Intuitive Introductory Statistics 1st

Edition Douglas A. Wolfe

Visit to download the full and correct content document:
https://textbookfull.com/product/intuitive-introductory-statistics-1st-edition-douglas-a-w
olfe/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Introductory Business Statistics (Fall 2020 Corrected

Edition) Alexander Holmes

https://textbookfull.com/product/introductory-business-
statistics-fall-2020-corrected-edition-alexander-holmes/

Introductory Statistics Using SPSS Herschel (Hersch) E.

(Edmond) Knapp

https://textbookfull.com/product/introductory-statistics-using-
spss-herschel-hersch-e-edmond-knapp/

IBM SPSS for Introductory Statistics Use and

Interpretation Sixth Edition Morgan

https://textbookfull.com/product/ibm-spss-for-introductory-
statistics-use-and-interpretation-sixth-edition-morgan/

The Bayesian way: introductory statistics for

economists and engineers First Edition Nyberg

https://textbookfull.com/product/the-bayesian-way-introductory-
statistics-for-economists-and-engineers-first-edition-nyberg/
Working the Lode Going for the Gold 1 1st Edition Layla
Wolfe Karen Mercury Wolfe Layla

https://textbookfull.com/product/working-the-lode-going-for-the-
gold-1-1st-edition-layla-wolfe-karen-mercury-wolfe-layla/

Answering questions with data: Introductory Statistics

for Psychology Students Matthew J. C. Crump

https://textbookfull.com/product/answering-questions-with-data-
introductory-statistics-for-psychology-students-matthew-j-c-
crump/

Adolescent Dating Violence Theory Research and

Prevention 1st Edition David A Wolfe

https://textbookfull.com/product/adolescent-dating-violence-
theory-research-and-prevention-1st-edition-david-a-wolfe/

The Glass House Bettina Wolfe

https://textbookfull.com/product/the-glass-house-bettina-wolfe/

A Companion to the War Film 1st Edition Douglas A.

Cunningham

https://textbookfull.com/product/a-companion-to-the-war-film-1st-
edition-douglas-a-cunningham/
Springer Texts in Statistics

Douglas A. Wolfe
Grant Schneider

Intuitive
Introductory
Statistics
Springer Texts in Statistics

Series editors
R. DeVeaux
S.E. Fienberg
I. Olkin
Springer Texts in Statistics (STS) includes advanced textbooks from 3rd- to 4th-year under-
graduate courses to 1st- to 2nd-year graduate courses. Exercise sets should be included. The
series editors are currently Stephen Fienberg and Richard D. De Veaux. George Casella
and Ingram Olkin were editors of the series for many years.

More information about this series at http://www.springer.com/series/417

Douglas A. Wolfe • Grant Schneider

Intuitive Introductory
Statistics
Douglas A. Wolfe Grant Schneider
Department of Statistics Upstart Network
The Ohio State University San Carlos, CA, USA
Columbus, OH, USA

ISSN 1431-875X ISSN 2197-4136 (eBook)

Springer Texts in Statistics
ISBN 978-3-319-56070-0 ISBN 978-3-319-56072-4 (eBook)
DOI 10.1007/978-3-319-56072-4

Library of Congress Control Number: 2017950163

# Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter
developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed
to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional
affiliations.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To our wives, Marilyn and Jingjing, for their patience and support
through the lengthy preparation of this text
Preface

Understanding and interpreting data, both experimental data and the vast
quantity of observational data that is now routinely available in the public
domain, need not be linked solely to comprehension of a prescribed set of
mathematical expressions. Introducing students to the beauty of statistics can
be both motivational and interesting without being encumbered by equations.
In fact, it is our view that the goal of an introductory statistics course should
be to foster an appreciation for the role of statistics and associated data
analysis approaches in our everyday lives rather than to prepare the students
to be statistical analysts—if necessary, there will be time enough for that
secondary emphasis as they begin to concentrate on their chosen fields of
study. An introductory course should foster appreciation for the relevance
and importance of using statistical methods for summarizing and interpreting
data, but not at the expense of enjoying the process. While it is, of course,
necessary to include common statistics such as the sample mean, standard
deviation, correlation, etc. in an introductory statistics course, they do little to
motivate and “grab” the students and can be introduced after a student has
come to appreciate the nature of the basic information that is provided in
collections of data. This desire leads us in this book to emphasize counting

vii
viii Preface

and ranking approaches as initial tools for eliciting information from data
collections rather than as a fallback only when sample means and standard
deviations are not effective. Not only can students easily understand such
counting and ranking techniques, but they also find them to be quite valuable
as they explain their analyses to others.
The second point of emphasis in this text is that examples should not be
chosen simply to illustrate how a statistical procedure can be applied. Dull,
contrived examples can make even interesting statistical analyses uninterest-
ing. On the other hand, an example to which students easily relate can go a
long way in both piquing and sustaining their interest in the associated
statistical analysis. While they learn a statistical technique, they also unearth
some information of interest to them in its own right. We all learn best when
we see the relevance of a topic in our own lives. We have worked hard to
motivate the statistical discussions in our text through experimental settings
and data collections that we ourselves find interesting and that raise questions
that we can use statistical tools to address. While we are the first to confess
that our view of the world may not match up completely with that of high
school juniors and seniors or college freshmen and sophomores, we have tried
to include data sets in both the examples and the exercises that are both real
(not just realistic) and of general relevance for the students.
A third new feature for this text is the way in which we have chosen to
present the chapter exercises. In addition to the usual set of exercises at the
end of each section (and there are many), we have included a substantial
number of comprehensive exercises at the end of each chapter that include
conceptual exercises, data analysis/computational exercises, exercises involv-
ing hands-on student activities, and exercises associated with a student’s use
of the Internet to access interesting and relevant data sets and statistical
analyses. This provides instructors with a wide variety of exercises to chal-
lenge the individual interests of students in their classes.
Preface ix

