Download as pdf or txt
Download as pdf or txt
You are on page 1of 54

Introduction to Probability and

Statistics - Metric Version, 15e 15th


Edition William Mendenhall - eBook
PDF
Go to download the full and correct content document:
https://ebooksecure.com/download/introduction-to-probability-and-statistics-metric-ver
sion-15e-ebook-pdf/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Introduction to Probability and Statistics (Metric


Version) 15th Edition William Mendenhall - eBook PDF

https://ebooksecure.com/download/introduction-to-probability-and-
statistics-metric-version-ebook-pdf/

Introduction to Probability and Statistics Metric


Edition 1925-2009) Mendenhall - eBook PDF

https://ebooksecure.com/download/introduction-to-probability-and-
statistics-metric-edition-ebook-pdf/

(eBook PDF) Introduction to Probability and Statistics


3rd by William Mendenhall

http://ebooksecure.com/product/ebook-pdf-introduction-to-
probability-and-statistics-3rd-by-william-mendenhall/

(eBook PDF) Introduction to Probability and Statistics


15th Edition

http://ebooksecure.com/product/ebook-pdf-introduction-to-
probability-and-statistics-15th-edition/
Statistics for Business and Economics: Metric Version -
eBook PDF

https://ebooksecure.com/download/statistics-for-business-and-
economics-metric-version-ebook-pdf/

Introduction to Probability and Statistics 14th Edition


(eBook PDF)

http://ebooksecure.com/product/introduction-to-probability-and-
statistics-14th-edition-ebook-pdf/

(eBook PDF) Introduction to Probability and Statistics


4rd Canadian Edition

http://ebooksecure.com/product/ebook-pdf-introduction-to-
probability-and-statistics-4rd-canadian-edition/

(eBook PDF) Introduction to Probability and Statistics


for Engineers and Scientists 5th Edition

http://ebooksecure.com/product/ebook-pdf-introduction-to-
probability-and-statistics-for-engineers-and-scientists-5th-
edition/

Understanding basic statistics 8 metric version Edition


Charles Henry Brase - eBook PDF

https://ebooksecure.com/download/understanding-basic-statistics-
ebook-pdf-2/
■ Table 4 Critical Values of t, pages 692–963

  df t.100 t.050 t.025 t.010 t.005 df


a     1 3.078 6.314 12.706 31.821 63.657 1
ta     2 1.886 2.920   4.303   6.965   9.925 2
    3 1.638 2.353   3.182   4.541   5.841 3
    4 1.533 2.132   2.776   3.747   4.604 4
    5 1.476 2.015   2.571   3.365   4.032 5

    6 1.440 1.943   2.447   3.143   3.707 6


    7 1.415 1.895   2.365   2.998   3.499 7
    8 1.397 1.860   2.306   2.896   3.355 8
    9 1.383 1.833   2.262   2.821   3.250 9
  10 1.372 1.812   2.228   2.764   3.169 10

  11 1.363 1.796   2.201   2.718   3.106 11


  12 1.356 1.782   2.179   2.681   3.055 12
  13 1.350 1.771   2.160   2.650   3.012 13
  14 1.345 1.761   2.145   2.624   2.977 14
  15 1.341 1.753   2.131   2.602   2.947 15

  16 1.337 1.746   2.120   2.583   2.921 16


  17 1.333 1.740   2.110   2.567   2.898 17
  18 1.330 1.734   2.101   2.552   2.878 18
  19 1.328 1.729   2.093   2.539   2.861 19
  20 1.325 1.725   2.086   2.528   2.845 20

  21 1.323 1.721   2.080   2.518   2.831 21


  22 1.321 1.717   2.074   2.508   2.819 22
  23 1.319 1.714   2.069   2.500   2.807 23
  24 1.318 1.711   2.064   2.492   2.797 24
  25 1.316 1.708   2.060   2.485   2.787 25

  26 1.315 1.706   2.056   2.479   2.779 26


  27 1.314 1.703   2.052   2.473   2.771 27
  28 1.313 1.701   2.048   2.467   2.763 28
  29 1.311 1.699   2.045   2.462   2.756 29
  30 1.310 1.697   2.042   2.457   2.750 30

  31 1.309 1.696   2.040   2.453   2.744 31


  32 1.309 1.694   2.037   2.449   2.738 32
  33 1.308 1.692   2.035   2.445   2.733 33
  34 1.307 1.691   2.032   2.441   2.728 34
  35 1.306 1.690   2.030   2.438   2.724 35

  36 1.306 1.688   2.028   2.434   2.719 36


  37 1.305 1.687   2.026   2.431   2.715 37
  38 1.304 1.686   2.024   2.429   2.712 38
  39 1.304 1.685   2.023   2.426   2.708 39
  40 1.303 1.684   2.021   2.423   2.704 40

  45 1.301 1.679   2.014   2.412   2.690 45


  50 1.299 1.676   2.009   2.403   2.678 50
  55 1.297 1.673   2.004   2.396   2.668 55
  60 1.296 1.671   2.000   2.390   2.660 60
  65 1.295 1.669   1.997   2.385   2.654 65

  70 1.294 1.667   1.994   2.381   2.648 70


  80 1.292 1.664   1.990   2.374   2.639 80
  90 1.291 1.662   1.987   2.368   2.632 90
100 1.290 1.660   1.984   2.364   2.626 100
200 1.286 1.653   1.972   2.345   2.601 200

300 1.284 1.650   1.968   2.339   2.592 300


400 1.284 1.649   1.966   2.336   2.588 400
500 1.283 1.648   1.965   2.334   2.586 500
inf. 1.282 1.645 1.96   2.326   2.576 inf.

Source: Percentage points calculated using Minitab software.

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Body temperature and heart rate 542 Measurement error 263 Social Sciences
Breathing rates 74, 93, 441 Medical diagnostics 162
Achievement scores 581
Bulimia 392 Mercury concentration in dolphins 84, 594
Achievement tests 75, 515, 550
Calcium 440, 460, 464 Metal corrosion and soil acids 572
Adolescents and social stress 376
Calcium content 25 Metabolism and weight gain 552
Alcohol and altitude 442
Cancer survivor rates 12, 187 Monkey business 144 American Presidents-age at death 26
Cerebral blood flow 226, 263, 391, 419 MRIs 164 Animation helps 502
Chemical experiment 306, 515 Nematodes 549 Anxious infants 612
Chemotherapy 642 Omega-3 Fats 250 Back to work 17
Chicago weather 186 Ore samples 74 Books or iPads? 401
Chirping crickets 111, 523, 528 Parasites in foxes 94 Boomers, Xers and Millennial Men 375
Chloroform 92 PCBs 377 Catching a cold 318, 321, 630
Cholesterol 393 Pearl millet 340 Choosing a mate 158
Citrus red mite 325 pH in rainfall 326 Discovery-based teaching 627
Color preferences in mice 210 pH levels in water 660 Drug offenders 156
Cotton versus cucumber 580 Physical fitness 500, 612 Drug testing 156
Cure for insomnia 364 Plant density 208 Eye movement 642
Cure for the common cold 358 Plant genetics 157, 188, 234, 363 Faculty salaries 263, 486, 501
Deep-sea research 617 Plant science 537 Good at math? 460
Diabetes in children 208 Polluted seawater 67, 84 Graduate teaching assistants 628
Digitalis and calcium uptake 475 Pollution 326, 499, 678 Hospital survey 143
Disinfectants 402 Potassium levels 264 Household size 101, 617
Dissolved O2 content 392, 403, 459, 642 Potency of an antibiotic 353 Images and word recall 251, 654
Drugs for hypertension 92 Pulse rates 49, 227 Intensive care 195
Drug potency 420 Purifying organic compounds 392 Jury duty 136
E. coli outbreaks 196 Rain and snow 121 Laptops and learning 51, 524, 528
Early detection of breast cancer 363 Recovery rates 647 Math and art 677
Enzymes 401 Recurring illness 32, 91 Medical bills 189
Excedrin or Tylenol 318 Red blood cell count 25, 393 Memory experiments 412
FDA testing 175 Rh factor 233, 285 Midterm scores 118
Fossils 440 Ring-necked pheasants 440 Music in the workplace 412
Fruit flies 136 Runners and cyclists 402, 428, 443 No pass-no play rule for athletics 162
Geothermal power 542 San Andreas Fault 296 Organized religion 25
Genetic defects 233 Screening tests 162 Political corruption 326
Gestation times 121, 226, 523 Sea urchins 440 Preschool 33
Glucose tolerance 464 Seed treatments 199 Racial bias 250
Good tasting medicine 619, 665 Selenium 311, 326 Reaction times 410, 441, 442,
Ground or air 411 Shade or sun? 440 497, 498
Gulf oil spill 48 Slash pine seedlings 474 Reducing hostility 458
Hazardous waste 26, 123 Sleep deprivation 515, 523 Same-sex marriage 284, 306
Healthy eating 358, 579 Smoking 331, 392 SAT scores 92, 187, 313, 359,
Healthy teeth 401, 411 Sodium hydroxide 439 376, 427
Heart rate and exercise 465, 659 Spraying fruit trees 358 Smoking and cancer 157
Hormone therapy and Alzheimer’s disease 368 Sunflowers 227, 332 Snacking and TV 242
Human body temperatures 50, 95, 242, 264, 307, Survival times 32, 74, 85 Social ambivalence 92
313, 353, 359 Swampy sites 459, 464, 659 Social Security numbers 74
Hungry rats 297 Sweet potato whitefly 363 Social skills training 110, 541, 671
Impurities 428, 439 Tai Chi and fibromyalgia 251, 368 Spending patterns 612
Iodine concentration 331 Taste test for PTC 189 Starting salaries 312, 321, 359
Jigsaw puzzles 654 Tay-Sachs disease 188 Student ratings 671
Lead levels in blood 647 Titanium 402 Teaching biology 312
Lead levels in drinking water 358 Toxic chemicals 664 Test interviews 119, 515
Less red meat 321, 579 Weights of babies 225, 263, 305, 352 Unbiased choices 144, 174, 199
Lobsters 392, 541 Weights of turtles 642 Union Yes! 318
Long stemmed roses 92 Whitefly infestation 210, 499 Violent crime 162
Lung cancer 233 White tailed deer 376 Want to be President? 16

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Edition
Introduction
15 to Probability
and Statistics
Metric Version
William Mendenhall, III
1925–2009

Robert J. Beaver
University of California, Riverside, Emeritus

Barbara M. Beaver
University of California, Riverside, Emerita

Australia • Brazil • Mexico • Singapore • United Kingdom • United States

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
This is an electronic version of the print textbook. Due to electronic rights restrictions,
some third party content may be suppressed. Editorial review has deemed that any suppressed
content does not materially affect the overall learning experience. The publisher reserves the right
to remove content from this title at any time if subsequent rights restrictions require it. For
valuable information on pricing, previous editions, changes to current editions, and alternate
formats, please visit www.cengage.com/highered to search by ISBN#, author, title, or keyword for
materials in your areas of interest.

Important Notice: Media content referenced within the product description or the product
text may not be available in the eBook version.

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Introduction to Probability and Statistics, © 2020, 2013, 2009 Cengage Learning, Inc.
Fifteenth Edition, Metric Version WCN: 02-300
William Mendenhall, III, Robert J. ALL RIGHTS RESERVED. No part of this work covered by the copyright
­Beaver, Barbara M. Beaver herein may be reproduced or distributed in any form or by any means,
except as permitted by U.S. copyright law, without the prior written
Metric Version prepared by Qaboos Imran
permission of the copyright owner.
International Product Director: Timothy L.
Anderson

Senior Product Assistant: Alexander Sham For product information and technology assistance, contact us at
Cengage Customer & Sales Support, 1-800-354-9706
Content Manager: Marianne Groth
or support.cengage.com.
Associate Marketing Manager: Tori Sitcawich
For permission to use material from this text or product, submit all
Associate Content Managers: Abby DeVeuve, requests online at www.cengage.com/permissions.
Amanda Rose

Manufacturing Planner: Doug Bertke


ISBN: 978-0-357-11446-9
IP Analyst: Reba Frederics
Cengage International Offices
IP Project Manager: Carly Belcher
Asia Australia/New Zealand
Production Service/Compositor: SPi Global www.cengageasia.com www.cengage.com.au
Inventory Analyst: Sarah Ginsberg tel: (65) 6410 1200 tel: (61) 3 9685 4111

Art Director: Vernon T. Boes Brazil India


www.cengage.com.br www.cengage.co.in
Senior Designer: Diana Graham tel: (55) 11 3665 9900 tel: (91) 11 4364 1111
Cover Image: mikroman6/Getty Images
Latin America UK/Europe/Middle East/Africa
www.cengage.com.mx www.cengage.co.uk
tel: (52) 55 1500 6000 tel: (44) 0 1264 332 424

Represented in Canada by
Nelson Education, Ltd.
tel: (416) 752 9100 / (800) 668 0671
www.nelson.com

Cengage is a leading provider of customized learning solutions with


office locations around the globe, including Singapore, the United
Kingdom, Australia, Mexico, Brazil, and Japan. Locate your local office
at: www.cengage.com/global.

