Commissioned by the Statistical Society of Canada (SSC), Statistics in Action: A Canadian Outlook helps both general readers and users of statistics better appreciate the scope and importance of statistics. It presents the ways in which statistics is used while highlighting key contributions that Canadian statisticians are making to science, technology, business, government, and other areas. The book emphasizes the role and impact of computing in statistical modeling and analysis, including the issues involved with the huge amounts of data being generated by automated processes.
The first two chapters review the development of statistics as a discipline in Canada and describe some major contributions to survey methodology made by Statistics Canada, one of the world's premier official statistics agencies. The book next discusses how statistical methodologies, such as functional data analysis and the Metropolis algorithm, are applied in a wide variety of fields, including risk management and genetics. It then focuses on the application of statistical methods in medicine and public health as well as finance and e-commerce. The remainder of the book addresses how statistics is used to study critical scientific areas, including difficult-to-access populations, endangered species, climate change, and agricultural forecasts.

Foreword xiii

Preface xv

About the Editor xix

Contributors xxi

1 Canadians Studying Abroad and the Development of Statistics in Canada David R. Bellhouse and Stephen E. Fienberg 1

1.1 Introduction 1

1.2 Statistics' Rise as a Modern Field in North America 2

1.3 Statistical Education in Canada between the Wars 4

1.4 World War II as a Watershed 8

1.5 A Second Early Generation of Canadian Statisticians 9

1.6 Generation of the 1950s and Early 1960s 10

      1.6.1 Princeton Connection 11

      1.6.2 More Migration from Toronto 12

      1.6.3 Manitoba Influence 13

      1.6.4 Other Notable Canadians Who Went Abroad 14

1.7 A Paradigm Shift 15

Bibliography 17

2 Some of Statistics Canada's Contributions to Survey Methodology Jean-Francois Beaumont, Susie Fortier, Jack Gambino, Mike Hidiroglou, and Pierre LavaIlee 19

2.1 Introduction 19

2.2 Sample Design 21

      2.2.1 Record Linkage 21

      2.2.2 Multiple Frames 22

      2.2.3 Indirect Sampling 23

      2.2.4 Optimal Stratification 23

      2.2.5 Power Allocation 25

2.3 Edit and Imputation 25

2.4 Estimation 27

      2.4.1 Generalized Estimation System 29

      2.4.2 Outliers 30

      2.4.3 Composite Estimation 30

      2.4.4 Small Area Estimation 31

2.5 Variance Estimation 32

      2.5.1 Taylor Linearization 33

      2.5.2 Replication Methods 34

      2.5.3 Variance Estimation in the Presence of Imputation 35

2.6 Data Analysis 36

2.7 Time Series 37

      2.7.1 Seasonal Adjustment and Trend-Cycle Estimation Method 37

      2.7.2 Benchmarking and Reconciliation Methods 38

2.8 Conclusion 39

Bibliography 40

3 Watching Children Grow Taught Us All We Know James O. Ramsay and Michael Hermanussen 47

3.1 Introduction 47

3.2 Modeling Growth 49

3.3 Phase/Plane Plots of Derivative Interactions 50

3.4 Identification and Analysis of Phase or Tempo Variation 53

3.5 Conclusion 56

Bibliography 56

4 Modeling Dependence beyond Correlation Christian Genest and Johanna G. Našlehová 59

4.1 Beyond the Normal Brave Old World 59

4.2 Dangers of Ignoring Non-Normality 62

4.3 Copulas to the Rescue 64

      4.3.1 Fréchet's Problem 64

      4.3.2 Measuring Association 65

      4.3.3 Stress Testing 66

      4.3.4 Validate, Validate, Validate! 69

4.4 Proof of the Pudding Is in the Ranking 70

4.5 Applications and Future Challenges 74

Bibliography 75

5 Lasso and Sparsity in Statistics Robert J. Tibshirani 79

5.1 Sparsity, l1 Penalties and the Lasso 79

5.2 Some Background 82

5.3 A History for Coordinate Descent for the Lasso 84

5.4 An Example in Medicine 85

5.5 Nearly Isotonic Regression 87

5.6 Conclusion 89

Bibliography 89

6 Optimizing and Adapting the Metropolis Algorithm Jeffrey S. Rosenthal 93

6.1 Introduction 93

6.2 Monte Carlo Algorithms 94

6.3 Markov Chain Monte Carlo (MCMC) 95

6.4 Metropolis Algorithm 96

6.5 Goldilocks Principle 97

6.6 Optimal Scaling 100

6.7 Proposal Shape 101

6.8 Adaptive MCMC 103

6.9 Summary 105

Bibliography 106

7 Design of Computer Experiments for Optimization, Estimation of Function Contours, and Related Objectives Derek Bingham, Pritam. Ranjan, and William. J. Welch 109

