Modern Engineering Statistics

THOMAS P. RYAN, PHD, served on the Editorial Review Board of the Journal of Quality Technology from 1990 to 2006, including three years as the book review editor. He is the author of four books published by Wiley and is an elected Fellow of the American Statistical Association, the American Society for Quality, and the Royal Statistical Society. He currently teaches advanced courses on design of experiments and engineering statistics at statistics.com and serves as a consultant to Cytel Software Corporation.

• Preface xvii1. Methods of Collecting and Presenting Data 11.1 Observational Data and Data from Designed Experiments 31.2 Populations and Samples 51.3 Variables 61.4 Methods of Displaying Small Data Sets 71.5 Methods of Displaying Large Data Sets 161.6 Outliers 221.7 Other Methods 221.8 Extremely Large Data Sets: Data Mining 231.9 Graphical Methods: Recommendations 231.10 Summary 24References 24Exercises 252. Measures of Location and Dispersion 452.1 Estimating Location Parameters 462.2 Estimating Dispersion Parameters 502.3 Estimating Parameters from Grouped Data 552.4 Estimates from a Boxplot 572.5 Computing Sample Statistics with MINITAB 582.6 Summary 58Reference 58Exercises 583. Probability and Common Probability Distributions 683.1 Probability: From the Ethereal to the Concrete 683.3 Common Discrete Distributions 763.4 Common Continuous Distributions 923.5 General Distribution Fitting 1063.6 How to Select a Distribution 1073.7 Summary 108References 109Exercises 1094. Point Estimation 1214.1 Point Estimators and Point Estimates 1214.2 Desirable Properties of Point Estimators 1214.3 Distributions of Sampling Statistics 1254.4 Methods of Obtaining Estimators 1284.5 Estimating σθ 1324.6 Estimating Parameters Without Data 1334.7 Summary 133References 134Exercises 1345. Confidence Intervals and Hypothesis Tests—One Sample 1405.1 Confidence Interval for μ: Normal Distribution σ Not Estimated from Sample Data 1405.2 Confidence Interval for μ: Normal Distribution σ Estimated from Sample Data 1465.3 Hypothesis Tests for μ: Using Z and t 1475.4 Confidence Intervals and Hypothesis Tests for a Proportion 1575.5 Confidence Intervals and Hypothesis Tests for σ2 and σ 1615.6 Confidence Intervals and Hypothesis Tests for the Poisson Mean 1645.7 Confidence Intervals and Hypothesis Tests When Standard Error Expressions are Not Available 1665.8 Type I and Type II Errors 1685.9 Practical Significance and Narrow Intervals: The Role of n 1725.10 Other Types of Confidence Intervals 1735.11 Abstract of Main Procedures 1745.12 Summary 175Appendix: Derivation 176References 176Exercises 1776. Confidence Intervals and Hypothesis Tests—Two Samples 1896.1 Confidence Intervals and Hypothesis Tests for Means: Independent Samples 1896.2 Confidence Intervals and Hypothesis Tests for Means: Dependent Samples 1976.3 Confidence Intervals and Hypothesis Tests for Two Proportions 2006.4 Confidence Intervals and Hypothesis Tests for Two Variances 2026.5 Abstract of Procedures 2046.6 Summary 205References 205Exercises 2057. Tolerance Intervals and Prediction Intervals 2147.1 Tolerance Intervals: Normality Assumed 2157.2 Tolerance Intervals and Six Sigma 2197.3 Distribution-Free Tolerance Intervals 2197.4 Prediction Intervals 2217.5 Choice Between Intervals 2277.6 Summary 227References 228Exercises 2298. Simple Linear Regression Correlation and Calibration 2328.1 Introduction 2328.2 Simple Linear Regression 2328.3 Correlation 2548.4 Miscellaneous Uses of Regression 2568.5 Summary 264References 264Exercises 2659. Multiple Regression 2769.1 How Do We Start? 2779.2 Interpreting Regression Coefficients 2789.3 Example with Fixed Regressors 2799.4 Example with Random Regressors 2819.5 Example of Section 8.2.4 Extended 2919.6 Selecting Regression Variables 2939.7 Transformations 2999.8 Indicator Variables 3009.9 Regression Graphics 3009.10 Logistic Regression and Nonlinear Regression Models 3019.11 Regression with Matrix Algebra 3029.12 Summary 302References 303Exercises 30410. Mechanistic Models 31410.1 Mechanistic Models 31510.2 Empirical–Mechanistic Models 31610.3 Additional Examples 32410.4 Software 32510.5 Summary 326References 326Exercises 32711. Control Charts and Quality Improvement 33011.1 Basic Control Chart Principles 33011.2 Stages of Control Chart Usage 33111.3 Assumptions and Methods of Determining Control Limits 33411.4 Control Chart Properties 33511.5 Types of Charts 33611.6 Shewhart Charts for Controlling a Process Mean and Variability (Without Subgrouping) 33611.7 Shewhart Charts for