Professional Documents
Culture Documents
ebook download Introduction to Probability and Statistics - Metric Version, 15e 15th Edition William Mendenhall - eBook PDF all chapter
ebook download Introduction to Probability and Statistics - Metric Version, 15e 15th Edition William Mendenhall - eBook PDF all chapter
https://ebooksecure.com/download/introduction-to-probability-and-
statistics-metric-version-ebook-pdf/
https://ebooksecure.com/download/introduction-to-probability-and-
statistics-metric-edition-ebook-pdf/
http://ebooksecure.com/product/ebook-pdf-introduction-to-
probability-and-statistics-3rd-by-william-mendenhall/
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/
http://ebooksecure.com/product/introduction-to-probability-and-
statistics-14th-edition-ebook-pdf/
http://ebooksecure.com/product/ebook-pdf-introduction-to-
probability-and-statistics-4rd-canadian-edition/
http://ebooksecure.com/product/ebook-pdf-introduction-to-
probability-and-statistics-for-engineers-and-scientists-5th-
edition/
https://ebooksecure.com/download/understanding-basic-statistics-
ebook-pdf-2/
■ Table 4 Critical Values of t, pages 692–963
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
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
Represented in Canada by
Nelson Education, Ltd.
tel: (416) 752 9100 / (800) 668 0671
www.nelson.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.
Brief Contents
4 Probability 126
7 Sampling Distributions 245
8 Large-Sample Estimation 288
15 Nonparametric Statistics 633
Appendix i 681
data sources 714
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
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
Chapter Review 86
Technology Today 87
Reviewing What You’ve Learned 91
Case Study: The Boys of Summer 95
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
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
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
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
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
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
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
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
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
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.
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.
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
Study Aids z5
x 2 m 211.5 2 250
s
5
15
5 2 2.57
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.
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.
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?
• 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.
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.
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.
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.
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.