For product information: www.cengage.com/international


Visit your local office: www.cengage.com/global
Visit our corporate website: www.cengage.com

Printed in the United States of America


Print Number: 01   Print Year: 2018

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Brief Contents

introduction: WHAT IS STATISTICS?  1

1 Describing Data with Graphs  7

2 Describing Data with Numerical Measures  54

3 Describing Bivariate Data  96

4 Probability  126

5 Discrete Probability Distributions  167

6 The Normal Probability Distribution  212

7 Sampling Distributions  245

8 Large-Sample Estimation  288

9 Large-Sample Tests of Hypotheses  335

10 Inference from Small Samples  380

11 The Analysis of Variance  445

12 Simple Linear Regression and Correlation  503

13 Multiple Linear Regression Analysis  555

14 Analysis of Categorical Data  599

15 Nonparametric Statistics  633

Appendix i  681

data sources  714

answers to selected exercises  727

index  745

iii

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Contents

Introduction: What Is Statistics? 1


The Population and the Sample 3
Descriptive and Inferential Statistics 3
Achieving the Objective of Inferential Statistics: The Necessary Steps 4
Keys for Successful Learning 5

1 Describing Data with Graphs 7


1.1 Variables and Data 8
Types of Variables 9
Exercises  11
1.2 Graphs for Categorical Data 12
Exercises  15
1.3 Graphs for Quantitative Data 17
Pie Charts and Bar Charts 17
Line Charts 19
Dotplots 20
Stem and Leaf Plots 20
Interpreting Graphs with a Critical Eye 22
Exercises  24
1.4 Relative Frequency Histograms 27
Exercises  31
Chapter Review 35
Technology Today 35
Reviewing What You’ve Learned 47
Case Study: How Is Your Blood Pressure? 53

2 Describing Data with Numerical Measures 54


Introduction 55
2.1 Measures of Center 55
Exercises  59
2.2 Measures of Variability 61
Exercises  66
iv

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Contents v

2.3 Understanding and Interpreting the Standard Deviation 67


Tchebysheff’s Theorem 67
The Empirical Rule 69
Approximating s Using the Range 71
Exercises  73
2.4 Measures of Relative Standing 76
z-Scores 76
Percentiles and Quartiles 77
The Five-Number Summary and the Box Plot 80
Exercises  83

Chapter Review 86
Technology Today 87
Reviewing What You’ve Learned 91
Case Study: The Boys of Summer 95

3 Describing Bivariate Data 96


Introduction 97
3.1 Describing Bivariate Categorical Data 97
Exercises  99
3.2 Describing Bivariate Quantitative Data 101
Scatterplots 101
The Correlation Coefficient 104
The Least-Squares Line 106
Exercises  109

Chapter Review 112


Technology Today 112
Reviewing What You’ve Learned 118
Case Study: Are Your Clothes Really Clean? 124

4 Probability 126
Introduction 127
4.1 Events and the Sample Space 127
Exercises  130
4.2 Calculating Probabilities Using Simple Events 131
Exercises  134
4.3 Useful Counting Rules 137
Using the TI-83/84 Plus Calculator 142
Exercises  142

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
vi Contents

4.4 Rules for Calculating Probabilities 144


Calculating Probabilities for Unions and Complements 146
Calculating Probabilities for Intersections 148
Exercises  155
4.5 Bayes’ Rule 158
Exercises  161
Chapter Review 163
Reviewing What You’ve Learned 163
Case Study: Probability and Decision Making in the Congo 166

5 Discrete Probability Distributions 167


5.1 Discrete Random Variables and Their ­Probability Distributions 168
Random Variables 168
Probability Distributions 168
The Mean and Standard Deviation for a Discrete Random Variable 170
Exercises  174
5.2 The Binomial Probability Distribution 176
Exercises  185
5.3 The Poisson Probability Distribution 189
Exercises  194
5.4 The Hypergeometric Probability Distribution 196
Exercises  198

Chapter Review 200


Technology Today 201
Reviewing What You’ve Learned 206
Case Study: A Mystery: Cancers Near a Reactor 211

6 The Normal Probability Distribution 212


6.1 Probability Distributions for Continuous Random Variables 213
The Continuous Uniform Probability Distribution 215
The Exponential Probability Distribution 216
Exercises  217
6.2 The Normal Probability Distribution 218
The Standard Normal Random Variable 219
Calculating Probabilities for a General Normal Random Variable 222
Exercises  225
6.3 The Normal Approximation to the Binomial Probability Distribution 228
Exercises  232

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Contents vii

Chapter Review 235


Technology Today 235
Reviewing What You’ve Learned 241
Case Study: “Are You Going to Curve the Grades?” 244

7 Sampling Distributions 245


Introduction 246
7.1 Sampling Plans and Experimental Designs 246
Exercises  249
7.2 Statistics and Sampling Distributions 252
Exercises  254
7.3 The Central Limit Theorem and the Sample Mean 255
The Central Limit Theorem 255
The Sampling Distribution of the Sample Mean 258
Standard Error of the Sample Mean 259
Exercises  262
7.4 Assessing Normality 264
7.5 The Sampling Distribution of the Sample Proportion 268
Exercises  271
7.6 A Sampling Application: Statistical Process Control (Optional) 273
A Control Chart for the Process Mean: The x Chart 274
A Control Chart for the Proportion Defective: The p Chart 276
Exercises  278

Chapter Review 280


Technology Today 281
Reviewing What You’ve Learned 284
Case Study: Sampling the Roulette at Monte Carlo 287

8 Large-Sample Estimation 288


8.1 Where We’ve Been and Where We’re Going 289
Statistical Inference 289
Types of Estimators 290
8.2 Point Estimation 291
Exercises  296
8.3 Interval Estimation 298
Constructing a Confidence Interval 298
Large-Sample Confidence Interval for a Population Mean  300

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
viii Contents

Interpreting the Confidence Interval 301


Large-Sample Confidence Interval for a Population Proportion p 303
Using Technology 304
Exercises  304
8.4 Estimating the Difference Between Two Population Means 307
Exercises  311
8.5 Estimating the Difference Between Two Binomial Proportions 313
Using Technology 316
Exercises  316
8.6 One-Sided Confidence Bounds 319
Exercises  320
8.7 Choosing the Sample Size 322
Exercises  325

Chapter Review 326


Technology Today 327
Reviewing What You’ve Learned 330
Case Study: How Reliable Is That Poll? CBS News: How and Where America Eats 333

9 Large-Sample Tests of Hypotheses 335


Introduction 336
9.1 A Statistical Test of Hypothesis 336
Exercises  339
9.2 A Large-Sample Test About a Population Mean 340
The Essentials of the Test 340
Calculating the p-Value 344
Two Types of Errors 348
The Power of a Statistical Test 349
Exercises  352
9.3 A Large-Sample Test of Hypothesis for the ­Difference Between
Two Population Means 354
Hypothesis Testing and Confidence Intervals 356
Exercises  357
9.4 A Large-Sample Test of Hypothesis
for a Binomial Proportion 360
Statistical Significance and Practical Importance 362
Exercises  363
9.5 A Large-Sample Test of Hypothesis for the ­Difference
Between Two Binomial Proportions 365
Exercises  367

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Contents ix

9.6 Concluding Comments on Testing Hypotheses 369

Chapter Review 370


Technology Today 371
Reviewing What You’ve Learned 375
Case Study: An Aspirin a Day . . . ? 378

10 Inference from Small Samples 380


Introduction 381
10.1 Student’s t Distribution 381
Assumptions behind Student’s t Distribution 384
Exercises  385
10.2 Small-Sample Inferences Concerning a Population Mean 386
Exercises  390
10.3 Small-Sample Inferences for the Difference Between Two Population Means:
Independent Random Samples 394
Exercises  400
10.4 Small-Sample Inferences for the Difference Between Two Means: A Paired-­
Difference Test 404
Exercises  409
10.5 Inferences Concerning a Population Variance 413
Exercises  419
10.6 Comparing Two Population Variances 421
Exercises  427
10.7 Revisiting the Small-Sample Assumptions 429

Chapter Review 430


Technology Today 431
Reviewing What You’ve Learned 439
Case Study: School Accountability—Are We Doing Better? 443

11 The Analysis of Variance 445


11.1 The Design of an Experiment 446
Basic Definitions 446
What Is an Analysis of Variance? 447
The Assumptions for an Analysis of Variance 448
Exercises  448
11.2 The Completely Randomized Design: A One-Way Classification 449
Partitioning the Total Variation in the Experiment 450

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
x Contents

Testing the Equality of the Treatment Means 453


Estimating Differences in the Treatment Means 455
Exercises  458
11.3 Ranking Population Means 461
Exercises  464
11.4 The Randomized Block Design: A Two-Way Classification 465
Partitioning the Total Variation in the Experiment 466
Testing the Equality of the Treatment and Block Means 469
Identifying Differences in the Treatment and Block Means 471
Some Cautionary Comments on Blocking 472
Exercises  473
11.5 The a 3 b Factorial Experiment: A Two-Way Classification 477
The Analysis of Variance for an a 3 b Factorial Experiment 479
Exercises  483
11.6 Revisiting the Analysis of Variance Assumptions 486
Residual Plots 487
11.7 A Brief Summary 489

Chapter Review 490


Technology Today 490
Reviewing What You’ve Learned 497
Case Study: How to Save Money on Groceries! 502

12 Simple Linear Regression and Correlation 503


Introduction 504
12.1 Simple Linear Regression 504
A Simple Linear Model 505
The Method of Least Squares 507
Exercises  509
12.2 An Analysis of Variance for Linear Regression 511
Exercises  514
12.3 Testing the Usefulness of the Linear ­Regression Model 516
Inferences About b, the Slope of the Line of Means 516
The Analysis of Variance F-Test 519
Measuring the Strength of the Relationship:
The Coefficient of Determination 520
Interpreting the Results of a Significant Regression 521
Exercises  522

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Contents xi

12.4 Diagnostic Tools for Checking the Regression Assumptions 525


Dependent Error Terms 525
Residual Plots 525
Exercises  526
12.5 Estimation and Prediction Using the Fitted Line 530
Exercises  534
12.6 Correlation Analysis 537
Exercises  540

Chapter Review 543


Technology Today 544
Reviewing What You’ve Learned 549
Case Study: Is Your Car “Made in the U.S.A.”? 553

13 Multiple Linear Regression Analysis 555


Introduction 556
13.1 The Multiple Regression Model 556
13.2 Multiple Regression Analysis 558
The Method of Least Squares 558
The Analysis of Variance 559
Testing the Usefulness of the Regression Model 561
Interpreting the Results of a Significant Regression 562
Best Subsets Regression 563
Checking the Regression Assumptions 564
Using the Regression Model for Estimation and Prediction 564
Exercises  565
13.3 A Polynomial Regression Model 567
Exercises  570
13.4 Using Quantitative and Qualitative Predictor Variables in a Regression
­Model 573
Exercises  578
13.5 Testing Sets of Regression Coefficients 582
13.6 Other Topics in Multiple Linear Regression 584
Interpreting Residual Plots 584
Stepwise Regression Analysis 586
Binary Logistic Regression 587
Misinterpreting a Regression Analysis 587
13.7 Steps to Follow When Building a Multiple Regression Model 589

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
xii Contents

Chapter Review 589


Technology Today 590
Reviewing What You’ve Learned 592
Case Study: “Made in the U.S.A.”—Another Look 598

14 Analysis of Categorical Data 599


14.1 The Multinomial Experiment and the Chi-Square Statistic 600
14.2 Testing Specified Cell Probabilities: The Goodness-of-Fit Test 602
Exercises  604
14.3 Contingency Tables: A Two-Way Classification 606
The Chi-Square Test of Independence 607
Exercises  611
14.4 Comparing Several Multinomial Populations: A Two-Way Classification with
Fixed Row or Column Totals 614
Exercises  616
14.5 Other Topics in Categorical Data Analysis 619
The Equivalence of Statistical Tests 619
Other Applications of the Chi-Square Test 620

Chapter Review 621


Technology Today 622
Reviewing What You’ve Learned 627
Case Study: Who Is the Primary Breadwinner in Your Family? 631

15 Nonparametric Statistics 633


Introduction 634
15.1 The Wilcoxon Rank Sum Test: Independent Random Samples 634
Normal Approximation for the Wilcoxon Rank Sum Test 638
Exercises  641
15.2 The Sign Test for a Paired Experiment 643
Normal Approximation for the Sign Test 644
Exercises  646
15.3 A Comparison of Statistical Tests 648
15.4 The Wilcoxon Signed-Rank Test for a Paired Experiment 648
Normal Approximation for the Wilcoxon Signed-Rank Test 652
Exercises  653
15.5 The Kruskal–Wallis H-Test for Completely ­Randomized Designs 655
Exercises  658

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Contents xiii

15.6 The Friedman Fr-Test for Randomized Block Designs 660


Exercises  663
15.7 Rank Correlation Coefficient 666
Exercises  670
15.8 Summary 672

Chapter Review 672


Technology Today 673
Reviewing What You’ve Learned 676
Case Study: Amazon HQ2 680

Appendix I 681
Table 1 Cumulative Binomial Probabilities  682
Table 2 Cumulative Poisson Probabilities  688
Table 3 Areas under the Normal Curve  690
Table 4 Critical Values of t  692
Table 5 Critical Values of Chi-Square  694
Table 6 Percentage Points of the F Distribution  696
Table 7 Critical Values of T for the Wilcoxon Rank
Sum Test, n1 # n2  704
Table 8 Critical Values of T for the Wilcoxon Signed-Rank
Test, n 5 5(1)50  706
Table 9 Critical Values of Spearman’s Rank Correlation Coefficient
for a One-Tailed Test  707
Table 10 Random Numbers  708
Table 11 Percentage Points of the Studentized Range, q.05(k, df )  710

Data Sources 714

Answers to Selected Exercises 727

Index 745

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Preface

Every time you pick up a newspaper or a magazine, watch TV, or scroll through F ­ acebook,
you encounter statistics. Every time you fill out a questionnaire, register at an online
­website, or pass your grocery rewards card through an electronic scanner, your personal
information becomes part of a database containing your personal statistical information.
You can’t avoid it! In this digital age, data collection and analysis are part of our day-to-day
activities. If you want to be an educated consumer and citizen, you need to understand how
statistics are used and misused in our daily lives.
This international metric version is designed for classrooms and students outside of
the United States. The units of measurement used in selected examples and exercises have
been changed from U.S. Customary units to metric units. We did not update problems that
are specific to U.S. Customary units, such as passing yards in football or data ­related to
specific publications.