The fourth and final unique feature of this introductory text is the inclu-
sion of the necessary R functions to enable instructors and students to ana-
lyze data sets without the tedium often accompanying many statistical
computations. Examples throughout the text still provide all of the details of
the associated statistical calculations so that students are fully aware of how
the various statistics elicit information from a data set(s). However, we also
provide the appropriate R procedures that can be used to make the same
calculations. It is easy enough for instructors to bypass the R procedures if
they choose, but including them as part of the basic course would permit their
students to apply the associated statistical procedures to large data sets where
the direct numerical calculations by hand would be prohibitive. The use of
these R programs also eliminates the need to include the normal, t, and χ 2
tables, as well as the relevant nonparametric null distribution tables as part of
the text. We have also organized all of the R programs used in the text into a
documented collection that is formally registered as an R package IIS specifi-
cally linked to this text.
The text is specifically designed for use in an AP statistics course for high
school juniors or seniors or in a one-semester or two-quarter introductory
precalculus statistics course for college freshmen or sophomores. How well
we have succeeded in reaching these groups of students will clearly deter-
mine the impact of our not-just-one-more introductory statistics textbook!
Many friends and colleagues have helped with both the initial develop-
ment and improvement of this text over the years. We owe a particular debt of
gratitude to Brad Hartlaub for his invaluable help in initiating this project in
the first place and his dedicated effort to move it forward over a number of
succeeding years. We also appreciated his input from teaching much of this
material in statistics classes at Kenyon College. Similarly, we owe thanks to
Deborah Rumsey and Elizabeth Stasny for their feedback from using early
drafts of the project in introductory statistics courses at The Ohio State
x Preface

University. We also owe thanks to Jungwon Byun, Ben Chang, Neha Hebbar,
Cindy Smith, and all of the attendees of the so-called “data party” at Upstart
for providing a fertile testing ground for the material and concepts. Finally,
we owe a special thank you to the computer support group in the Department
of Statistics at The Ohio State University for their patience in helping us work
through the various versions of MSWord that were confronted over the many
years in preparation of the text.
Our editors Michael Penn and Hannah Bracken, who originally signed us
to publish with Springer, and Rebekah McClure, who assumed our project
when Hannah returned to school, were dedicated from the start of the project
and provided tremendous support to see it through to publication. Our
production manager Christina Oliver skillfully guided the manuscript
through the production process.
To everyone who helped over the many years, our heartfelt thanks.

Columbus, OH, USA Douglas A. Wolfe

San Carlos, CA, USA Grant Schneider
Contents

1 Exploratory Data Analysis: Observing Patterns

and Departures from Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Interpreting Graphical Displays of Data Collections . . . . . . . 2
1.1.1 Construction of a Histogram . . . . . . . . . . . . . . . . . . . . 17
1.2 Numerically Summarizing One-Variable
Data Collections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
1.2.1 Effects of a Linear Transformation . . . . . . . . . . . . . . . 78
1.3 Comparing One-Variable Data Collections . . . . . . . . . . . . . . . 95
Chapter 1 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 124
1.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
1.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 128
1.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
1.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

2 Exploring Bivariate and Categorical Data . . . . . . . . . . . . . . . . . . . . 143

2.1 Exploring the Relationship Between Two
Quantitative Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

xi
xii Contents

2.1.1 Common Types of Relationships – No

Association, Positive Association,
Negative Association . . . . . . . . . . . . . . . . . . . . . . . . . . 144
2.1.2 Scatterplot Smoothing . . . . . . . . . . . . . . . . . . . . . . . . . . 152
2.1.3 Including a Third Variable on Scatterplots . . . . . . . . 161
2.2 Measuring the Strength of Association . . . . . . . . . . . . . . . . . . 169
2.2.1 Properties of r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
2.2.2 An Alternative Measure of Association . . . . . . . . . . . 178
2.3 Exploring the Relationship between Two Categorical
Variables (Frequency Tables) . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
Chapter 2 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 190
2.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
2.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 192
2.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
2.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

3 Designing a Survey or Experiment: Deciding What

and How to Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
3.1 Methods of Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
3.2 Planning and Conducting Surveys or Polls . . . . . . . . . . . . . . . 218
3.3 Planning and Conducting Experiments . . . . . . . . . . . . . . . . . . 227
Chapter 3 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 237
3.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
3.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 238
3.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240
3.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241

4 Understanding Random Events: Producing Models

Using Probability and Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
4.1 Probability as Relative Frequency: Law
of Large Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Contents xiii

4.2 Some Basic Probability Rules . . . . . . . . . . . . . . . . . . . . . . . . . . 253

4.2.1 Addition Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
4.2.2 Conditional Probability . . . . . . . . . . . . . . . . . . . . . . . . 257
4.2.3 Multiplication Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
4.3 Discrete Random Variables and Their Probability
Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
4.3.1 Binomial Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . 268
4.3.2 Geometric Distribution . . . . . . . . . . . . . . . . . . . . . . . . . 272
4.4 Simulating Probability Distributions . . . . . . . . . . . . . . . . . . . . 279
4.5 Expected Values and Standard Deviations
for Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
4.6 Combining Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . 291
4.7 Normal Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
4.7.1 Probability Calculations for Normal
Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299
4.7.2 Using Normal Distributions as Models
for Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Chapter 4 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 315
4.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
4.B. Data Analysis/ Computational . . . . . . . . . . . . . . . . . . . 317
4.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
4.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328