7.1 Introduction 109

7.2 Expected Improvement and Sequential Design: Basic Ideas 110

      7.2.1 Optimization 110

      7.2.2 Contour Estimation 114

7.3 Expected Improvement Criteria 116

      7.3.1 EI for Global Optimization 116

      7.3.2 EI for Contour Estimation 117

7.4 Gaussian Process Models and Predictive Distributions 118

7.5 Other EI-Based Criteria 120

7.6 Summary 122

Bibliography 123

8 Statistical Genetic Modeling and Analysis of Complex Traits Shelley B. Bull, Jinko Graham, and Celia M. T. Greenwood 125

8.1 Introduction and Overview 125

8.2 Essentials of Statistical Genetic Methods 127

      8.2.1 Modeling Genetic Sharing in Sibships and Relative Pairs 129

      8.2.2 Models for Genetic Transmission in Families 131

      8.2.3 Genetic Association in Unrelated Individuals 133

8.3 Advances in Statistical Genetic Methods 135

      8.3.1 Linkage with Covariate Data 135

      8.3.2 Effect Estimation in Genome-Wide Analysis 138

8.4 Commentary: Current and Emerging Issues 140

Bibliography 142

9 Bayesian Methods in Fisher's Statistical Genetics World Radu V. Craiu and Lei Sun 147

9.1 Background and Introduction 147

      9.1.1 Basic Genetic Terminology 148

      9.1.2 Statistical Set-Up of Genetic Association Studies: Two Intertwined Issues 148

      9.1.3 GWAS and an Example 150

9.2 Identification of Trait-Associated SNPs 151

      9.2.1 p-Value 151

      9.2.2 Bayes Factor 152

      9.2.3 Additional Considerations 153

9.3 Replication of a Significant Finding 155

      9.3.1 Conditional MLE vs. Bayesian Model Averaging 155

      9.3.2 Computational Considerations 157

9.4 Conclusion and Discussion 158

Bibliography 159

10 Bayesian Statistical Methodology for Observational Health Sciences Data Paul Gustafson 163

10.1 Introduction 163

10.2 Case-Control Studies 165

10.3 Bayesian Analysis 167

10.4 Exposure Misclassification 168

10.5 Going Further with Bayesian Analysis 172

Bibliography 174

11 Statistical Models for Disease Processes: Markers and Skeletal Complications in Cancer Metastatic to Bone Richard J. Cook 177

11.1 Skeletal Complications in Individuals Suffering from Cancer Metastatic to Bone 177

11.2 Bone Markers and Prognosis for Adverse Events 180

11.3 Joint Models for Dynamic Bone Markers and Adverse Events 183

11.4 Discussion 189

Bibliography 191

12 Analysis of Biased Survival Data: The Canadian Study of Health and Aging and beyond Masoud Asgharian, Christina Wolfson, and David B. Wolfson 193

12.1 Introduction 193

12.2 Nonparametric Estimation of the Survivor Function 196

12.3 Checking for Stationarity of the Incidence Process 200

12.4 Estimating the Incidence Rate 201

12.5 Covariates 202

12.6 Concluding Remarks 203

Bibliography 205

13 Assessing the Effect on Survival of Kidney Transplantation with Higher-Risk Donor Kidneys Douglas E. Schaubel and John D. Kalbfleisch 209