The Secret to Our Success


The first college course in introductory statistics that we ever took used Introduction to
Probability and Statistics by William Mendenhall. Since that time, this text—currently in
the fifteenth edition—has helped generations of students understand what statistics is all
about and how it can be used as a tool in their particular area of application. The secret to
the success of Introduction to Probability and Statistics is its ability to blend the old with
the new. With each revision we try to build on the strong points of previous editions, and
to look for new ways to motivate, encourage, and interest students using new technologies.

Hallmark Features of the Fifteenth Edition


The fifteenth edition keeps the traditional outline for the coverage of descriptive and
­inferential statistics used in previous editions. This revision maintains the straightforward
presentation of the fourteenth edition. We have continued to simplify the language in order to
make the text more readable—without sacrificing the statistical integrity of the presentation.
We want students to understand how to apply statistical procedures, and also to understand
• how to meaningfully describe real sets of data
• how to explain the results of statistical tests in a practical way
• how to tell whether the assumptions behind statistical tests are valid
• what to do when these assumptions have been violated

Exercises
As with all previous editions, the variety and number of real applications in the exercise
sets is a major strength of this edition. We have revised the exercise sets to provide new and
xv

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
xvi Preface

interesting real-world situations and real data sets, many of which are drawn from current
periodicals and journals. The fifteenth edition contains over 1900 exercises, many of which
are new to this edition. Exercises are graduated in level of difficulty; some, involving only
basic techniques, can be solved by almost all students, while others, involving practical
applications and interpretation of results, will challenge students to use more sophisticated
statistical reasoning and understanding. Exercises have been rearranged to provide a more
even distribution of exercises within each chapter and a new numbering system has been
introduced, so that numbering begins again with each new section.

Organization and Coverage


We believe that Chapters 1 through 10—with the possible exception of Chapter 3—should
be covered in the order presented. The remaining chapters can be covered in any order.
The analysis of variance chapter precedes the regression chapter, so that the instructor can
present the analysis of variance as part of a regression analysis. Thus, the most effective
presentation would order these three chapters as well.
Chapters 1–3 present descriptive data analysis for both one and two variables, using
MINITAB 18, Microsoft Excel 2016®, and TI-83/84 Plus graphics. Chapter 4 includes a full
presentation of probability. The last section of Chapter 4 in the fourteenth edition of the
text, “Discrete Random Variables and Their Probability Distributions” has been moved to
become the first section in Chapter 5. As in the fourteenth edition, the chapters on analysis
of variance and linear regression include both calculational formulas and computer print-
outs in the basic text presentation. These chapters can be used with equal ease by instructors
who wish to use the ­“hands-on” computational approach to linear regression and ANOVA
and by those who choose to ­focus on the interpretation of computer-generated statistical
printouts. This edition includes ­expanded coverage of the uniform and exponential distri-
butions in Chapter 5 and normal probability plots for assessing normality in Chapter 7,
in ­addition to an expanded t-table (Table 4 in ­Appendix I). New topics in Chapter 13 ­include
best subsets regression procedures and binary logistic regression.
One important feature in the hypothesis testing chapters involves the emphasis on
p-values and their use in judging statistical significance. With the advent of computer-­
generated p-values, these probabilities have become essential in reporting the results of a
statistical analysis. As such, the observed value of the test statistic and its p-value are pre-
sented together at the outset of our discussion of statistical hypothesis testing as equivalent
tools for decision-making. Statistical significance is defined in terms of preassigned values
of , and the p-value approach is presented as an alternative to the critical value approach
for testing a statistical hypothesis. Examples are presented using both the p-value and
­critical value approaches to hypothesis testing. Discussion of the practical interpretation
of statistical results, along with the difference between statistical significance and practical
significance, is emphasized in the practical examples in the text.

Special Features of the Fifteenth Edition


• NEED TO KNOW. . .: This edition again includes highlighted sections called “NEED
TO KNOW. . .” and identified by this icon. These sections provide in-
formation consisting of definitions, procedures, or step-by-step hints on problem solv-
ing for specific questions such as “NEED TO KNOW… How to ­Construct a Relative
Frequency Histogram?” or “NEED TO KNOW… How to Decide Which Test to Use?”
• Graphical and numerical data description includes both traditional and EDA ­methods,
using computer graphics generated by MINITAB 18 for Windows and MS Excel 2016.
• Calculator screen captures from the TI-84 Plus calculator have been used for several
examples, allowing students to access this option for data analysis.

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
42.5to use. Many researchers report
you wish 30.0the smallest possible significance level for
Ford Escape which41.5 28.0
their results are statistically significant. Empirical Rule should work relatively well. That is,
Hyundai Tucson 41.5 28.0 • approximately 68% of the measurements will fall between 16.1
Jeep Cherokee For example,
43.5 the TI-84 plus output for Example
30.0 9.7 (Figure 9.9) shows z 5 0.9090909091
Jeep Compass with p-value41.5
5 0.182. Detailed instructions for28.0the TI-83/84 plus as well as MINITAB•can approximately
be 95% of the measurements will fall between 10.6
Jeep Patriot found in the41.0
Technology Today section at the26.0end of this chapter. These results are consistent Preface
• approximately 99.7% of the measurements will xvii
fall between 5.1
Kia Sportage with our hand
41.5calculations to the second decimal
28.0 place. Based on this p-value, H 0 cannot
Mazda C-5 be rejected.42.0
The results are not statistically 27.5
significant.
Toyota RAV4 42.0 30.0 Figure 2.11 6/25
Figure 9.9
Volkswagen Tiguan
TI-84 plus output for
42.0 28.0 Relative frequency
histogram for Example 2.8

Relative Frequency
Example 9.7
4/25
1. Since the data involve two variables and a third labeling variable, enter the data into
the first three columns of an Excel spreadsheet, using the labels in the table. Select Data
➤ Data Analysis ➤ Descriptive Statistics, and click OK. Highlight or type the Input 2/25

range (the data in the second and third columns) into the Descriptive Statistics Dialog
box (Figure 2.19(a)). Type an Output location, make sure the boxes for “Labels in First 0
Row” and “Summary Statistics” are both checked, and click OK. The summary statistics 8.5 14.5 20.5 26.5 32.5
Scores
(Figure 2.19(b)) will appear in the selected location in your spreadsheet.
(a) Sometimes it is easy to confuse the significance(b) level  with the p-value (or observed
Using Tchebysheff’s Theorem and the Empirical Rule
significance level). They are both probabilities calculated as areas in the tails of the sampling
distribution of the test statistic. However, the significance level  is preset by the experi-
menter before collecting the data. The p-value is linked directly to the data and actually
describes how likely or unlikely the sample results are, assuming that H 0 is true. The smaller
Tchebysheff’s Theorem gives a lower bound
the p-value, the more unlikely it is that H 0 is true!
interval x 6 ks. At least 1 2 (1/k 2 )
probably more!
? Need to Know…
Rejection Regions, p-Values, and Conclusions distribution).
The significance level, a , lets you set the risk that you are willing to take of making
an incorrect decision in a test of hypothesis.
• To set a rejection region, choose a critical value of z so that the area in the
mate of the fraction of measurements falling within 1, 2, or 3
tail(s) of the z distribution is (are) either  for a one-tailed test or a /2 for a
mean.
two-tailed test. Use the right tail for an upper-tailed test and the left tail for a
lower-tailed test. Reject H 0 when the test statistic exceeds the critical value
2. You may notice that some ofand
the cells in the spreadsheet are overlapping. To adjust
• falls
Allinexamples
the rejectionand
this, highlight the affected columns and click the Home tab. In the Cells group,
exercises in the text that contain
region. printouts or calculator
Approximating screen
s Using the Range
• To find
choose Format ➤ AutoFit Column
a p-value,are
­captures
Width.
find based
You
the area in
may ontheMINITAB
want
tail “beyond” 18,
theMS Excel 2016,
the test statistic. If the or the TI-84 Plus calculator.
test is one-tailed, this is the p-value. Ifto
themodify appearance
test is two-tailed, this is only
thehalf
calculation of s
of the output by decreasing the Theseand
thedecimal
p-value outputs
accuracy are provided
in certain
must be doubled. H 0for
cells.
Reject some
Highlight
when exercises,
the appro-
the p-value is less thanwhile
a . other
measurements
exercises require the
lie within two
priate cells and click the Decreasestudent
Decimalto obtain
icon solutions
(Home tab, without
Number using
group)a computer.
to
modify the output. We have displayed 48 the accuracy
chapter 1 Describingto three
Data decimal places.
with Graphs

Name Length (km) Name Length (km) d. Use a bar graph to show the percentage of federal
Gulf fishing areas closed.
Superior 560 Titicaca 195
54428_ch09_hr_335-379.indd 347 Victoria 334 Nicaragua
54428_ch02_hr_054-095.indd 71
163 e. Use a line
9/4/18chart
12:10 to
PM show the amounts of dispersants

Huron 330 Athabasca 333 used. Is there any underlying straight line relation-
Michigan 491 Reindeer 229 ship over time?
Aral Sea 416 Tonle Sap 112
Tanganyika 672 Turkana 246
Baykal 632 Issyk Kul 184 DATA 7. election Results The 2016 election was a race
SET
Great Bear 307 Torrens 208 in which Donald Trump defeated Hillary Clinton
DS0129 8:37 AM
Nyasa 576 Vänern 146 9/24/18 and other candidates, winning 304 electoral votes,
Great Slave 477 Nettilling 107 or 57% of the 538 available. However, Trump only won
Erie 386 Winnipegosis 226
Winnipeg 426 Albert 160 46.1% of the popular vote, while Clinton won 48.2%.
Ontario 309 Nipigon 115 The popular vote (in thousands) for Donald Trump in
Balkhash 602 Gairdner 144 each of the 50 states is listed as follows18:
Ladoga 198 Urmia 144
Maracaibo 213 Manitoba 224 AL 1319 HI 129 MA 1091 NM 320 SD 228
Onega 232 Chad 280 AK 163 ID 409 MI 2280 NY 2820 TN 1523
Eyre 144 AZ 1252 IL 2146 MN 1323 NC 2363 TX 4685
AR 685 IN 1557 MS 701 ND 217 UT 515
Source: The World Almanac and Book of Facts 2017 CA 4484 IA 801 MO 1595 OH 2841 VT 95
CO 1202 KS 671 MT 279 OK 949 VA 1769
a. Use a stem and leaf plot to describe the lengths of CT 673 KY 1203 NE 496 OR 782 WA 1222
the world’s major lakes. DE 185 LA 1179 NV 512 PA 2971 WV 489
FL 4618 ME 336 NH 346 RI 181 WI 1405
b. Use a histogram to display these same data. How
GA 2089 MD 943 NJ 1602 SC 1155 WY 174
does this compare to the stem and leaf plot in part a?
c. Are these data symmetric or skewed? If skewed, a. By just looking at the table, what shape do you think
what is the direction of the skewing? the distribution for the popular vote by state will
have?
DATA 6. Gulf oil Spill Cleanup On April 20, 2010, the
SET b. Draw a relative frequency histogram to describe the
United States experienced a major environmental
DS0128 distribution of the popular vote for President Trump
disaster when a Deepwater Horizon drilling rig
in the 50 states.
exploded in the Gulf of Mexico. The number of person-
nel and equipment used in the Gulf oil spill cleanup, c. Did the histogram in part b confirm your guess in
beginning May 2, 2010 (Day 13) through June 9, 2010 part a? Are there any outliers? How can you explain
(Day 51) is given in the following table.17 them?