5 Sampling Distributions and Approximations . . . . . . . . . . . . . . . . . 331

5.1 The Sampling Distribution for a Sample Average . . . . . . . . . 333
5.1.1 Comparing Two Averages . . . . . . . . . . . . . . . . . . . . . . 339
5.2 Sampling Distributions for Proportions and Counts . . . . . . . 346
5.2.1 Comparing Two Proportions . . . . . . . . . . . . . . . . . . . . 350
xiv Contents

5.2.2 Comparing Several Proportions . . . . . . . . . . . . . . . . . 352

5.2.3 Using Ranks and Counts to Compare
Two Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
5.3 Approximating Sampling Distributions . . . . . . . . . . . . . . . . . 366
5.4 Simulating Sampling Distributions . . . . . . . . . . . . . . . . . . . . . 380
Chapter 5 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 397
5.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
5.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 400
5.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 413
5.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414

6 Statistical Inference: Estimating Probabilities

and Testing and Confirming Models . . . . . . . . . . . . . . . . . . . . . . . . 417
6.1 Point Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 420
6.2 Interval Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
6.3 Hypothesis Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466
Chapter 6 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 510
6.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
6.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 517
6.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
6.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533

7 Statistical Inference for the Center of a Population . . . . . . . . . . . . 537

7.1 Exact Inference for the Center of a Population
under a Minimal Assumption . . . . . . . . . . . . . . . . . . . . . . . . . . 539
7.2 Exact Inference for the Center of a Continuous
Population Under the Assumption of Population
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
7.3 Inference for the Center of a Normal Distribution–
Procedures Associated with the Sample Mean
and Sample Standard Deviation . . . . . . . . . . . . . . . . . . . . . . . . 574
Contents xv

7.4 Discussion of Methods of Inference for the Center

of a Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
7.5 Approximate Inference for the Center of a Population
when the Number of Sample Observations is Large . . . . . . . 604
7.6 Approximate Inference for the Median of an Arbitrary
Distribution – Bootstrapping the Sample Median . . . . . . . . . 616
Chapter 7 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 621
7.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621
7.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 623
7.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628
7.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630

8 Statistical Inference for Matched Pairs or Paired

Replicates Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
8.1 Inference for Continuous Paired Replicates
or Matched Pairs Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636
8.2 Inference for Qualitative Differences—Data
from Paired Replicates or Matched Pairs
Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
Chapter 8 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 655
8.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 655
8.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 657
8.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666
8.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 667

9 Statistical Inference for Two Populations–Independent

Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
9.1 Approximate Inference for the Difference
in Proportions for Two Populations . . . . . . . . . . . . . . . . . . . . . 671
xvi Contents

9.2 Inference for the Difference in Medians

for Any Two Continuous Populations . . . . . . . . . . . . . . . . . . . 687
9.3 Approximate Inference for the Difference in Means
for Two Populations–Procedures Based
on the Two Sample Averages and Sample
Standard Deviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708
9.4 Inference for the Difference in Means for Two Normal
Populations with Equal Variances--Procedures Based
on the Two Sample Averages and a Pooled
Sample Standard Deviation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 734
9.5 Discussion of the Methods of Inference for the
Difference Between the Centers of Two Populations
with Independent Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748
Chapter 9 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 749
9.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
9.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . . 750
9.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766
9.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769

10 Statistical Inference for Two-Way Tables

of Count Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773
10.1 General Test for Differences in Population
Proportions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776
10.2 Test for Association (Independence) between
Two Categorical Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787
10.3 Exact Procedure for Testing Equality
of Two Population Proportions . . . . . . . . . . . . . . . . . . . . . . . . 801
10.4 Goodness-of-fit Test for Probabilities in a
Multinomial Distribution with I > 2 Categories . . . . . . . . . . 808
Contents xvii

Chapter 10 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 817

10.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817
10.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . 820
10.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832
10.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834

11 Statistical Inference for Bivariate Populations . . . . . . . . . . . . . . . . 839

11.1 Correlation Procedures for Bivariate Normal
Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840
11.2 Rank-Based Correlation Procedures . . . . . . . . . . . . . . . . . . . . . 851
11.3 Fitting a Least Squares Line to Bivariate Data . . . . . . . . . . . . 860
11.4 Linear Regression Inference for Normal Populations . . . . . . 867
11.5 Rank-Based Linear Regression Inference . . . . . . . . . . . . . . . . 874
Chapter 11 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 883
11.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883
11.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . 889
11.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901
11.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 902

12 Statistical Inference for More Than Two Populations . . . . . . . . . 907

12.1 One-way Rank-Based General Alternatives
ANOVA for More Than Two Populations . . . . . . . . . . . . . . . 909
12.2 One-way General Alternatives ANOVA for More
Than Two Normal Populations . . . . . . . . . . . . . . . . . . . . . . . . 916
12.3 One-way Rank-Based Ordered Alternatives
ANOVA for More Than Two Populations . . . . . . . . . . . . . . . 926
Chapter 12 Comprehensive Exercises . . . . . . . . . . . . . . . . . . . . . . . . . 935
12.A. Conceptual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935
12.B. Data Analysis/Computational . . . . . . . . . . . . . . . . . . . 936
12.C. Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 942
12.D. Internet Archives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944
xviii Contents

Appendix A: Listing of Datasets Usage Locations

Throughout IIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947

Appendix B: Listing of R Functions Usage Locations

Throughout IIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 955

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 967
 Exploratory Data Analysis: Observing
Patterns and Departures from Patterns 1

Data, data, everywhere and we are forced to look at it. We see it in the
newspapers (“63% of the people polled support the president’s decision
to. . . ”), on the news (“scientists at a major research university report that
treatment of Parkinson’s disease with a combination of drugs A and B has the
potential to extend remission of the disease by an average of 2 years. . .”), from
the government (“the nation’s trade deficit narrowed last month, for the first
time in. . . ”), in sports (“over the past two years left-handed hitters have a
batting average of .359 against . . . ”), in finance (“the Dow Jones Industrial
Average rose again today to a new record high, marking the seventh consec-
utive day of record highs, but the broader market. . .”), and social settings
(“despite the robust economy, the percentage of families living below the
poverty line has not dropped substantially over the past six months. . .”), just
to scratch the surface.
We are an information-oriented society–we demand facts in all aspects of
our lives, but we are often overwhelmed by the magnitude of the response.