13.1 Introduction 209

13.2 Study Population and Notation 211

13.3 Analysis Based on Time-Dependent Treatment Indicator 213

13.4 Modification to Time-Dependent Analysis 216

13.5 Sequential Stratification 217

13.6 Discussion 219

Bibliography 222

14 Risk-Adjusted Monitoring of Outcomes in Health Care Stefan H. Steiner 225

14.1 Introduction 225

14.2 Motivation and Background 226

14.3 Monitoring Industrial Processes 227

14.4 Monitoring Outcomes in Health Care: Issues and Challenges 229

14.5 Monitoring Outcomes in Health Care: Methods 231

      14.5.1 Variable Life-Adjusted Display Chart 232

      14.5.2 Exponentially Weighted Moving Average Chart 232

      14.5.3 Risk-Adjusted Cumulative Sum Chart 234

14.6 Monitoring Outcomes in Health Care: Uses and Future 237

Bibliography 239

15 Statistics in Financial Engineering Bruno Rémillard 243

15.1 Introduction 243

15.2 Modeling 244

      15.2.1 Black—Scholes—Merton Model 244

      15.2.2 Extensions of the BSM Model 246

      15.2.3 Choice of Distribution for the Innovations 247

      15.2.4 Challenge of Model Validation 249

15.3 Applications to Financial Engineering 250

      15.3.1 Portfolio Management 250

      15.3.2 Option Pricing 252

      15.3.3 Risk Management 253

15.4 Final Comment 255

Bibliography 256

16 Making Personalized Recommendations in E-Connnerce Mu Zhu 259

16.1 Introduction 259

16.2 Nearest Neighbors 261

16.3 Matrix Factorization 261

16.4 Matrix Completion 262

16.5 Content-Boosted Matrix Factorization 264

16.6 Discussion 267

Bibliography 267

17 What Do Salmon and Injection Drug Users Have in Common? Laura L. E. Cowen, Wendell O. Challenger, and Carl J. Schwarz 269

17.1 Methodology Background 269

17.2 Closed Population Models and Two-Sample Experiments 270

      17.2.1 Victoria's Injection Drug Users 273

      17.2.2 Rotary Screw Traps and the Time Stratified Petersen Estimator 274

      17.2.3 Extending the Closed Population Models 276

17.3 Open Populations 277

      17.3.1 Novel Applications — Pacific Ocean Shelf Tracking 278

17.4 Additional Complications 282

17.5 Concluding Remarks 283

Bibliography 284

18 Capture–Recapture Methods for Estimating the Size of a Population: Dealing with Variable Capture Probabilities Louis-Paul Rivest and Sophie Baillargeon 289

18.1 Estimating Abundance with Marked Units 289

18.2 Datasets 292

18.3 Estimation of Population Size Using Aggregated Data 293

      18.3.1 Homogeneity Model 293

      18.3.2 A Probability Plot and a Lower Bound Estimate 295

      18.3.3 Parametric Models with Random Capture Probabilities 298

18.4 Modeling Capture Probabilities with Unit Level Covariates 300

18.5 Discussion 302

Bibliography 303

19 Challenges in Statistical Marine Ecology Joanna Mills Flemming and Christopher A. Field 305

19.1 Introduction 305

19.2 Sustaining North Atlantic Cod Stocks 307

      19.2.1 Results 309

19.3 Conserving the Endangered Hammerhead Shark 312

      19.3.1 Results 315

19.4 Conclusions 317

Bibliography 318

20 Quantifying the Human and Natural Contributions to Ob-served Climate Change Francis W. Zwiers, Gabriele C. Hegerl, Xuebin Zhang, and Qiuzi Wen 321

20.1 Introduction 321

20.2 Scope of the Problem 322

20.3 A Weighted Regression Approach 324

20.4 Role of the Global Climate Models 327

20.5 A Slightly More Detailed Regression Model 329

20.6 Interpreting the Results 331

20.7 Discussion 335

Bibliography 338

21 Data Hungry Models in a Food Hungry World: An Interdis-ciplinary Challenge Bridged by Statistics Louis Kouadio and Nathaniel Newlands 341

21.1 Big Challenges for Future Global Crop Production 341

21.2 Differing Needs for Crop Yield "Outlook" Forecasts 342

21.3 Use of Satellite Remote Sensing Data 344

21.4 Regional-Scale Crop Outlooks 348

21.5 More Reliable Forecasting 353

Bibliography 354

Index 357

Jerald F. Lawless is distinguished professor emeritus in the Department of Statistics and Actuarial Science at the University of Waterloo, where he was a faculty member from 1972 to 2007. His research interests include biostatistics, survival and event history analysis, reliability, and general methodology. He is the author of three books and numerous research articles. He has also con-sulted and collaborated widely in fields that include manufacturing, medicine, risk and safety analysis, social sciences, and law. He is a past editor of Tech-nometrics and a former president of the Statistical Society of Canada (SSC). He has received the SSC Gold Medal for research and the Shewhart Medal of the American Society for Quality. He is an elected fellow of the Royal Society of Canada, a fellow of the American Statistical Association, and a fellow of the Institute of Mathematical Statistics.

中科院文献情报中心