Day 13 Day 26 Day 39 Day 51 DATA 8. election Results, continued Refer to Exercise 7.
SET
Number of personnel (1000s) 3.0 17.5 20.0 24.0 Listed here is the percentage of the popular vote
DS0130
Federal Gulf fishing areas closed 3% 8% 25% 32% received by President Trump in each of the
Booms laid (kilometers) 74 504 1030 1454 50 states18:
Dispersants used (1000 liters) 590 1893 3293 4326
Vessels deployed (100s) 1.0 6.0 14.0 35.0 AL 62 HI 30 MA 33 NM 40 SD
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
62
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to removeAK 51 content
additional ID at any
59time ifMIsubsequent
47 rights
NY restrictions
37 TN 61
require it.
xviii Preface

The Role of Computers and Calculators in the


Fifteenth Edition—Technology Today
Computers and scientific or graphing calculators are now common tools for college students
in all disciplines. Most students are accomplished users of word processors, spreadsheets, and
databases, and they have no trouble navigating through software p­ ackages in the Windows en-
vironment. Many own either a scientific or a graphing calculator, very often one of the many
calculators made by Texas Instruments.™ We believe, however, that advances in computer
technology should not turn statistical analyses into a “black box.” Rather, we choose to use
the computational shortcuts that modern technology provides to give us more time to empha-
size statistical reasoning as well as the understanding and interpretation of statistical results.
In this edition, students will be able to use computers both for standard statistical analy-
ses and as a tool for reinforcing and visualizing statistical concepts. Both MS Excel 2016 and
MINITAB 18 are used exclusively as the computer packages for statistical analysis along with
procedures available using the TI-83 or TI-84 Plus calculators. However, we have chosen to
isolate the instructions for generating computer and calculator output into individual sections
called Technology Today at the end of each chapter. Each discussion uses numerical examples
to guide the student through the MS Excel commands and option necessary for the procedures
presented in that chapter, and then present the equivalent steps and commands needed to pro-
duce the same or similar results using MINITAB and the TI-83/84 Plus. We have included screen
captures from MS Excel, MINITAB 18, and the TI-84 Plus, so that the student can actually
Technology Today 89 work
through these sections as “mini-labs.”
2. Now
If you do not need click on the Variables
“hands-on” knowledge box and of MINITAB,
select MSfrom
both columns Excel,
the listoronthe
the left. (You can Plus, or if
TI-83/84
click on the Graphs option and choose one of several graphs if you like. You may also click
you are using another oncalculator or software package, you may choose to skip these
the Statistics option to select the statistics you would like to see displayed.) Click OK. sections and
simply use the printouts as guides
A display for
of descriptive the basic
statistics
Technology Today 87
forunderstanding of computer
both columns will appear in the Session or calculator
window (see outputs.
Figure 2.20(b)). You may print this output using File ➤ Print Session Window if you choose.
3. To examine the distribution of the two variables and look for outliers, you can create box
technOLOgy tOday plots using the command Graph ➤ Boxplot ➤ One Y ➤ Simple. Click OK. Select the
appropriate column of measurements in the Dialog box (see Figure 2.21(a)). You can
Numerical Descriptive measures in Excel
88 change the appearance of the box plot in several ways. Scale ➤ Axes and Ticks will
chapter 2 Describing Data with Numerical Measures
MS Excel provides most of the basic descriptive statistics presented in Chapter 2 using a
single command on the Data tab. Other descriptive allow you
statistics cantobetranspose
calculatedthe axes
using theand orient the box plot horizontally, when you check the
Function Library group on the Formulas tab. box marked “Transpose value and category scales.” Multiple Graphs provides printing
3. Notice that the sample quartiles, Q1 and Q 3, are not given in the Excel output in
options for multiple box plots. Labels will let you annotate the graph with titles and
e x a m p l e 2. 17 Figure 2.19(b). You can calculateentered
the quartiles using the function command. Place your
The following data are the front and rear leg rooms (in footnotes.
inches) for 10Ifdifferent
you have data into the worksheet as a frequency distribution (values
compact sports
utility vehicles :
13 cursor into inanone
empty
column,cellfrequencies
and select in Formulas ➤ More
another), the Functions
Data Options will➤ Statistical
allow ➤be read
the data to
QUARTILE.EXC. in that Highlight
format. The theplot
box appropriate
for the cellsleg
front in rooms
the box
is marked
shown “Array”
in Figure and type
2.21(b).
Make & Model Front Leg Room
an integer (0Rear Leg Room
5 min, 1 5 first quartile, 2 5 median, 3 5 third quartile, or 4 5 max) in the
Chevrolet Equinox 42.5 4. Save this 30.0 worksheet in a file called “Leg Room” before exiting MINITAB. We will use it
Ford Escape 41.5 box marked “Quart.”
again 28.0
The quartile (calculated using this textbook’s method) will appear
in Chapter 3.
Hyundai Tucson 41.5 in the cell you have
28.0 chosen. Using the two quartiles, you can calculate the IQR and
Jeep Cherokee 43.5 30.0
Jeep Compass Figure 2.21 41.5 construct a box plot
28.0 by hand.
Jeep Patriot 41.0 26.0
Kia Sportage 41.5 (a) 28.0 (b)
Numerical Descriptive measures in MINITAB
Mazda C-5
Toyota RAV4
42.0
42.0
27.5
30.0
Volkswagen Tiguan MINITAB provides most
42.0 28.0 of the basic descriptive statistics presented in Chapter 2 using a
single command in the drop-down menus.
1. Since the data involve two variables and a third labeling variable, enter the data into
the first three columns of an Excel spreadsheet, using the labels in the table. Select Data
➤ Data e xAnalysis
a m p l e➤ 2Descriptive
.18 Statistics, and click OK. Highlight or type the Input
range (the data in the second and The following
third columns) data areDescriptive
into the the front Statistics
and rearDialog
leg rooms (in inches) for 10 different compact sports
utilitylocation,
box (Figure 2.19(a)). Type an Output vehicles 13
make : sure the boxes for “Labels in First
Row” and “Summary Statistics” are both checked, and click OK. The summary statistics
(Figure 2.19(b)) will appear in the selected
Make location in your spreadsheet.
& Model Front Leg Room Rear Leg Room
Figure 2.19 (a) Chevrolet Equinox (b) 42.5 30.0
Ford Escape 41.5 28.0
Hyundai Tucson 41.5 28.0
Jeep Cherokee 43.5 30.0
Jeep Compass 41.5 28.0
Jeep Patriot 41.0 26.0
Numerical Descriptive
Kia Sportage measures on the41.5 TI-83/84 Plus Calculators 28.0
Mazda C-5 The TI-83/84 Plus calculators
42.0 can be used to calculate descriptive
27.5 statistics and create box
Toyota RAV4 plots using the stat ➤ CALC
42.0 and 2nd ➤ stat plot commands.
30.0
Volkswagen Tiguan 42.0 28.0
e x a m p l e 2 .1 9The following
1. Since the data involvedata
twoarevariables
the front and
andrear leg rooms
a third (in inches)
labeling for enter
variable, 10 different compact
the data into sports
the first utility
three vehicles
columns :of a MINITAB worksheet, using the labels in the table. Using the
13

2. You may notice that some of the cells in the spreadsheet


drop-down are overlapping.
menus, select To adjust
Stat ➤ Basic Statistics ➤ Display Descriptive Statistics. The
Make
this, highlight the affected columns and click the& Home
Model tab. In the Cells group,
Front Leg Room Rear Leg Room
Dialog box is shown in Figure 2.20(a).
choose Format ➤ AutoFit Column Width. You may want to modify the appearance
Chevrolet Equinox 42.5 30.0
of the output by decreasing the decimal accuracy FordinEscape
certain cells. Highlight the appro- 41.5 28.0
Figure 2.20
priate cells and click the Decrease DecimalHyundai icon (Home
Tucson tab, Number group) to 41.5 28.0
modify the output. We(a) have displayed the accuracy
Jeep to three decimal places.
Cherokee 43.5 30.0may be suppressed from the eBook and/or eChapter(s).
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights,(b)
some third party content
Editorial review has deemed that any suppressed content does not materially affectJeep Compass
the overall 41.5reserves the right to remove additional
learning experience. Cengage Learning 28.0
content at any time if subsequent rights restrictions require it.
The probability of observing 211 or fewer who prefer her brand is
P( x # 211) 5 p(0) 1 p(1) 1 ... 1 p(210) 1 p(211)
Preface xix
The normal approximation to this probability is the area to the left of 211.5 under a normal
curve with a mean of 250 and a standard deviation of 15. First calculate

Study Aids z5
x 2 m 211.5 2 250
s
5
15
5 2 2.57

The many and varied Then


exercises in the text provide the best learning tool for students em-
barking on a first coursePin statistics. The answers to all odd-numbered exercises are given
( x ≤ 211) ≈ P ( z < 2 2.57) 5 .0051
in the back of the text. Each application exercise has a title, making it easier for students
The probability
and instructors to immediately of observing
identify a sample
both value of 211
the context oforthe
less when
problemp 5 .10and
is so the
smallarea
that you
of
can conclude that one of two things has occurred: Either you have observed an unusual sample
application. All of theeven
basic exercises have been rewritten and all of the applied exercises
though really p 5 .10, or the sample reflects that the actual value of p is less than .10 and
restructured accordingperhaps
to ­increasing
closer to thedifficulty. New
observed sample exercises
proportion, have5been
211/2500 .08. introduced, dated
exercises have been deleted, and a new numbering system has been introduced within each
section.
6.3 exercises
the basics 12. P( x $ 6) and P( x . 6) when n 5 15 and p 5 .5
Normal Approximation? Can the normal approxima- 13. P(4 # x # 6) when n 5 25 and p 5 .2
tion be used to approximate probabilities for the bino-
mial random variable x, with values for n and p given 14. P( x $ 7) and P( x 5 5) when n 5 20 and p 5 .3
in Exercises 1–4? If not, is there another approximation 15. P( x $ 10) when n 5 20 and p 5 .4
that you could use?
1. n 5 25 and p 5 .6 2. n 5 45 and p 5 .05 applying the basics
3. n 5 25 and p 5 .3 4. n 5 15 and p 5 .5 16. A USA Today snapshot found that 47% of Ameri-
Using the Normal Approximation Find the mean and cans associate “recycling” with Earth Day.9 Suppose a
standard deviation for the binomial random variable x random sample of n 5 50 adults are polled and that the
using the information in Exercises 5–11. Then use the 47% figure is correct. Use the normal curve to approxi-
correction for continuity and approximate the probabili- mate the probabilities of the following events.
Students
ties should
using the normal be encouraged to use the
approximation. “NEED TO KNOW. . .” sections as they
occur
5. P ( xin
. the text.n 5The
9) when placement
25 and p 5 .6 of these sections is intended to answer questions as they
Recycling 47% Cleaning
would
6. P (6 normally arise
246 chapter 7 Sampling
# x # 9) when n 5in discussions.
Distributions
25 and p 5 .3 In addition, there are numerous hintslocal
23% called
parks,
beaches,etc.
“NEED
A TIP?” that appear in the margins of the text. The tips are short and 30%concise.
7. P (20 , x , 25) when n 5 100 and p 5 .2
8. P ( x . 22) when n 5 100 and p 5 .2
introduction Planting a tree

Taking care
9. P ( x $ 22) when n 5 100 and In the p5 previous
.2 three chapters, you have learned a lot about probabilityof distributions,
the Earth
such
as the binomial and normal distributions. The shape of the normal What do you relate most to Earthis
distribution determined
Day?
10.●P ( x #a 25)
● Need Tip?when n 5 100by and 5 .2 m and its standard deviation s , while the shape of the binomial distribution is
itspmean
parameter ⇔ population determined by p. These numerical descriptive measures—called parameters—are needed
P (355
11.Statistic ⇔# x # 360) when n 5 400 and p 5 .9
Sample a. Fewer than 30 individuals associate “recycling” with
to calculate the probability of observing sample results.
How Good Is 86YourCHAPTER
Approximation? Using
In practical
2 Describing Data with Table
situations,
Numerical in may beEarth
1 you
Measures
Day?
able to decide which type of probability distribution
Appendix I, find the exact valuesto usefor as the binomial
a model, but the values ofb.the
prob- More than 20that
parameters individuals
specify its associate
exact form “recycling”
are unknown.with
abilities in 26.Exercises 12–15. Then approximate Earth Day?
Snapshots Here
Here are are two
a few examples:
snapshots fromthe USA • Twenty-two percent of all fans are willing to pay
probabilitiesToday.
using the normal approximation with the c. More$75 than 10 individuals do not of associate
• The person conducting an opinion poll isorsure
more for the
that a ticket to one
responses the top
to his 100 concert
“agree/dis-
correction for continuity. Compare
• About 12% of America’s your
agree” answers.
questions
volunteers spend will follow
more “recycling”
tours. with Earth Day?
thana binomial distribution, but p, the proportion of those
Finally, sections
5 hours percalled Key Concepts
week volunteering.
who “agree” and Formulas
in the population, is Identify
unknown. appear inx each
the variable chapterandasany
being measured, a review
percentiles you can determine from this information.
in outline form of the material covered in that chapter.
• Fifty-eight percent of
• all
An cars in operation
agricultural are at
researcher least
believes that the yield per acre of a variety of wheat is
8 years old. approximately normally distributed, but the mean m and standard deviation s of the
yields are unknown.
CHAPTER REVIEW
In these cases, you must rely on the sample to learn about these parameters. The proportion of
54428_ch06_hr_212-244.indd 232
those who “agree” in the pollster’s sample provides information about the actual value of p.
2. The Empirical Rule can be used only for rela-
Key Concepts and Formulas
The mean and standard deviation of the researcher’s sample approximate the actual values of
tively mound-shaped data sets. Approximately
I.
m and s . If you want the sample to provide reliable
Measures of the Center of a Data Distribution 68%,information
95%, and 99.7% about the population,
of the measurements however, are
you must select your sample in a certain way! within one, two, and three standard deviations of
1. Arithmetic mean (mean) or average
the mean, respectively.
a. Population: m
∑ xi IV. Measures of Relative Standing
b. Sample of n measurements: x 5
7.1 Sampling plans and experimental Designs
n
1. Sample z-score: z 5
x2x
2. Median; position of the median 5 .5(n 1) s
3. Mode The way a sample is selected is called the sampling plan or p%
2. pth percentile; experimental design. Know-
of the measurements are
4. The medianing maythe be sampling
preferred toplan used in
the mean if thea particular situationsmaller, will often
and (100 2 allow
p)% are you to measure the
larger.
data are highlyreliability
skewed.or goodness of your inference. 3. Lower quartile, Q1; position of Q1 5 .25 (n 1)
Simple random sampling is a commonly used sampling plan in which every sample
II. Measures of Variability 4. Upper quartile, Q ; position of Q 5 .75 (n 1)
of size n has the same chance of being selected. For example,3 suppose you 3want to select
1. Range: R 5alargest sample 2 smallest
of size n 5 2 from a population containing 5. Interquartile
N 5 range:
4 objects.IQR If Q3 2four
5 the Q1 objects are
2. Variance identified by the symbols x 1, x 2 , x 3 , and x
V. 4 , there
The are six
Five-Numberdistinct pairs
Summary thatandcouldBox bePlots
selected,
a. Population as of
listed in Table 7.1. If the sample of n 5 1.
N measurements: 2 observations
The five-number is selected
summary: so that each of these
six samples has the same chance—one out of six or Min 1/6—of Q1 selection,
Median then Q3 the Max resulting
∑( x i 2 m ) 2
s 2 5 sample is called a simple random sample, or just a random sample.
N One-fourth of the measurements in the data set
lie between each of the four adjacent pairs of
b. Sample of n measurements:
numbers.
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole ■●Table 7.1Due toWays
or in part. of Selecting a SamplecontentofmaySize
( ∑ xi ) the right to remove additional2.content
electronic
2 rights, some third party 2 fromfrom
be suppressed 4 Objects
the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning
2 ∑ xi2 2 reserves Box atplots areifused
any time for detecting
subsequent outliers
rights restrictions and
require it.
xx Preface

Instructor Resources
WebAssign
WebAssign for Mendenhall/Beaver/Beaver’s Introduction to Probability and Statistics, 15th
Edition, Metric Version is a flexible and fully customizable online instructional solution
that puts powerful tools in the hands of instructors, empowering you to deploy assignments,
­instantly assess individual student and class performance, and help your students master the
course concepts. With WebAssign’s powerful digital platform and Introduction to Probability
and Statistics’s specific content, you can tailor your course with a wide range of assignment
settings, add your own questions and content, and access student and course analytics and
communication tools.