# Springer International Publishing AG 2017 1

D.A. Wolfe, G. Schneider, Intuitive Introductory Statistics, Springer Texts in Statistics,
DOI 10.1007/978-3-319-56072-4_1
2 1 Exploratory Data Analysis: Observing Patterns and Departures from Patterns

However, it is vitally important that we are able to properly interpret the data
we encounter, as well as to understand the circumstances behind the collec-
tion of the numbers. For example, what is it that is causing such a large
increase in sales for a particular item produced by a major snack-food com-
pany? Which states are the safest to drive in and what issues are directly
related to that safety? What percentage of college students are involved in
“binge drinking” and are there consequences? What factors directly affect the
growth of pine seedlings? What is a safe dosage for a new medication to treat
Alzheimer’s disease?
This chapter is about visualizing, summarizing, and interpreting existing
collections of data. In later chapters we turn to the other set of issues regard-
ing data collection and analysis themselves, including what, where, when,
how, and from whom.

1.1 Interpreting Graphical Displays of Data Collections

Data are data are data–we all know what they are–it is as simple as that–
right? Actually, nothing could be further from the truth. The word data itself
is plural, referring to the fact that it corresponds to more than a single piece of
information. The American Heritage College Dictionary (1993) defines the
word data as “Factual information, esp. information organized for analysis.”
or “Values derived from scientific experiments.” While one or the other of
these definitions adequately describes the typical collection of data we might
encounter, neither definition helps us understand or interpret a specific set
of data.
The discipline of statistics offers an organized set of principles and
procedures for addressing these many facets of data. One way to visualize
the role of statistics in the collection and interpretation of data is to think of the
interaction between data and statistics as an interlocking jigsaw puzzle –
sometimes straightforward, sometimes quite intricate, in their connections.
1.1 Interpreting Graphical Displays of Data Collections 3

In order to properly obtain and interpret a collection of data, it is important

that we have taken the necessary steps to produce a fully complete puzzle. As
we have all experienced, visualizing only portions of a jigsaw puzzle can
produce misleading glimpses of the overall picture depicted in the puzzle.
Thus, it is vital that we understand all of the various data-statistics
connections so that every one of the puzzle pieces is in its proper place. This
is the only way to ensure that a study or investigation leads to a correct
representation and interpretation of the problem of interest. Figure 1.1
provides a visualization of this data-statistics jigsaw puzzle.
The types and varieties of data are as numerous as the studies that lead to
their collection. As a result, the very first step in any study or investigation is
to clearly understand what type and how much data should be obtained in
order to address the question of interest. Once this decision has been reached,
the next natural consideration is how to properly collect the appropriate data.
These data-statistics connections relating to the design of an experiment and
collection of the data are depicted as red pieces in the Fig. 1.1 jigsaw puzzle.
Such issues will be discussed further in Chap. 3.
The second set of considerations in the data-statistics puzzle address the
question of how to summarize and describe a data collection obtained from
a properly designed study. How can the data be clustered into useful inter-
pretive categories? What are the appropriate ways to summarize, describe,
and compare the data? These types of data-statistics connections that relate to
summarizing and describing a data collection are depicted as yellow pieces in
Fig. 1.1. Issues related to the major data types and proper methods for
summarizing and describing each type are the topics of Chaps. 1 and 2.
The remaining connections in the data-statistics puzzle deal with appro-
priate methods for analyzing the various types of data and reaching proper
conclusions from these analyses. The data-statistics puzzle pieces associated
with this aspect of data analysis and interpretation are colored blue in Fig. 1.1
and are discussed in detail in Chaps. 6, 7, 8, 9, 10, 11, and 12.
4 1 Exploratory Data Analysis: Observing Patterns and Departures from Patterns

Fig. 1.1 Data-statistics jigsaw puzzle: an accurate picture only when we

have made all of the proper data-statistics connections

What visual
displays can How much data
What should
be used to do we need to
we collect?
effectively collect?
present the data?

Do the data Are these data from an

How should we
contain unusual observational
summarize
observations study or an
the data
or outliers? experiment?
collected?

Were there any Is the normal Can you identify

problems in the model the distinguishing
data collection appropriate for features of these data?
process? these data?

What inferences, How would you

if any, are explain, in plain
you hoping to How will the
English, the overall patterns
make with data be used?
or trends in these
these data? data to a friend?

We collect data because we are interested in the characteristics of some

group or groups of people, places, things, or events. For example, we might
want to know about temperatures in the month of June in Los Angeles, the
intended functions of some clay artifacts excavated at an archaeological site,
or the numbers of eggs for ruby-throated hummingbirds. In these contexts,
temperature, intended function, and number of eggs are called variables.
1.1 Interpreting Graphical Displays of Data Collections 5

Definition 1.1 A variable is an attribute of interest in a study. It can

represent the outcomes of a designed experiment, the responses from a
survey, or characteristics of a set of people, places, or things.