Controlling a Process Mean and Variability (With Subgrouping) 34411.8 Important Use of Control Charts for Measurement Data 34911.9 Shewhart Control Charts for Nonconformities and Nonconforming Units 34911.10 Alternatives to Shewhart Charts 35611.11 Finding Assignable Causes 35911.12 Multivariate Charts 36211.13 Case Study 36211.14 Engineering Process Control 36411.15 Process Capability 36511.16 Improving Quality with Designed Experiments 36611.17 Six Sigma 36711.18 Acceptance Sampling 36811.19 Measurement Error 36811.20 Summary 368References 369Exercises 37012. Design and Analysis of Experiments 38212.1 Processes Must be in Statistical Control 38312.2 One-Factor Experiments 38412.3 One Treatment Factor and at Least One Blocking Factor 39212.4 More Than One Factor 39512.5 Factorial Designs 39612.6 Crossed and Nested Designs 40512.7 Fixed and Random Factors 40612.8 ANOM for Factorial Designs 40712.9 Fractional Factorials 40912.10 Split-Plot Designs 41312.11 Response Surface Designs 41412.12 Raw Form Analysis Versus Coded Form Analysis 41512.13 Supersaturated Designs 41612.14 Hard-to-Change Factors 41612.15 One-Factor-at-a-Time Designs 41712.16 Multiple Responses 41812.17 Taguchi Methods of Design 41912.18 Multi-Vari Chart 42012.19 Design of Experiments for Binary Data 42012.20 Evolutionary Operation (EVOP) 42112.21 Measurement Error 42212.22 Analysis of Covariance 42212.23 Summary of MINITAB and Design-Expert® Capabilities for Design of Experiments 42212.24 Training for Experimental Design Use 42312.25 Summary 423Appendix A Computing Formulas 424Appendix B Relationship Between Effect Estimates andRegression Coefficients 426References 426Exercises 42813. Measurement System Appraisal 44113.1 Terminology 44213.2 Components of Measurement Variability 44313.3 Graphical Methods 44913.4 Bias and Calibration 44913.5 Propagation of Error 45413.6 Software 45513.7 Summary 456References 456Exercises 45714. Reliability Analysis and Life Testing 46014.1 Basic Reliability Concepts 46114.2 Nonrepairable and Repairable Populations 46314.3 Accelerated Testing 46314.4 Types of Reliability Data 46614.5 Statistical Terms and Reliability Models 46714.6 Reliability Engineering 47314.7 Example 47414.8 Improving Reliability with Designed Experiments 47414.9 Confidence Intervals 47714.10 Sample Size Determination 47814.11 Reliability Growth and Demonstration Testing 47914.12 Early Determination of Product Reliability 48014.13 Software 48014.14 Summary 481References 481Exercises 48215. Analysis of Categorical Data 48715.1 Contingency Tables 48715.2 Design of Experiments: Categorical Response Variable 49715.3 Goodness-of-Fit Tests 49815.4 Summary 500References 500Exercises 50116. Distribution-Free Procedures 50716.1 Introduction 50716.2 One-Sample Procedures 50816.3 Two-Sample Procedures 51216.4 Nonparametric Analysis of Variance 51416.5 Exact Versus Approximate Tests 51916.6 Nonparametric Regression 51916.7 Nonparametric Prediction Intervals and Tolerance Intervals 52116.8 Summary 521References 521Exercises 52217. Tying It All Together 52517.1 Review of Book 52517.2 The Future 52717.3 Engineering Applications of Statistical Methods 528Reference 528Exercises 528Answers to Selected Excercises 533Appendix: Statistical Tables 562Table A Random Numbers 562Table B Normal Distribution 564Table C t-Distribution 566Table D F-Distribution 567Table E Factors for Calculating Two-Sided 99% Statistical Intervals for a Normal Population to Contain at Least 100p% of the Population 570Table F Control Chart Constants 571Author Index 573Subject Index 579

"Overall this is an excellent book, which defines a broader mandate than many of its competing texts. By providing, clear, understandable discussion of the basics of statistics through to more advanced methods commonly used by engineers, this book is an essential reference for practitioners, and an ideal text for a two semester course introducing engineers to the power and utility of statistics." (The American Statistician, August 2008) "In this book on modern engineering statistics, Ryan does an excellent job of providing the appropriate statistical concepts and tools using engineering resources.... Highly recommended. Lower- and upper-division undergraduates" (CHOICE, April 2008)"This self-contained volume motivates an appreciation of statistical techniques within the context of engineering; many datasets that are used in the chapters and exercises are from engineering sources. This book is ideal for either a one- or two-semester course in engineering statistics." (Computing Reviews, April 2008)