MindTap Reader
Available via WebAssign, MindTap Reader is Cengage’s next-generation eBook. MindTap
Reader provides robust opportunities for students to annotate, take notes, navigate, and
interact with the text. Instructors can edit the text and assets in the Reader, as well as add
videos or URLs.

Cognero
Cengage Learning Testing, powered by Cognero, is a flexible, online system that allows
you to import, edit, and manipulate content from the text’s Test Bank or elsewhere—­
including your own favorite test questions; create multiple test versions in an instant; and
deliver tests from your LMS, your classroom, or wherever you want.

Instructor Solutions Manual


This time-saving online manual provides complete solutions to all the problems in the text.
You can download the solutions manual from the Instructor Companion Website.

Instructor Companion Website


Everything you need for your course in one place! This collection of book-specific class
tools is available online via www.cengage.com/login. Access and download PowerPoint
presentations, images, Instructor Solutions Manual, data sets, and more.

SnapStat
Tell the story behind the numbers with SnapStat in WebAssign. Designed with students
to bring stats to life, SnapStat uses interactive visuals to perform complex analysis online.
Labs and Projects in WebAssign allow students to crunch their own data or choose from
pre-existing data sets to get hands-on with technology and see for themselves that Statistics
is much more than just numbers.

Student Resources
WebAssign
WebAssign for Mendenhall/Beaver/Beaver’s Introduction to Probability and Statistics,
15th Edition, Metric Version lets you prepare for class with confidence. Its online learning

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Preface xxi

platform for your math, statistics, and science courses helps you practice and absorb what
you learn. Videos and tutorials walk you through concepts when you’re stuck, and instant
feedback and grading let you know where you stand—so you can focus your study time
and perform better on in-class assignments. Study smarter with WebAssign!

MindTap Reader
Available via WebAssign, MindTap Reader is Cengage’s next-generation eBook. ­MindTap
Reader provides robust opportunities for students to annotate, take notes, navigate, and
­interact with the text. Annotations captured in MindTap are automatically tied to the
­Notepad app, where they can be viewed chronologically and in a cogent, linear fashion.

Online Technology Guides


Online Technology Guides, accessed via www.cengage.com, provide step-by-step ­instructions
for completing problems using common statistical software.

SnapStat
Learn the story behind the numbers with SnapStat in WebAssign. Designed with students
to bring stats to life, SnapStat uses interactive visuals to perform complex analysis online.
Labs and Projects in WebAssign allow you to crunch your own data or choose from pre-
existing data sets to get hands-on with technology and see for yourself that Statistics is
much more than just numbers.

Acknowledgments
The authors are grateful to Catherine Van Der Laan and the editorial staff of Cengage
Learning for their patience, assistance, and cooperation in the preparation of this edition.
Thanks are also due to fifteenth edition reviewers Olcay Akman, Matt Harris, ­Zhongming
Huang, Bo Kai, Sarah Miller, and Katie Wheaton. We wish to thank ­authors and organiza-
tions for allowing us to reprint selected material; ­acknowledgments are made wherever
such material appears in the text.
Robert J. Beaver
Barbara M. Beaver

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Introduction
What Is ­Statistics?

What is statistics? Have you ever met a statistician? Do


you know what a statistician does? Maybe you are think-
ing of the person who sits in the broadcast booth at the
Rose Bowl, recording the number of pass completions,
yards rushing, or interceptions thrown on New Year’s Day.
Or maybe just hearing the word statistics sends a shiver of
fear through you. You might think you know nothing about
statistics, but almost every time you turn on the news or
scroll through your favorite news app, you will find statis-
tics in one form or other! Here are some examples that we
Andrea Ricordi, Italy/Moment/Getty Images
found just before the 2017 November elections:

• Northam Heads Into Virginia Governor’s Race With A Small Lead. The first
major statewide elections since President Trump was inaugurated take place on
Tuesday…And while the race’s final result by itself isn’t likely to tell us much
about the national political environment, it is likely to have a big effect on the 2018
midterms. Polls show a fairly close race, with Northam slightly favored to win [over
Ed Gillespie]. An average of the last 10 surveys give Northam a 46 percent-to-43
percent advantage. Over the past month, there has been a tightening of the race, with
Gillespie closing what had been a 6-point lead. In the individual polls, though, there
is a fairly wide spread. Northam has led by as much as 17 percentage points
(a Quinnipiac University survey) and has trailed by as much as 8 points (a Hampton
University poll).1
—www.fivethirtyeight.com
• Why Trump Has a Lock on the 2020 GOP ­Nomination. In interviews with nearly
three-dozen GOP strategists and fundraisers over the past several tumultuous weeks,
virtually everyone told me that…they expect Trump to coast to the GOP nomina-
tion in 2020…the hurdles to a 2020 primary challenge are vivid when considering
a recent Washington Post/ABC News poll that found 91% of Trump voters said
they’d vote for him again…This ABC News/Washington Post poll was conducted by
landline and cellular telephone Oct. 29-Nov. 1, 2017, in English and Spanish, among
a random national sample of 1005 adults. Results have a margin of sampling error of
3.5 points, including the design effect.2
—www.cnn.com

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
2 Introduction What Is Statistics?

Articles similar to these can be found in all forms of news media, and, just before a presi-
dential or congressional election, a new poll is reported almost every day. These articles are
very familiar to us; however, they might leave you with some unanswered questions. How
were the people in the poll selected? Will these people give the same response tomorrow?
Will they give the same response on election day? Will they even vote? Are these people
representative of all those who will vote on election day? It is the job of a statistician to ask
these questions and to find answers for them in the language of the poll.
Most Believe “Cover-Up” of JFK Assassination Facts
A majority of the public believes the assassination of President John F. Kennedy was part of a
larger conspiracy, not the act of one individual. In addition, most Americans think there was a
cover-up of facts about the 1963 shooting. Almost 50 years after JFK’s assassination, a FOX
news poll shows many Americans disagree with the government’s conclusions about the killing.
The Warren Commission found that Lee Harvey Oswald acted alone when he shot Kennedy,
but 66 percent of the public today think the assassination was “part of a larger conspiracy” while
only 25 percent think it was the “act of one individual.”
“For older Americans, the Kennedy assassination was a traumatic experience that began
a loss of confidence in government,” commented Opinion Dynamics President John Gorman.
“Younger people have grown up with movies and documentaries that have pretty much pushed
the ‘conspiracy’ line. Therefore, it isn’t surprising there is a fairly solid national consensus that
we still don’t know the truth.”
(The poll asked): “Do you think that we know all the facts about the assassination of
­President John F. Kennedy or do you think there was a cover-up?”

We Know All the Facts (%) There Was a Cover-Up (Not Sure)
All 14 74 12
Democrats 11 81 8
Republicans 18 69 13
Independents 12 71 17

—www.foxnews.com3

When you see an article like this one, do you simply read the title and the first paragraph,
or do you read further and try to understand the meaning of the numbers? How did the
authors get these numbers? Did they really interview every American with each political
affiliation? It is the job of the statistician to answer some of these questions.
Hot News: 98.68F Not Normal
After believing for more than a century that 98.6°F was the normal body temperature for
humans, researchers now say normal is not normal anymore.
For some people at some hours of the day, 99.9°F could be fine. And readings as low as
96°F turn out to be highly human.
The 98.6°F standard was derived by a German doctor in 1868. Some physicians have always
been suspicious of the good doctor’s research. His claim: 1 million readings—in an epoch
without computers.
So Mackowiak & Co. took temperature readings from 148 healthy people over a three-day
period and found that the mean temperature was 98.2°F. Only 8 percent of the readings were
98.6°F.
—The Press-Enterprise4

What questions do you have when you read this article? How did the researcher select the
148 people, and how can we be sure that the results based on these 148 people are accurate
when applied to the general population? How did the researcher arrive at the normal “high”
and “low” temperatures given in the article? How did the German doctor record 1 million
temperatures in 1868? This is another statistical problem with an application to everyday life.
Statistics is a branch of mathematics that has applications in almost every part of our
daily life. It is a new and unfamiliar language for most people, however, and, like any

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
  Descriptive and Inferential Statistics 3

new language, statistics can seem overwhelming at first glance. But once the language of
statistics is learned and understood, it provides a powerful tool for data analysis in many
different fields of application.

The Population and the Sample


In the language of statistics, one of the most basic concepts is sampling. In most statistical
problems, a specified number of measurements or data—a sample—is drawn from a much
larger body of measurements, called the population.

Sample

Population

For the body-temperature experiment, the sample is the set of body-temperature mea-
surements for the 148 healthy people chosen by the experimenter. We hope that the sample
is representative of a much larger body of measurements—the population—the body tem-
peratures of all healthy people in the world!
Which is more important to us, the sample or the population? In most cases, we are
interested primarily in the population, but identifying each member of the population may
be difficult or impossible. Imagine trying to record the body temperature of every healthy
person on earth or the presidential preference of every registered voter in the United States!
Instead, we try to describe or predict the behavior of the population on the basis of
information obtained from a representative sample from that population.
The words sample and population have two meanings for most people. For example,
you read that a Gallup poll conducted in the United States was based on a sample of
1823 people. Presumably, each person interviewed is asked a particular question, and that
person’s response represents a single measurement in the sample. Is the sample the set of
1823 people, or is it the 1823 responses that they give?
In statistics, we distinguish between the set of objects on which the measurements are
taken and the measurements themselves. To experimenters, the objects on which measure-
ments are taken are called experimental units. The sample survey statistician calls them
elements of the sample.

Descriptive and Inferential Statistics


When first presented with a set of measurements—whether a sample or a population—you
need to find a way to organize and summarize it. The branch of statistics that gives us tools
for describing sets of measurements is called descriptive statistics. You have seen descrip-
tive statistics in many forms: bar charts, pie charts, and line charts presented by a political
candidate; numerical tables in the media; or the average rainfall amounts on your favorite
weather app. Computer-generated graphics and numerical summaries are commonplace in
our everyday communication.

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
4 Introduction What Is Statistics?

Definition
Descriptive statistics are procedures used to summarize and describe the important
characteristics of a set of measurements.

If the set of measurements is the entire population, you need only to draw conclusions
based on the descriptive statistics. However, it might be too expensive or too time consum-
ing to identify each member of the population. Maybe listing the entire population would
destroy it—for example, measuring the amount of force required to cause a football helmet
crack. For these or other reasons, you may have only a sample from the population. By
looking at the sample, you want to answer questions about the population as a whole. The
branch of statistics that deals with this problem is called inferential statistics.

Definition
Inferential statistics are procedures used to make inferences (that is, draw conclusions,
make predictions, make decisions) about a population from information contained in a
sample drawn from this population.

The objective of inferential statistics is to make inferences about a population from


information contained in a sample.

Achieving the Objective of Inferential


Statistics: The Necessary Steps
How can you make inferences about a population using information contained in a sample?
The task becomes simpler if you organize the problem into a series of logical steps.
1. Specify the questions to be answered and identify the population of interest. In
the Virginia election poll, the objective is to determine who will get the most votes
on election day. So, the population of interest is the set of all votes in the Virginia
election. When you select a sample, it is important that the sample be representative
of this population, not the population of voter preferences on some day prior to
the election.
2. Decide how to select the sample. This is called the design of the experiment or the
sampling procedure. Is the sample representative of the population of interest? For
example, if a sample of registered voters is selected from the city of San Francisco,
will this sample be representative of all voters in California? Will it be the same as
a sample of “likely voters”—those who are likely to actually vote in the election?
Is the sample large enough to answer the questions posed in step 1 without wasting
time and money on additional information? A good sampling design will answer
the questions posed with minimal cost to the experimenter.
3. Select the sample and analyze the sample information. No matter how much
information the sample contains, you must use an appropriate method of analysis to
obtain it. Many of these methods, which depend on the sampling procedure in
step 2, are explained in the text.