Variables assume different values across the group or groups we are

studying. Thus, June temperatures in Los Angeles are different from day to
day and from year to year, and different hummingbirds lay different numbers
of eggs. To understand how a variable varies across a group, it is important to
obtain a ‘representative collection’ of variable values from the group.

Definition 1.2 A particular measurement of an attribute is called an

observation. The goal of a scientific study is to collect enough represen-
tative observations to tell us what we want to know about an attribute.

Each observation is a value of the variable of interest. For example, an

observation at an archaeological site might be: “the clay artifact found in this
location has a ceremonial function”. On the other hand, an ornithologist
might record: “this particular nest contains 3 eggs”. Data sets encountered
on a daily basis are nothing more than collections of observations on
attributes of interest to the collectors of the data–but what types of variables
make up these data collections? The first division that we make is between
quantitative and categorical variables.

Definition 1.3 A categorical variable is one for which the associated

observations are simply listings of physical characteristics or traits of the
subjects or objects being studied. For example, eye color is a categorical
variable, with categories brown, blue, green, etc. Categorical data then
correspond to observed sample counts in each of the possible categories.
6 1 Exploratory Data Analysis: Observing Patterns and Departures from Patterns

Definition 1.3 (continued)

In the case of eye color, the data would be the number of sample subjects
with brown eyes, the number with blue eyes, the number with green
eyes, etc.

Categorical variables abound in the way we describe people or objects to

others. The classifications tall, young, poor, Hispanic, college graduate, ath-
lete, male, and HIV positive are but a few of the traits that might be important
in studying a human population.
For tabulation purposes, arbitrary numerical values are sometimes
assigned to the classes of possible characteristics or traits for a categorical
variable. However, arithmetic manipulations with such numerical labels are
not meaningful. For example, although we might arbitrarily assign the num-
ber 1 to brown eyes, the number 2 to blue eyes, the number 3 to green eyes,
etc., arithmetic computations with these labels would not make sense; that is,
the sum of 4 for one brown-eyed and one green-eyed subject would not have
the same interpretation as the sum of 4 for two blue-eyed subjects.

Definition 1.4 A quantitative variable is one for which the associated

observations will be numerical and, therefore, such that the usual arith-
metic manipulations make sense. Quantitative data then correspond to
the measured numerical values of a quantitative variable.

We are very familiar with many quantitative variables as well, including

weight, height, time between failures of an electronic device, blood pressure,
number of eggs laid by a hummingbird, and percentage of tornadoes in a
given year which caused at least one death. It makes sense to do arithmetic,
such as adding and dividing by the number of observations to find an average
value, on all of these variables.
1.1 Interpreting Graphical Displays of Data Collections 7

In addition to variables that are clearly either categorical or quantitative, it

is not uncommon to encounter ‘borderline’ variables. For example, consider
an experiment in which each participant is asked to rank a set of five dish
detergents, from most preferred to least preferred. For each participant the
realized data (known as ordinal data) will be a ranking from 1 (best) to
5 (worst) for the five competing detergents. While such data clearly provide
more than simple categorical information (e. g., the detergent labeled as 1 by a
participant is actually preferred to and not just different from a detergent that
he labels 3), they are not quite quantitative, as we have described it in
Definition 1.4. One of the missing features is that numerical calculations are
not valid on these ordinal rank data. Thus, for example, a particular dish
detergent might receive two rankings of 1 and two rankings of 5 from four
participants in the study, while a second detergent might receive all four
rankings of 3 from the same participants. Even though both detergents
received an average ranking of 3, the four participants clearly do not view
the two detergents as equivalent. The averaging process is simply not a valid
operation for such ranking data. While our major emphasis in this text will be
on categorical and numerical data, we will at times also discuss how to deal
with ordinal data.

Graphical Displays of Categorical Data The main purpose of some studies is

to see how a set of data is distributed across a small set of categories or classes. If
each observation falls into exactly one of the classes, we say that the classes
partition the data collection. For example, the classes urban, suburban, and
rural partition new housing construction, and the categories cats, dogs, and
“other” partition domestic animals. The classes or categories in a partition are
exhaustive and exclusive, meaning that they include every possible observation
and they do not overlap, respectively.
As you learned in Definition 1.3, observations that can be placed into
categories are called categorical data. Gender, hair color, and make of
8 1 Exploratory Data Analysis: Observing Patterns and Departures from Patterns

automobile are examples of categorical data. For such data we count how
many observations fall into each of the categories. This collection of category
frequency counts describes how the data collection is distributed across the
partition.
Computation of the relative frequencies, corresponding to the frequency
counts divided by the number of items in the sample, for each of the
categories provides an adequate numerical summary measure for most
collections of categorical data. However, there are two graphical techniques,
bar graphs and pie charts that can be used to provide visual summarization of
categorical data. Bar graphs can be used to display either the actual frequency
counts or the relative frequencies (sample percentages) in a sample. Pie charts,
on the other hand, are designed solely to visualize relative frequencies (sam-
ple percentages). Both techniques can be particularly effective when we are
interested in comparing a number of different collections of data using the
same partitioning categories. Examples 1.1 and 1.2 show how to construct bar
graphs and relative-frequency pie charts for categorical data.

