW, FORBES, and McLEAN Statistical Techniques for Manpower Planning, Second Edition BATES and WATTS Nonlinear Regression Analysis and Its Applications BECHHOFER, SANTNER, and GOLDSMAN Design and Analysis of Experiments for Statistical Selection, Screening, and Multiple Comparisons BELSLEY Conditioning Diagnostics: Collinearity and Weak Data in Regression BELSLEY, KUH, and WELSCH Regression Diagnostics: Identifying Influential Data and Sources of Collinearity BERRY Bayesian Analysis in Statistics and Econometrics: Essays in Honor of Arnold Zellner BERRY, CHALONER, and GEWEKE Bayesian Analysis in Statistics and Econometrics: Essays in Honor of Arnold Zellner BHAT Elements of Applied Stochastic Processes, Second Edition BHATTACHARYA and WAYMIRE Stochastic Processes with Applications BIEMER, GROVES, LYBERG, MATHIOWETZ, and SUDMAN Measurement Errors in Surveys BIRKES and DODGE Alternative Methods of Regression BLOOMFIELD Fourier Analysis of Time Series: An Introduction BOLLEN Structural Equations with Latent Variables BOULEAU Numerical Methods for Stochastic Processes
file:///G|/page_562.html (1 of 2)7/11/2009 2:12:32 PM

page_562

BOX R. A. Fisher, the Life of a Scientist BOX and DRAPER Empirical Model-Building and Response Surfaces BOX and DRAPER Evolutionary Operation: A Statistical Method for Process Improvement BOX, HUNTER, and HUNTER Statistics for Experimenters: An Introduction to Design, Data Analysis, and Model Building BOX and LUCENO Statistical Control by Monitoring and Feedback Adjustment BROWN and HOLLANDER Statistics: A Biomedical Introduction BUCKLEW Large Deviation Techniques in Decision, Simulation, and Estimation BUNKE and BUNKE Nonlinear Regression, Functional Relations and Robust Methods: Statistical Methods of Model Building CHATTERJEE and HADI Sensitivity Analysis in Linear Regression CHATTERJEE and PRICE Regression Analysis by Example, Second Edition CLARKE and DISNEY Probability and Random Processes: A First Course with Applications, Second Edition COCHRAN Sampling Techniques, Third Edition *COCHRAN and COX Experimental Designs, Second Edition CONOVER Practical Nonparametric Statistics, Second Edition CONOVER and IMAN Introduction to Modern Business Statistics *Now available in a lower priced paperback edition in the Wiley Classics Library. Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

< previous page

page_562

next page >

file:///G|/page_562.html (2 of 2)7/11/2009 2:12:32 PM

page_563

< previous page

page_563
Applied Probability and Statistics (Continued)

next page >

Page 563

CORNELL Experiments with Mixtures, Designs, Models, and the Analysis of Mixture Data, Second Edition COX A Handbook of Introductory Statistical Methods *COX Planning of Experiments COX, BINDER, CHINNAPPA, CHRISTIANSON, COLLEDGE, and KOTT Business Survey Methods CRESSIE Statistics for Spatial Data, Revised Edition DANIEL Applications of Statistics to Industrial Experimentation DANIEL Biostatistics: A Foundation for Analysis in the Health Sciences, Sixth Edition DAVID Order Statistics, Second Edition *DEGROOT, FIENBERG, and KADANE Statistics and the Law *DEMING Sample Design in Business Research DETTE and STUDDEN The Theory of Canonical Moments with Applications in Statistics, Probability, and Analysis DILLON and GOLDSTEIN Multivariate Analysis: Methods and Applications DODGE and ROMIG Sampling Inspection Tables, Second Edition DOWDY and WEARDEN Statistics for Research, Second Edition DRAPER and SMITH Applied Regression Analysis, Second Edition DUNN Basic Statistics: A Primer for the Biomedical Sciences, Second Edition DUNN and CLARK Applied Statistics: Analysis of Variance and Regression, Second Edition ELANDT-JOHNSON and JOHNSON Survival Models and Data Analysis EVANS, PEACOCK, and HASTINGS Statistical Distributions, Second Edition FISHER and VAN BELLE Biostatistics: A Methodology for the Health Sciences FLEISS The Design and Analysis of Clinical Experiments FLEISS Statistical Methods for Rates and Proportions, Second Edition FLEMING and HARRINGTON Counting Processes and Survival Analysis FLURY Common Principal Components and Related Multivariate Models GALLANT Nonlinear Statistical Models GLASSERMAN and YAO Monotone Structure in Discrete-Event Systems

file:///G|/page_563.html (1 of 2)7/11/2009 2:12:32 PM

page_563

GNANADESIKAN Methods for Statistical Data Analysis of Multivariate Observations, Second Edition GREENWOOD and NIKULIN A Guide to Chi-Squared Testing GROSS and HARRIS Fundamentals of Queueing Theory, Second Edition GROVES Survey Errors and Survey Costs GROVES, BIEMER, LYBERG, MASSEY, NICHOLLS, and WAKSBERG Telephone Survey Methodology HAHN and MEEKER Statistical Intervals: A Guide for Practitioners HAND Discrimination and Classification *HANSEN, HURWITZ, and MADOW Sample Survey Methods and Theory, Volume 1: Methods and Applications *HANSEN, HURWITZ, and MADOW Sample Survey Methods and Theory, Volume II: Theory HEIBERGER Computation for the Analysis of Designed Experiments HELLER MACSYMA for Statisticians HINKELMAN and KEMPTHORNE: Design and Analysis of Experiments, Volume I: Introduction to Experimental Design HOAGLIN, MOSTELLER, and TUKEY Exploratory Approach to Analysis of Variance HOAGLIN, MOSTELLER, and TUKEY Exploring Data Tables, Trends and Shapes HOAGLIN, MOSTELLER, and TUKEY Understanding Robust and Exploratory Data Analysis HOCHBERG and TAMHANE Multiple Comparison Procedures HOCKING Methods and Applications of Linear Models: Regression and the Analysis of Variables *Now available in a lower priced paperback edition in the Wiley Classics Library. Start of Citation[PU]John Wiley & Sons, Inc. (US)[/PU][DP]1998[/DP]End of Citation

next page >

file:///G|/page_564.html (2 of 2)7/11/2009 2:12:32 PM

page_565

< previous page

page_565
Applied Probability and Statistics (Continued)

next page >

Page 565

page_572
Page 572

CUSTOMER NOTE: IF THIS BOOK IS ACCOMPANIED BY SOFTWARE, PLEASE READ THE FOLLOWING BEFORE OPENING THE PACKAGE. This software contains files to help you utilize the models described in the accompanying book. By opening the package, you are agreeing to be bound by the following agreement: This software product is protected by copyright and all rights are reserved by the author and John Wiley & Sons, Inc. You are licensed to use this software on a single computer. Copying the software to another medium or format for use on a single computer does not violate the U. S. Copyright Law. Copying the software for any other purpose is a violation of the U. S. Copyright Law. This software product is sold as is without warranty of any kind, either express or implied, including but not limited to the implied warranty of merchantability and fitness for a particular purpose. Neither Wiley nor its dealers or distributors assumes any liability of any alleged or actual damages arising from the use of or the inability V to use this software. (Some states do not allow the exclusion of implied warranties, so the exclusion may not apply to you.)