Copyright 2020 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s).
Editorial review has deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.
Another random document with
no related content on Scribd:
terres, — me voici toute la Constitution Britannique sur le dos
comme si je vendais des bombes. Je n’y comprends rien.
— Pas plus que n’y comprend rien la Great Buchonian —
apparemment. (J’étais en train de feuilleter les lettres.) Voici le
directeur général du trafic, qui déclare absolument incompréhensible
qu’un homme… Juste Ciel ! Wilton, pour le coup, ça y est ! »
Je ris tout bas, en continuant ma lecture.
« Qu’est-ce qu’il y a encore de drôle ? demanda mon hôte.
— Il y a que vous, ou Howard en votre nom, auriez fait stopper le
train du Nord de trois heures quarante.
— A qui le dites-vous ! Ils étaient tous après moi, depuis le
conducteur de la machine.
— Mais, c’est le train de trois heures quarante — l’« Induna » —
vous avez sûrement entendu parler de l’« Induna » de la Great
Buchonian ?
— Comment diable pourrais-je reconnaître un train d’un autre ! Il
s’en amène un à peu près toutes les deux minutes.
— Fort vrai. Mais il se trouve que c’est l’« Induna », le seul,
l’unique train de toute la ligne. Il est réglé à quatre-vingt-dix
kilomètres à l’heure. Il fut inauguré vers 1860, et ne s’est jamais vu
dans l’obligation de stopper…
— Ah oui, je sais ! Depuis l’arrivée de Guillaume le Conquérant
ou depuis que le roi Charles se cacha dans la cheminée de la
locomotive. Vous ne valez pas mieux que le reste de ces insulaires.
S’il est en marche depuis ce temps-là, il est temps qu’on l’arrête une
fois de temps à autre. »
L’Américain commençait à suinter par tous les pores chez Wilton,
et ses petites mains nerveuses s’agitaient sans repos :
« Supposez que vous arrêtiez l’Empire State Express, ou le
Western Cyclone.
— Supposons que je l’aie fait. Je connais Otis Harvey — ou l’ai
connu. Je lui enverrais un télégramme, et il comprendrait que je
n’avais pas autre chose à faire. C’est précisément ce que j’ai dit à la
compagnie fossile dont il s’agit.
— Vous avez donc répondu à leurs lettres sans prendre l’avis
d’un homme de loi ?
— Naturellement.
— Oh, bon sang de bon sang ! Continuez, allons, Wilton.
— Je leur ai écrit que je serais fort heureux de voir leur président
et de lui expliquer toute l’affaire en trois mots ; mais cela n’a pas eu
l’air de les arranger. C’est à croire que leur président est quelque
chose comme un dieu. Il était trop occupé, et — mais, vous pouvez
le lire vous-même — ils demandaient des explications. Oui, le chef
de gare d’Amberley Royal — en général, il rampe devant moi —
demandait une explication, et promptement, encore. Le grand
sachem de Saint Botolph en demandait trois ou quatre, et le Tout-
Puissant Mamamouchi, qui graisse les locomotives, en demandait
une chaque jour que Dieu fait. Je leur ai dit — je leur ai dit cinquante
fois — que si j’ai arrêté leur sacro-saint express, c’est que je voulais
« l’aborder [12] ». Est-ce qu’ils croient que c’était pour lui tâter le
pouls ?
[12] Américanisme, pour dire : monter dedans.

— Vous n’avez pas dit cela ?