Example 1.1. Archaeological Excavations at Naco Valley Ed Schortman and

Pat Urban, Professors in the Department of Anthropology and Sociology at
Kenyon College in Gambier, Ohio, periodically organize and supervise
expeditions for undergraduate students from a number of institutions to
study the history of various regions in Central America. One such study
involved the excavation of individual structures at Site 128 in the northwest
portion of Naco Valley in Honduras. During the period AD 600–1000 (known
as the Late Classic period) this site was the administrative home to a cadre of
bureaucratic functionaries (lesser nobility) and their supporters. Among other
things, the expedition group was interested in the types of artifacts that were
being used by these inhabitants in order to find out if there were differences in
such usage across different structures within Site 128 (possibly occupied by
persons of differing levels of power or wealth, etc.). In one such archaeological
1.1 Interpreting Graphical Displays of Data Collections 9

study, the scientists and their students counted the numbers of the following
types of objects found at a number of different structures within a particular
excavation site:

1. Ocarinas (multi-note clay musical instruments)

2. Figurines (fired clay anthropomorphic and zoomorphic effigies)
3. Incensarios (incense burners)
4. Ground stone tools (used to process grain, primarily corn, into flour)
5. Stamps (for cloth decoration)
6. Sherd disks (possibly spindle whorls for processing thread)
7. Candelaros (for both practical lighting and ceremonial activities)
8. Jewelry (personal adornments made of clay).

The observed frequency counts for two of the structures at this site are
presented in Table 1.1.
We are interested in the artifacts used by the inhabitants of these structures
during the period of their existence, as partitioned into the eight noted artifact
categories or classes. To construct bar charts for these categorical data, we
label the horizontal baseline with the numbers as assigned to the artifact
categories in Table 1.1. (Note that these labeling numbers are purely arbitrary,
as they could have been assigned in any order to the eight categories.) Then
we construct rectangles with common widths above each of the labels, with
the height of a particular rectangle corresponding to the observed frequency
count for the associated category. The bar graphs for Structures 13 and 17 are
presented in Fig. 1.2a and b.
Several things are visually clear from the bar graphs. First, Categories
2 (figurines) and 3 (incensarios) dominate the types of artifacts found at both
of the structures, perhaps indicating that each of them might have played
some sort of ceremonial role for the inhabitants of the site. However, it is also
clear from the bar graphs that the relative importance of Categories 2 and
3 are reversed at the two structures, suggesting that the associated ceremonial
10 1 Exploratory Data Analysis: Observing Patterns and Departures from Patterns

Table 1.1 Artifacts recovered at two different structures during the exca-
vation of Site 128 in Naco Valley, Honduras, 1996
Frequency count Frequency count
Artifact at structure #13 at structure #17
1. Ocarinas 14 21
2. Figurines 56 53
3. Incensarios 73 35
4. Ground stone tools 21 12
5. Candelaros 10 12
6. Stamps 5 9
7. Sherd disks 5 0
8. Jewelry 4 7
Source: Schortman and Urban (1998).

Fig. 1.2 Bar graphs for the Naco Valley artifact recovery data. (a) Structure
13. (b) Structure 17

roles may have differed to some extent. (We will leave that discussion for the
archaeologists!) Finally, there appears to be the possibility of differences in the
use of stamps, sherd disks and jewelry at the two structures as well.
A second form of visual representation for categorical data is provided by
the pie chart. It is most commonly used when we wish to pictorially display
the sample relative frequencies, or percentages, rather than the raw
frequencies for the various classes of the partition. It is particularly effective
for displaying differences between two populations with respect to the same
categories.
1.1 Interpreting Graphical Displays of Data Collections 11

Example 1.2. Motor-Vehicle Deaths by Type of Accident In their annual

report on accident statistics, the National Safety Council (1996) reported the
breakdown of motor-vehicle deaths by type of accident for the period of years
from 1913 through 1995. The population of deaths by motor-vehicle accident
is partitioned into the following nine exhaustive categories:

1. Pedestrians
2. Other motor vehicles
3. Railroad trains
4. Streetcars
5. Pedalcycles
6. Animal-drawn vehicle or animal
7. Fixed objects
8. Noncollision accidents
9. Nontraffic deaths

Do you feel there would be a noticeable change in the relative percentages

for these nine categories from, say, 1949 to 1985? One way to visually help us
answer this question is to create pie charts for the data from those years. The
dataset traffic_accidents contains the motor-vehicle accident data for the inclu-
sive period 1924–1995. Accessing this file, we find the counts for the years
1949 and 1985 displayed in Table 1.2.
Summing, we find that the total number of reported motor-vehicle deaths
in 1949 and 1985 are 32,536 and 47,019, respectively. To create relative
frequencies (or percentages) for the nine categories for each of the years, we
divide the observed frequencies by the total numbers of reported motor-
vehicle deaths in the respective years. For example, the relative frequency
(or percentage) for pedestrian deaths in 1949 is 8800/32,536 ¼ .2705, or
slightly greater than 27%. These relative frequencies are then depicted visu-
ally in the two pie charts displayed in Fig. 1.3, where each relative frequency is
Another random document with
no related content on Scribd:
 "And the—good-night."

"G'night."

CHAPTER XVIII

"But how absurd," said Mary, "when we've only just met."

"Oh, but come; it was about two years ago that you let me call you
'Mary.'"

"True," said Mary thoughtfully.

"And you can't say we aren't suited to one another. We both play
without the chalk line."

"So we do. Yet ... Oh, I can't say all at once. Give me a little time."

"I'll give you three of Jane's hoops. That's about six months."

CHAPTER XX

"Twenty-eight," said the North Rutland champion. "That's what I won

the championship by, I remember."

"It's a good winning score," I said. "Do they play much in North
Rutland?"

"I'm afraid it's been very slow for you and Miss Middleton," said Jane.

"Not exactly slow," I said.

"We've been talking a lot of nonsense," explained Miss Middleton.

 "Not exactly nonsense," said I.