— Lui tâter le pouls ? Naturellement, non.
— Non. « L’aborder. »
— Qu’est-ce que vous vouliez donc que je dise ?
— Mon cher Wilton, à quoi servent Mrs. Sherborne et les Clay, et
tous ces tas de gens occupés depuis quatre ans à faire de vous un
Anglais, si, la première fois que quelque chose vous tourmente, vous
retournez à votre patois ?
— J’en ai soupé, de Mrs. Sherborne et de toute la smala.
L’Amérique est assez bonne pour moi. Qu’est-ce qu’il fallait dire ?
« S’il vous plaît » ou « tous mes remerciements », ou quoi ? »
Il n’y avait pas moyen maintenant de se tromper sur la nationalité
de l’homme. La parole, le geste et le pas qu’on lui avait
soigneusement inculqués s’en étaient allés avec le masque
d’emprunt de l’indifférence. C’était le fils incontesté du plus jeune
des peuples, ce peuple qui eut pour prédécesseurs les Peaux-
Rouges. Sa voix s’était élevée au cocorico aigu et rauque de ses
semblables, lorsqu’ils se trouvent sous l’empire de quelque
surexcitation. Ses yeux rapprochés montraient tour à tour la peur
injustifiée, l’ennui hors de raison, des éclairs de pensée subits et
sans fondement, la convoitise de l’enfant pour une vengeance
immédiate, et le douloureux égarement qu’il manifeste lorsqu’il se
cogne contre la vilaine, la méchante table. Et de l’autre côté c’était,
je le savais, la compagnie, aussi incapable que Wilton de
comprendre.
« Et dire que je pourrais acheter trois vieilles lignes de chemins
de fer comme la leur, gronda-t-il en jouant avec un couteau à papier,
et en allant et venant de façon inquiète.
— Vous ne leur avez pas dit cela, j’espère ! »
Il ne répondit pas ; mais, en parcourant les lettres, je m’aperçus
que Wilton devait leur avoir dit nombre de choses surprenantes. La
Great Buchonian avait commencé par demander des explications
sur l’arrêt de son Induna, et trouvé une certaine dose de légèreté
dans les explications offertes. Elle avait alors avisé « Mr W.
Sargent » de vouloir bien mettre son avoué en rapport avec celui de
la compagnie, ou quelque phrase juridique dans ce goût-là.
« Et vous ne l’avez pas fait ? demandai-je, en levant les yeux.
— Non. Ils me traitaient absolument comme un bambin qui joue
sur la voie. Il n’y avait nul besoin d’un avoué. Cinq minutes de
conversation, et tout eût été bâclé. »
Je revins à la correspondance. La Great Buchonian regrettait
qu’en raison de leurs occupations aucun de ses directeurs ne pût
accepter l’invitation de Mr. W. Sargent de venir discuter le cas entre
deux trains. La Great Buchonian avait soin de faire remarquer que
sa façon d’agir ne cachait nulle mauvaise intention, et que ce n’était
point une question d’argent. Elle avait pour devoir de protéger les
intérêts de sa ligne, et ces intérêts n’étaient point protégés si on
laissait s’établir un précédent suivant lequel il devenait loisible à un
sujet quelconque de la reine d’arrêter un train en pleine marche. De
son côté (et il s’agissait là d’une autre branche de correspondance,
cinq directeurs de services différents au moins se trouvant impliqués
dans l’affaire), la Compagnie admettait qu’il y avait peut-être doute
fondé quant aux devoirs des rapides en certains cas exceptionnels,
et que la question pourrait être réglée devant les tribunaux jusqu’à
ce qu’intervînt, si nécessaire, un décret définitif de la Chambre des
Lords.
« Cela m’a cassé bras et jambes, dit Wilton, qui lisait par-dessus
mon épaule. Je savais bien que je finirais par buter à la Constitution
Britannique. La Chambre des Lords — Grand Dieu ! Et, à tout
prendre, je ne suis pas un des sujets de la reine.
— Mais j’avais dans l’idée que vous vous étiez fait naturaliser. »
Wilton rougit fortement et expliqua que la Constitution
Britannique en verrait bien d’autres avant qu’il retirât ses papiers.
« Quel effet tout cela vous fait-il ? demanda Wilton. La Great
Buchonian ne vous paraît-elle pas quelque peu gâteuse ?
— J’ignore. Vous avez fait quelque chose que jamais avant vous
personne n’a songé à faire, et la compagnie ne sait que penser. Je
vois qu’on propose d’envoyer l’avoué ainsi qu’un autre officier
ministériel de la compagnie pour discuter officieusement l’affaire.
Puis, il y a une autre lettre vous conseillant d’élever un mur de
quatorze pieds, couronné de verre de bouteille, au fond du jardin.
— Vous parlez d’insolence britannique ! Celui qui recommande
cela (c’est encore un de ces gros enflés de fonctionnaire) déclare
que je tirerai grand agrément du fait de regarder le mur monter de
jour en jour ! Avez-vous jamais vu fiel semblable ? Je leur ai offert en
argent de quoi acheter toute une nouvelle collection de voitures et
faire une pension au conducteur pendant trois générations ; mais
cela ne semble pas être ce qu’ils demandent. Ils s’attendent sans
doute à me voir aller à la Chambre des Lords faire intervenir un
règlement, et entre temps construire des murs. Sont-ils tous
devenus fous à lier ? On dirait que je fais profession d’arrêter les
trains. Comment aurais-je été fichu de reconnaître leur vieil Induna
de n’importe quel train omnibus ? J’ai pris le premier qui s’en venait,
et j’ai été pour cela déjà mis en prison et condamné à l’amende.
— C’était pour avoir voulu flanquer une tripotée au contrôleur.
— Il n’avait pas le droit de me tirer par derrière quand j’étais déjà
passé à moitié d’une fenêtre.
— Qu’avez-vous l’intention de faire ?
— Leur avoué et l’autre officier ministériel (est-ce qu’ils ne
peuvent pas se fier à leurs hommes sans les atteler en paire !)
arrivent ici ce soir. Je leur ai déclaré que j’étais pris, en règle
générale, jusqu’après dîner, mais qu’ils pouvaient envoyer toute la
direction si cela pouvait les consoler. »
Or, les visites après dîner, pour question d’affaires ou de plaisir,
font partie des habitudes de la petite ville en Amérique, et non de
celles de l’Angleterre, où la fin de la journée est sacrée pour le plus
petit propriétaire. En vérité, Wilton Sargent avait hissé là le drapeau
étoilé de la rébellion !
« N’est-il pas temps pour vous, Wilton, de vous apercevoir du
comique de la situation ? demandai-je.
— Qu’est-ce que vous voyez de comique à harceler un citoyen
américain, rien que parce qu’il se trouve être multimillionnaire — le
pauvre diable ! (Il resta silencieux un moment, et reprit :) Cela va
sans dire. Maintenant, je comprends ! (Il fit un tour sur lui-même et
se campa devant moi d’un air furieux :) Cela saute aux yeux. Ces
lascars-là sont en train de poser leurs jalons pour m’écorcher.
— Ils déclarent d’une façon explicite qu’ils ne recherchent pas
l’argent !
— Tout cela, c’est de la blague. C’est comme leur façon de
mettre W. Sargent, avec un simple W sur leurs adresses. Ils savent
bien que je suis le fils du vieux. Comment n’ai-je pas déjà pensé à
cela ?
— Un instant. S’il vous prenait la fantaisie de grimper sur le dôme
de Saint-Paul, et d’offrir une récompense à tout Anglais capable de
dire qui était et ce que faisait Wilton Sargent, votre père, il n’y en
aurait pas vingt dans tout Londres pour oser le prétendre.
— C’est leur provincialisme d’insulaires, alors. Je ne m’en fais
pas de bile pour un sou. Le vieux aurait mis la Great Buchonian à la
côte avant son petit déjeuner, rien que pour en prendre une épave et
déboucher sa pipe. Pardieu, je vais le faire, et sans plus tarder ! Je
vais leur apprendre que ce n’est pas une raison de faire les fendants
avec un étranger, parce qu’on a arrêté un de leurs petits joujoux de
trains, et — voilà quatre ans que je dépense ici cinquante mille livres
sterling au moins par an. »
Je m’applaudissais de ne pas être son homme de loi. Je relus la
correspondance, particulièrement la lettre qui lui recommandait —
presque avec compassion, je crus m’en apercevoir — de construire
un mur de brique de quatorze pieds au fond de son jardin ; et à
moitié route de ma lecture, une idée me frappa, qui me remplit de
douce gaîté.
Le valet de pied introduisit deux personnages, en redingote,
pantalonnés de gris, rasés de frais, à la parole et à la démarche
posées. Il était presque neuf heures, mais on eût dit qu’ils sortaient
du bain. Je me demandai pourquoi le plus âgé, qui était en même
temps le plus grand, me lança comme un regard d’intelligence, ni
pourquoi il me serra la main avec une chaleur qui n’avait rien
d’anglais.
« Voici qui simplifie la situation », dit-il tout bas.
Et, comme j’ouvrais de grands yeux, il murmura à son
compagnon :
« Je crains d’être bien peu utile pour le moment. Peut-être Mr.
Folsom ferait-il mieux de causer de l’affaire avec Mr. Sargent.
— C’est pourquoi je suis ici », repartit Wilton.
L’homme de loi eut un aimable sourire, et déclara qu’il ne voyait
pas de raison pour que toutes difficultés ne se trouvassent aplanies
en deux minutes de conversation. Il avait pris, en s’asseyant en face
de Wilton, un air on ne peut plus gracieux. Son compère me fit
remonter la scène. Le mystère s’approfondissait, mais je suivis
docilement, et entendis Wilton dire en riant d’un air gêné :
« Cette affaire est cause pour moi d’insomnie, Mr. Folsom.
Finissons-en d’une façon ou d’une autre, pour l’amour de Dieu ! »
« Ah ! A-t-il beaucoup souffert de l’insomnie ces derniers temps ?
me demanda mon compagnon, à moi, avec une petite toux
préliminaire.
— Je ne saurais dire, répliquai-je.
— Alors, je suppose que c’est tout récemment que vous êtes
entré en fonctions ici ?
— Je suis arrivé ce soir. J’avoue que je n’ai ici aucune fonction
particulière.
— Je comprends. Rien que pour observer le cours des
événements — en cas. »
Il hocha la tête.
« Justement. »
L’observation, après tout, est mon métier.
Il eut de nouveau un léger accès de toux, et en vint alors à
l’affaire.
— Mais, — je ne vous demande cela que comme simple
renseignement, — vous apercevez-vous que les hallucinations
soient constantes ?
— Quelles hallucinations !
— Elles ne sont pas toujours les mêmes, alors. Celle-ci offre un
caractère particulièrement curieux en ce sens que… Mais dois-je
comprendre que le type d’hallucination n’est pas toujours le même ?
Par exemple, Mr. Sargent croit qu’il peut acheter la Great Buchonian.
— Vous a-t-il écrit cela ?
— Il en a fait l’offre à la Compagnie — sur une demi-feuille de
papier à lettre. Or, ne serait-il pas allé par hasard à l’autre extrême,
et se croirait-il en danger de devenir indigent ? L’économie étrange
qu’il a entendu faire en se servant d’une demi-feuille de papier
montre que quelque idée de la sorte a pu lui traverser l’esprit ; et les
deux hallucinations peuvent coexister, mais c’est un cas assez rare.
Comme vous devez le savoir, l’hallucination de la grande fortune —
la folie des grandeurs, comme l’appellent, je crois, nos amis les
Français — en règle générale est constante, à l’exclusion de toutes
autres. »
Sur quoi j’entendis Wilton se servir de sa meilleure intonation
anglaise, à l’autre extrémité du cabinet :
« Mon cher monsieur, je l’ai déjà expliqué vingt fois, je voulais
avoir ce scarabée avant dîner. Supposez que vous ayez oublié de la
même façon une pièce importante de procédure ? »
« Ce trait de ruse est très significatif », murmura celui qu’il me
faut appeler mon confrère, puisqu’il y tenait.
« Je suis fort heureux, cela va sans dire, de faire votre
connaissance ; mais vous eussiez seulement envoyé votre président
dîner ici que j’aurais pu terminer l’affaire en une demi-minute. Mieux,
j’aurais pu lui acheter la Buchonian dans le temps que vos clercs
m’envoyaient ces paperasses. »
Wilton laissa tomber lourdement sa main sur la correspondance
bleu et blanc, et l’homme de loi sursauta.
« Mais, à parler franchement, répliqua ce dernier, il est, si je peux
employer expression, tout à fait inconcevable, s’agirait-il des pièces
de procédure les plus importantes, que quiconque arrête le rapide
de trois heures quarante — l’Induna — notre Induna, mon cher
monsieur. »
« Absolument ! » répéta en écho mon compagnon.
Puis, baissant la voix, il ajouta pour moi :
« Vous remarquez, encore une fois, l’hallucination constante de
la grande fortune. C’est moi qu’on appela, lorsqu’il nous écrivit cela.
Vous comprenez l’impossibilité absolue pour la Compagnie de
continuer de faire passer ses trains à travers la propriété d’un
homme qui peut à tout moment se croire nanti de la mission divine
d’arrêter le trafic. Si seulement il nous avait adressés à son homme
de loi — mais, naturellement, c’est chose qu’il n’eût pas faite, étant
données les circonstances. Une calamité — une vraie calamité ! A
son âge ! En passant, il est curieux, ne trouvez-vous pas, de
remarquer dans la voix de ceux qui se trouvent affligés de ce mal cet
accent d’absolue conviction — j’ose dire que cela vous fend le cœur
— et l’impuissance à suivre une idée. »
« Je ne vois pas bien ce que vous demandez, était en train de
dire Wilton à l’homme de loi.
— Il n’a pas besoin d’avoir plus de quatorze pieds de haut —
construction qui n’a rien de déplaisant, et il serait possible de faire
pousser des poiriers contre le côté exposé au soleil. (L’homme de loi
parlait sur un ton de voix plutôt paternel.) Rien n’est plus agréable
que de voir, comme on dit, pousser sa vigne et son figuier.
Considérez le profit et l’amusement que vous en tireriez. Si, en ce
qui vous concerne, vous trouviez le moyen de faire cela, nous
pourrions, de notre côté, arranger tous les détails avec votre homme
de loi, et il serait possible que la Compagnie entrât pour une part
dans les frais. Voilà, je crois, l’affaire réduite à sa plus simple
expression. Si vous consentez, mon cher monsieur, à prendre en
considération ce projet de mur, et que vous vouliez bien nous donner
les noms de vos hommes de loi, j’ose vous assurer que vous
n’entendrez plus parler de la Great Buchonian.
— Mais pourquoi défigurer ma pelouse par un mur de brique
neuf ?
— Le silex gris est extrêmement pittoresque.
— De silex gris, alors, si vous le voulez ainsi. Pourquoi diable
faut-il que je m’en aille construire des tours de Babel rien que parce
que j’ai « stoppé » l’un de vos trains — une seule et unique fois ? »
« L’expression dont il s’est servi dans sa troisième lettre était qu’il
voulait « l’aborder », me dit à l’oreille mon compagnon. C’était fort
curieux — une hallucination empruntée à la vie maritime, qui vient se
heurter, pourrait-on dire, à une hallucination empruntée à la vie de la
terre ferme. Dans quel monde merveilleux il doit évoluer — et
évoluera avant la fin. Si jeune cependant — si jeune ! »
« Eh bien, voulez-vous que je vous le dise en bon anglais, le
diable m’emporte si je vais me mettre à gâcher du plâtre pour vous
faire plaisir. Vous pouvez mettre sur pied tous vos bataillons, aller
devant la chambre des Lords et en ressortir, et obtenir vos
règlements au mètre courant, si cela vous amuse, dit Wilton, en
s’échauffant. Mais, grand Dieu, mon cher monsieur, je ne l’ai fait
qu’une seule et unique fois !
— Nous n’avons jusqu’ici aucune garantie que vous ne
recommencerez pas ; et, avec notre trafic, il nous faut, quand ce ne
serait que pour nos voyageurs, exiger une garantie quelconque. Il
importe que votre cas ne puisse servir de précédent. Tout cela eût
été évité si vous nous eussiez adressés à votre représentant légal. »
L’homme de loi en appela autour de lui d’un regard circulaire. Un
point, et c’était tout.
« Wilton, demandai-je, voulez-vous, maintenant, me laisser dire
mon mot ?
— Tout ce que vous voulez, répondit Wilton. Il paraît que je ne
sais pas parler anglais. En tout cas, pas de mur. »
Il se renversa dans son fauteuil.
« Messieurs, dis-je lentement, car j’observai que le médecin
aurait de la peine à revenir sur son opinion, Mr. Sargent possède de
très gros intérêts dans les principaux réseaux de chemins de fer de
son pays.
— Son pays ? repartit l’homme de loi.
— A cet âge ? repartit le médecin.
— Certainement. Il en a hérité de son père, Mr. Sargent, un
Américain.
— Et fier de l’être, dit Wilton, comme s’il fût un sénateur de San
Francisco lâché sur l’Europe pour la première fois.
— Mon cher monsieur, dit l’homme de loi, en faisant mine de se
lever, pourquoi n’avoir pas donné à la compagnie connaissance de
ce fait — de ce fait vital — dès le début de notre correspondance ?
Nous eussions compris. Nous eussions fait la part des choses.
— Au diable votre part des choses ! Est-ce que vous me prenez
pour un Peau-Rouge ou pour un aliéné ? »
Les deux hommes baissèrent la tête.
« Si l’ami de Mr. Sargent eût commencé par nous faire part de
cela, dit le médecin, d’un ton fort sévère, les choses n’eussent pas
pris cette tournure. »
Hélas ! je m’étais fait de ce médecin un ennemi pour la vie.
« Vous ne m’en avez pas fourni l’occasion, repartis-je.
Maintenant, il est clair, vous comprenez, qu’un homme qui possède
plusieurs milliers de milles de voie ferrée, comme Mr. Sargent, est
disposé à traiter les chemins de fer plus cavalièrement que le
commun des mortels.
— Naturellement, naturellement. Monsieur est Américain ; tout
s’explique. Toutefois, il s’agissait de l’Induna ; mais je comprends fort
bien que les habitudes de nos cousins d’outre-mer diffèrent des
nôtres en ces matières. Dites-moi, arrêtez-vous toujours les trains
comme cela, aux États-Unis, Mr. Sargent ?
— Je le ferais si l’occasion s’en présentait ; mais je n’ai jamais
encore eu à le faire. Allez-vous prendre cela comme base de
complications internationales ?
— Vous n’avez plus à vous soucier en rien de l’affaire. Nous
voyons qu’il n’y a guère de probabilités pour que votre action
établisse un précédent, seule chose dont nous avions peur.
Maintenant que vous comprenez que nous ne pouvons faire concilier
notre méthode avec ce genre d’arrêts inattendus, nous sommes
persuadés que…
— Je ne crois pas rester maintenant ici assez longtemps pour
arrêter un autre train, dit Wilton, l’air rêveur.
— Vous retournez donc à nos cousins de l’autre côté de — ah —
la mare aux harengs, comme vous l’appelez ?
— Non, monsieur. L’Océan — l’Océan Atlantique du Nord. Il a
trois mille milles de large, et trois milles de profondeur par endroits.
Je voudrais qu’il en eût dix mille.
— Ce n’est pas que je sois moi-même fort épris des voyages en
mer ; mais je crois que c’est le devoir de tout Anglais d’étudier une
fois dans sa vie la grande branche de notre race anglo-saxonne, de
l’autre côté de l’océan, dit l’homme de loi.
— Si jamais vous venez chez nous, et que l’idée vous prenne
d’arrêter un train de mon réseau, je vous — je veillerai à ce que vous
vous en tiriez, dit Wilton.
— Merci — ah, merci. Vous êtes trop aimable. Je suis sûr que
j’aurais beaucoup de plaisir… »
« Nous avons négligé ce fait, murmura le médecin à mon oreille,
que votre ami a proposé d’acheter la Great Buchonian.
— Il est riche de quelque chose comme vingt ou trente millions
de dollars — quatre ou cinq millions de livres, répondis-je, sachant
qu’il serait inutile d’expliquer.
— Vraiment ! C’est une fortune énorme, mais la Great Buchonian
n’est pas à vendre.
— Peut-être ne tient-il plus à l’acheter, maintenant.
— Ce serait de toutes façons impossible, dit le médecin. »
« Voilà qui est bien typique ! murmura l’homme de loi, lequel
passait mentalement en revue toute l’affaire. Je me suis toujours
aperçu, d’après ce que j’en ai lu, que vos compatriotes étaient gens
pressés. Ainsi, vous auriez fait quarante milles pour aller à Londres
et autant pour en revenir — avant dîner — tout cela pour aller
chercher un scarabée ? Voilà qui est bien américain ! Mais vous
parlez tout à fait comme un Anglais, Mr. Sargent.
— C’est un défaut auquel on peut remédier. Il est une seule
question que je voudrais vous poser. Vous avez déclaré
inconcevable que quelqu’un arrête un train sur votre réseau ?
— Et c’est la vérité — absolument inconcevable.
— Quelqu’un de sain d’esprit, n’est-ce pas ?
— C’est ce que j’ai voulu dire, naturellement. C’est-à-dire, à
moins qu’il ne s’agisse d’un A…
— Merci. »
Les deux hommes s’en allèrent. Wilton, sur le point de bourrer
une pipe, s’arrêta, prit à la place un de mes cigares, et resta
silencieux un quart d’heure.
Puis il dit :
« Auriez-vous un horaire des départs de Southampton sur
vous ? »

Loin des tours de pierre grise, des cèdres ténébreux, des


chemins de gravier impeccables, et des pelouses épinard de Holt
Hangars, coule un fleuve appelé l’Hudson, sur les bords mal peignés
duquel se pressent les palais de ces gens dont l’opulence défie les
rêves de l’avarice. C’est là, où d’une rive à l’autre la sirène du
remorqueur répond au cri déchirant de la locomotive, que vous
trouverez, avec une installation complète de lumière électrique,
d’habitacles nickelés, et l’adjonction d’une… calliope à son sifflet à
vapeur, le Columbia, yacht également à vapeur, de douze cents
tonneaux, allant sur la mer, amarré à son embarcadère privé, prêt à
mener à son bureau, à la vitesse moyenne de dix-sept nœuds, — et
gare aux chalands — Wilton Sargent, Américain.
LA TOMBE DE SES ANCÊTRES

Certaines gens vous diront que si, dans toute l’Inde, il n’était
qu’une miche de pain, elle se verrait, en égales portions, partagée
entre les Plowdens, les Trevors, les Beadons, et les Rivett-Carnacs.
Ce qui revient à dire que certaines familles servent l’Inde de
génération en génération comme les dauphins se suivent en file à
travers les mers.
Prenons un cas aussi petit qu’obscur. Il y a toujours eu pour le
moins un représentant des Chinns du Devonshire dans l’Inde
Centrale ou ses environs depuis le temps du lieutenant-artificier
Humphrey Chinn, du régiment européen de Bombay, lequel assista
à la prise de Seringapatam, en 1799. Alfred Ellis Chinn, le frère
cadet de Humphrey, commanda un régiment de grenadiers de
Bombay, de 1804 à 1813, époque à laquelle il assista à pas mal de
démêlés ; et en 1834 John Chinn, de la même famille — nous
l’appellerons John Chinn Premier — se distingua comme
administrateur de sang-froid en temps de trouble, en un lieu appelé
Mundesour. Il mourut jeune, mais laissa sa griffe sur le pays
nouveau, et l’honorable comité des Directeurs de l’honorable East
India Company [13] , après avoir personnifié ses vertus en une
décision pompeuse, paya les frais de sa tombe dans les monts des
Satpuras.
[13] Compagnie semblable à notre Compagnie des
Indes.
Il eut pour successeur son fils, Lionel Chinn, lequel, à peine sorti
de la vieille petite demeure du Devonshire, se trouva grièvement
blessé au cours de l’Insurrection. Ce dernier Chinn passa son temps
d’activité dans un rayon de cent cinquante milles autour de la tombe
de John Chinn, et parvint au commandement d’un régiment de petits
montagnards sauvages, dont la plupart avaient connu son père. Son
fils John naquit dans le petit cantonnement aux toits de chaume, aux
murs de boue, qui se trouve encore aujourd’hui situé à quatre-vingts
milles de la station de chemin de fer la plus rapprochée, au cœur
d’un pays tout en broussailles et en tigres. Le colonel Lionel Chinn
servit trente ans et se retira. A la traversée du canal de Suez, son
paquebot croisa le transport à destination de l’étranger, qui emportait
son fils en Orient pour l’acquit de ses devoirs de famille.
Les Chinns ont plus de chance que le reste des humains en ce
qu’ils savent exactement ce qu’ils ont à faire. Un Chinn intelligent
passe son examen pour le service civil de Bombay, et s’en va dans
l’Inde Centrale, où tout le monde est enchanté de le voir. Un Chinn
moyen entre dans le service de la Police ou dans les Eaux et Forêts,
et, tôt ou tard, lui aussi fait son apparition dans l’Inde Centrale ; c’est
ce qui a donné naissance au dicton : « l’Inde Centrale est habitée
par les Bhils, les Mairs [14] et les Chinns, tous gens du même
acabit. » La race est à petits os, brune et silencieuse, et les plus
bornés d’entre eux sont encore de bons fusils. John Chinn II était
plutôt intelligent, mais en sa qualité de fils aîné il entra dans l’armée,
suivant une autre tradition des Chinns. Son devoir était de demeurer
dans le régiment de son père pour toute la durée de sa vie, bien que
le corps fût de ceux que bien des gens eussent payé cher pour
éviter. Il s’agissait d’irréguliers, de petits hommes bruns, noirauds,
vêtus de vert olive à garniture de cuir noir, et leurs amis les
appelaient les « Wuddars », nom qui s’applique à une race de gens
de basse caste, laquelle déterre les rats pour les manger. Mais les
Wuddars n’en tiraient nul grief. C’étaient les Wuddars, et leurs chefs
d’orgueil se résumaient à ceux-ci :
[14] Bhils et Mairs, antiques races de l’Inde Centrale.
Premièrement, ils possédaient moins d’officiers anglais que
n’importe quel régiment indigène. Secondement, à la parade, leurs
lieutenants n’étaient pas montés, comme il est de règle générale,
mais marchaient sur leurs deux jambes, à la tête de leurs hommes.
Or, il faut à l’homme qui peut se maintenir au niveau des Wuddars à
leur pas de marche bon souffle et bon jarret. Troisièmement,
c’étaient les plus pukka shikarries (épatants chasseurs) de toute
l’Inde. Quatrièmement — et indéfiniment — c’étaient les Wuddars —
connus jadis sous le nom de Chinn’s Irregular Bhil Levies, mais
maintenant, désormais et pour toujours, les Wuddars.
Nul Anglais n’était de leur mess, qui ne le fût pour l’amour d’eux
ou par tradition de famille. Les officiers employaient, pour parler à
leurs soldats, une langue qu’il n’y avait pas deux cents hommes
blancs dans l’Inde pour comprendre ; et les hommes étaient leurs
enfants, tous tirés des Bhils, lesquels sont peut-être la race la plus
étrange qui soit parmi les nombreuses races étranges de l’Inde. Ce
furent, et ils le sont restés dans le cœur, de véritables sauvages,
sournois, méfiants, pleins de superstitions à ne pas croire. Les races
que dans le pays nous appelons indigènes trouvèrent le Bhil en
possession de la terre la première fois qu’ils firent irruption dans
cette partie du monde, il y a des milliers d’années. Les livres les
appellent Préariens, Aborigènes, Dravidiens, que sais-je encore ; et,
en fin de compte, c’est comme cela que les Bhils s’appellent eux-
mêmes. Lorsqu’un chef rajpoute, dont les bardes peuvent chanter la
généalogie en remontant à douze siècles en arrière, se trouve porté
au trône, son investiture n’est point complète qu’on ne l’ait marqué
au front du sang tiré des veines d’un Bhil. Les Rajpoutes prétendent
que la cérémonie ne signifie rien du tout, mais le Bhil n’ignore pas
que c’est la dernière, dernière ombre de ses droits de jadis comme
antique possesseur du sol.
Des siècles d’oppression et de massacre ont fait du Bhil un larron
cruel et à moitié détraqué, ainsi qu’un voleur de bétail ; et, à l’arrivée
de l’Anglais, il semblait presque aussi ouvert à la civilisation que les
tigres de ses jungles. Mais John Chinn Premier, père de Lionel,
grand-père de notre John, pénétra dans son pays, vécut avec lui,
apprit sa langue, tua les daims qui ravageaient ses pauvres récoltes,
et gagna sa confiance, de sorte que quelques Bhils apprirent à
labourer et semer, tandis que d’autres se laissaient enrôler au
service de la Compagnie pour policer leurs frères.
Lorsqu’ils comprirent que se mettre à l’alignement ne voulait pas
dire « être exécuté sur l’heure », ils acceptèrent le métier de soldat
comme un genre fatigant, mais amusant, de sport, et apportèrent du
zèle à tenir en obéissance les petits Bhils restés sauvages. C’était le
tranchant du coin. John Chinn Premier leur donna par écrit la
promesse que si, à partir d’une certaine date, ils se conduisaient
bien, le gouvernement fermerait les yeux sur les fautes passées ; et
comme John Chinn était connu pour n’avoir jamais manqué à sa
parole — une fois il avait promis de faire pendre un Bhil considéré
chez lui comme invulnérable, et le fit pendre par-devant sa tribu pour
sept meurtres avérés — les Bhils se rangèrent aussi tranquillement
qu’il leur était donné de faire. Ce fut un travail lent, invisible, pareil à
celui qu’on pratique par toute l’Inde aujourd’hui ; et bien que l’unique
récompense de John Chinn arriva, comme je l’ai dit, sous la forme
d’une tombe aux frais du gouvernement, le petit peuple des
montagnes jamais ne l’oublia.
Le colonel Lionel Chinn les connaissait et les aimait, lui aussi, et
avant la fin de son service, ils se trouvaient devenus, pour des Bhils,
gens fort civilisés. C’est à peine si l’on pouvait distinguer maints
d’entre eux des fermiers hindous de basse caste ; mais dans le sud,
où John Chinn Premier était enterré, les plus sauvages s’en tenaient
encore aux sommets des monts des Satpuras, entretenant une
légende suivant laquelle un jour Jan Chinn, comme ils l’appelaient,
reviendrait aux siens. En attendant ils se méfiaient de l’homme blanc
et de ses façons. Au moindre motif d’excitation, les voilà qui
partaient pillant à l’aventure, et tuant de temps à autre ; mais se
trouvaient-ils ensuite l’objet d’un maniement discret, qu’ils se
désolaient comme des enfants, et promettaient de ne plus jamais,
jamais recommencer.
Les Bhils du régiment — les porteurs d’uniforme — donnaient
l’exemple de maintes vertus, mais avaient besoin qu’on se prêtât à
leurs volontés. Tant qu’on ne les emmenait pas comme rabatteurs
après les tigres, ils s’ennuyaient et éprouvaient une sorte de
nostalgie ; et leur froide audace — tous les Wuddars chassent le
tigre à pied, c’est la marque de leur caste — faisait l’étonnement des
officiers eux-mêmes. Ils vous suivaient de près un tigre blessé avec
autant d’insouciance que s’il se fût agi d’un moineau à l’aile brisée ;
et cela, à travers un pays tout en cavernes, crevasses et cavités, où
un fauve pouvait tenir une douzaine d’hommes à sa merci. De temps
à autre, on apportait à la caserne quelque petit homme la tête
fracassée ou les côtes arrachées ; mais ses compagnons n’en
apprenaient pas pour si peu la prudence ; ils se contentaient de
régler au tigre son compte.

Le siège de derrière d’une charrette à deux roues versa le jeune


John Chinn, ses étuis à fusil cascadant tout à l’entour de sa
personne, devant la verandah du mess isolé des Wuddars. Le svelte
petit gaillard, au nez en bec de corbin, semblait aussi délaissé
qu’une chèvre égarée lorsqu’il secoua la poussière qui lui
blanchissait les genoux, et que la charrette s’éloigna en cahotant sur
la route éblouissante. Mais au fond du cœur il se sentait satisfait.
Après tout, c’était le lieu où il était né, et les choses n’avaient guère
changé depuis qu’enfant on l’avait envoyé en Angleterre, il y avait
quinze ans.
Quelques constructions de plus ; mais l’air, l’odeur et le soleil
étaient les mêmes ; et les petits hommes en vert, qui traversaient le
terrain d’exercice, lui étaient on ne peut plus familiers. Trois
semaines plus tôt, John Chinn eût déclaré qu’il ne se rappelait pas
un mot de la langue Bhil, et voici qu’à la porte du mess il s’aperçut
que ses lèvres balbutiaient des phrases qu’il ne comprenait pas —
des bouts de vieux contes de nourrice, et les abrégés de tels et tels
ordres que son père donnait aux hommes.
Le colonel le regarda monter les marches, et se prit à rire.
« Voyez donc ! dit-il au major. Pas besoin de demander à quelle
famille appartient le petiot. C’est un pukka (vrai) Chinn. On dirait son
père entre 1850 et 1860.
— Espérons qu’il aura le tir aussi juste, repartit le major. Il a
apporté assez de ferblanterie avec lui.
— Ce ne serait pas un Chinn, s’il ne l’avait pas. Regardez-le se
moucher. Le piton des Chinns ! Il agite son mouchoir comme son
père. C’en est la seconde édition — mot pour mot.
— Un conte de fées, ma parole ! fit le major, en regardant à
travers les lattes des jalousies. Si c’est l’héritier légitime de son père,
il… Tenez, le vieux Chinn ne pouvait pas plus passer devant ce store
sans le tripoter, que…
— Son fils ! repartit le colonel, en sursautant.
— Ah, par exemple ! » s’écria le major.
Le regard du jeune homme s’était arrêté sur un store de roseau
fendu qui pendait de travers entre les piliers de la verandah, et
machinalement il en avait saisi le bord pour le redresser. Le vieux
Chinn avait maudit trois fois par jour ce store pendant nombre
d’années ; il ne pouvait jamais l’avoir à sa satisfaction. Son fils
pénétra dans le vestibule au milieu d’un silence ébahi. On lui fit bon
accueil, d’abord à cause de son père, et, après examen, à cause de
lui-même. Il ressemblait d’une façon ridicule au portrait du colonel,
là, sur le mur, et lorsqu’il se fut un peu rincé la gorge, il gagna ses
quartiers de ce pas de jungle, court et silencieux, propre à l’aîné.
« Parlez-moi d’hérédité, dit le major. C’est là le résultat de trois
générations parmi les Bhils.
— Et les hommes le savent, dit un capitaine. Ils sont restés, la
langue pendante, à attendre ce garçon-là. A moins qu’il ne leur
donne littéralement des coups de bâton, voilà des gens qui vont se
coucher dans la poussière par compagnies pour l’adorer.
— Rien comme d’avoir eu un père avant vous, repartit le major.
Je suis un parvenu avec mes garnements. Il n’y a que vingt ans que
je suis dans le régiment, et mon vénéré père n’était qu’un simple
bourgeois. Il ne faut pas chercher à approfondir la pensée d’un Bhil.
Mais pourquoi ce bijou de serviteur, que le jeune Chinn a amené
avec lui, se sauve-t-il à travers la campagne avec son paquet ? »
Il s’avança dans la verandah, et cria après l’homme — un de ces
serviteurs typiques de sous-lieutenant nouveau venu, qui parlent
anglais et vous mettent dedans en proportion.
« Qu’est-ce qu’il y a ? cria-t-il.
— Plein de mauvaises gens, ici. Moi m’en aller, moussu, fut-il
répondu. Eux avoir pris les clefs du Sahib et déclarer qu’ils vont tirer.
— Pas mal — c’est bien cela. Comme ils savent jouer du pied,
ces brigands de par là-haut. C’est quelqu’un qui l’aura fait mourir de
peur. »
Le major s’en alla en flânant regagner ses quartiers afin de
s’habiller pour le repas du mess.
Le jeune Chinn, marchant comme un homme en rêve, avait fait le
tour du cantonnement tout entier avant de gagner son minuscule
cottage. Les quartiers du capitaine, dans lesquels il était né, le
retinrent quelques moments ; ensuite, il regarda sur le terrain
d’exercices le puits au bord duquel il s’asseyait le soir avec sa
bonne, et l’église de dix pieds de large sur quatorze de long où les
officiers se rendaient au service si quelque chapelain d’une religion
officielle quelconque se trouvait passer par là. Tout cela paraissait
fort petit, comparé aux gigantesques constructions sur lesquelles il
avait coutume de lever le nez, mais c’était bien le même lieu.
De temps à autre il croisait un groupe de soldats silencieux qui
saluaient. Tout aussi bien eût-on pu les prendre pour ces hommes
qui l’avaient promené sur leur dos lorsqu’il étrennait sa première
culotte. Une faible lueur brillait dans sa chambre ; au moment où il y
entra, des mains lui étreignirent les pieds, et un murmure partit du
sol.
« Qui est-ce ? demanda le jeune Chinn, sans savoir qu’il parlait
en langue bhil.
— Je vous ai porté dans mes bras, Sahib, alors que j’étais un
homme fort, et vous, un petit — qui pleurait, pleurait, pleurait ! Je

You might also like