"Oh, it was," said Miss Middleton, "you know it was."

"I suppose it was," I sighed. "Well, we'll try again to-morrow."

"Right," said the champion. "But I shall use my other mallet."

GARDENING

There may be gardeners who can appear to be busy all the year round—
doing even in the winter their little bit under glass. But for myself I wait
reverently until the twenty-second of March is here. Then, spring having
officially arrived, I step out on to the lawn, and summon my head-gardener.

"James," I say, "the winter is over at last. What have we got in that big
brown-looking bed in the middle there?"

"Well, sir," he says, "we don't seem to have anything, do we, like?"

"Perhaps there's something down below that hasn't pushed through yet?"

"Maybe there is."

"I wish you knew more about it," I say angrily; "I want to bed out the
macaroni there. Have we got a spare bed, with nothing going on
underneath?"

"I don't know, sir. Shall I dig 'em up and have a look?"

"Yes, perhaps you'd better," I say.

Between ourselves, James is a man of no initiative. He has to be told

everything.
 However, mention of him brings me to my first rule for young
gardeners—

Never sow Spring Onions and New Potatoes in the

same bed

I did this by accident last year. The fact is, when the onions were given
to me I quite thought they were young daffodils; a mistake anyone might
make. Of course, I don't generally keep daffodils and potatoes together; but
James swore that the hard round things were tulip bulbs. It is perfectly
useless to pay your head-gardener half-a-crown a week if he doesn't know
the difference between potatoes and tulip bulbs. Well, anyhow, there they
were in the herbaceous border together, and they grew up side by side; the
onions getting stronger every day and the potatoes more sensitive. At last,
just when they were ripe for picking, I found that the young onions had
actually brought tears to the eyes of the potatoes—to such an extent that the
latter were too damp for baking or roasting, and had to be mashed. Now, as
everybody knows, mashed potatoes are beastly.

THE RHUBARB BORDER

gives me more trouble than all the rest of the garden. I started it a year
ago with the idea of keeping the sun off the young carnations. It acted
excellently, and the complexion of the flowers was improved tenfold. Then
one day I discovered James busily engaged in pulling up the rhubarb.

"What are you doing?" I cried. "Do you want the young carnations to go
all brown?"

"I was going to send some in to the cook," he grumbled.

 "To the cook! What do you mean? Rhubarb isn't a vegetable."

"No, it's a fruit."

I looked at James anxiously. He had a large hat on, and the sun couldn't
have got to the back of his neck.

"My dear James," I said, "I don't pay you half-a-crown a week for being
funny. Perhaps we had better make it two shillings in future."

However, he persisted in his theory that in the spring people stewed

rhubarb in tarts and ate it!

Well, I have discovered since that this is actually so. People really do
grow it in their gardens, not with the idea of keeping the sun off the young
carnations but under the impression that it is a fruit. Consequently, I have
found it necessary to adopt a firm line with my friends' rhubarb. On arriving
at any house for a visit, the first thing I say to my host is, "May I see your
rhubarb bed? I have heard such a lot about it."

"By all means," he says, feeling rather flattered, and leads the way into
the garden.

"What a glorious sunset," I say, pointing to the west.

"Isn't it," he says, turning round; and then I surreptitiously drop a pint of
weed-killer on the bed.

Next morning I get up early and paint the roots of the survivors with
iodine.

Once my host, who for some reason had got up early too, discovered
me.

"What are you doing?" he asked.

"Just painting the roots with iodine," I said, "to prevent the rhubarb
falling out."
 "To prevent what?"

"To keep the green fly away," I corrected myself. "It's the new French
intensive system."

But he was suspicious, and I had to leave two or three stalks untreated.
We had those for lunch that day. There was only one thing for a self-
respecting man to do. I obtained a large plateful of the weed and emptied
the sugar basin and the cream jug over it. Then I took a mouthful of the
pastry, gave a little start, and said, "Oh, is this rhubarb? I'm sorry, I didn't
know." Whereupon I pushed my plate away and started on the cheese.

ASPARAGUS

Asparagus wants watching very carefully. It requires to be tended like a

child. Frequently, I wake up in the middle of the night and wonder if James
has remembered to put the hot-water bottle in the asparagus bed. Whenever
I get up to look I find that he has forgotten.

He tells me to-day that he is beginning to think that the things which are
coming up now are not asparagus after all, but young hyacinths. This is
very annoying. I am inclined to fancy that James is not the man he was. For
the sake of his reputation in the past I hope he is not.

POTTING OUT

I have spent a busy morning potting out the nasturtiums. We have them
in three qualities—mild, medium and full. Nasturtiums are extremely
peppery flowers, and take offence so quickly that the utmost tact is required
to pot them successfully. In a general way all the red or reddish flowers
should be potted as soon as they are old enough to stand it, but it is
considered bad form among horticulturists to pot the white.
 James has been sowing the roses. I wanted all the pink ones in one bed,
and all the yellow ones in another, and so on; but James says you never can
tell for certain what colour a flower is going to be until it comes up. Of
course, any fool could tell then.

"You should go by the picture on the outside of the packet," I said.

"They're very misleading," said James.

"Anyhow, they must be all brothers in the same packet."

"You might have a brother with red hair," says James.

I hadn't thought of that.

GRAFTING

Grafting is when you try short approaches over the pergola in

somebody's else's garden, and break the best tulip. You mend it with a
ha'penny stamp and hope that nobody will notice; at any rate not until you
have gone away on the Monday. Of course in your own garden you never
want to graft.

I hope at some future time to be allowed—even encouraged—to refer to

such things as the most artistic way to frame cucumbers, how to stop
tomatoes blushing (the hom[oe]opathic method of putting them next to the
French beans is now discredited), and spring fashions in fox gloves. But for
the moment I have said enough. The great thing to remember in gardening
is that flowers, fruits and vegetables alike can only be cultivated with
sympathy. Special attention should be given to backward and delicate
plants. They should be encouraged to make the most of themselves. Never
forget that flowers, like ourselves, are particular about the company they
keep. If a hyacinth droops in the celery bed, put it among the pansies.

But above all, mind, a firm hand with the rhubarb.

 GOLF

CHAPTER I

The documents in the case are these:

Him to Me

"Come and play golf on Thursday. What is your handicap? I expect you
will be too good for me."

Me to Him

"MY GOOD THOMAS,—Don't be silly. I will play you at cricket,

tennis, lawn tennis, football (both codes), croquet, poker-patience, high
diving and here-we-go-round-the-mulberry-bush. If you insist, I will take
you on at prisoners' base and billiards. Moreover I can dance the pavane.
Yours ever,

"ADOLPHUS

"P.S.—Anyhow, I haven't any clubs."

Him to Me
 "MY DEAR ASS,—I gather that you aren't a golfer; well, why not
begin on Thursday? There will be nobody else playing probably. Meet me
at Victoria eleven-five. My brother is away, and I will lend you his clubs."

Me to Him

(Telegram)

"Is your brother out of England? Wire reply."

Him to Me

"Yes. Sicily."

Me to Him

"Right you are then."

CHAPTER II

"You know," I said to Thomas in the train, "I have played a little on a
very small island off the coast of Scotland, but it was such a very small
island that we never used a driver at all, or—what's that other thing called?
—a brassy. We should have been into the sea in no time. But I rather fancy
myself with a putter."

"You might go round with a putter to-day."

"I might, but I sha'n't. I expect to use the wooden clubs with great ease
and dexterity. And I think you will find that I can do some business with the
mashie. What's a niblick?"

"The thing you get out of bunkers with."

"Then I sha'n't want that."

CHAPTER III

The fateful moment arrived. Thomas presented me with a ball called the
Colonel, and a caddie offered me Thomas's brother's driver. He also asked
me what sort of tee I should like.

I leant upon my club and looked at him. Then I turned to Thomas.

"Our young friend Hector," I said, "is becoming technical. Will you
explain?"

"Well, do you want a high or a low one?"

"I want to hit this Colonel ball very hard in the direction of that flag.
What do you recommend?"

"Well, that's just as you——"

"I think a medium one. Slow to medium."

The preliminaries being arranged, I proceeded to address the ball. My

own instinct was to take the address as read and get to business as soon as
possible, but in the presence of an expert like Hector I did not dare to omit
the trimmings. As it was, after every waggle I felt less and less like hitting
the Colonel. When at last I did let fly it was with feelings of relief that I
discovered, on returning an eye to the spot, that the tee was indeed empty. I
shaded my eyes and gazed into the middle distance.

"No," said Thomas, "it's more to the right." He indicated a spot in the
foreground, about ten yards E.N.E. "There you are."
 "That isn't my ball."

"Yes, sir," said Hector, grinning.

"May I have it back?"

Thomas laughed and smote his own into the blue. "You go on from
there," he said.

"I'm still aiming at the same flag?"

"Go on, you ass."

I went on. The ball again rolled ten yards to the east.

"I don't know why we're going in this direction," I said. "If I get much
further east I shall have to send back Bartlett. You know, I don't believe the
Colonel is taking this seriously. He doesn't seem to me to be trying at all.
Has he ever been round the course before?"

"Never. He's quite new to it."

"There you are. He'll come down at the ditch for a certainty."

I played my third. A third time we went ten yards to the east—well,

perhaps with a touch of north in it again. And this time Hector gave a
sudden snort of laughter.

I leant upon my club, and stared him into gravity. Then I took Thomas
by the coat and led him on one side.

"There are, Thomas," I said, "other things than golf."

"There are," he agreed.

"A man may fail temporarily at the game and yet not be wholly
despicable."

"True."
 "He may, for instance, be able to dance the pavane with grace and
distinction."

"Quite so."

"Well then, will you take this giggling child away and explain to him
that I am not such an ass as I appear? Tell him that the intellectuals of
Brook Green think highly of my mental powers. Assure him that in many of
the best houses at Wandsworth Common I am held to be an amusing
raconteur. Remind him of my villanelle 'To Autumn.' For heaven's sake
make him understand that my reputation does not stand or fall with my
ability to use this brassy thing. I'm not a golf professional."

Thomas allowed himself to smile. "I will tell him," he said, "that you
are not a golf professional."

We veered right round to the east with my fourth and then I became
desperate.

"Why," I shouted, "do I hit the ball with a ridiculous club like this? I
could send it farther with a cricket bat. I could push it straighter with a
billiard cue. Where's that bag? I am going to have a lucky dip."

I dipped, and came up with what Thomas called a cleek. "Now then," I
said. I didn't stop to address the Colonel, I simply lashed at him. He flew
along the ground at a terrific pace.

"Well kept down," said Thomas admiringly.

"By Jove!" I cried, "that's never going to stop. See how he flies along ...
now he breasts the slope ... look, he is taking the water jump.... Ah, he has
crossed his legs, he's down."

"This," I said to Thomas as we walked after the Colonel, "is golf. A

glorious game."

"What nonsense," I said to Thomas, "they put in comic papers about

golf. All that about digging up the turf! ... and missing the ball! ... and
breaking the clubs! I mean, I simply don't see how one could! Let's see, I've
played four, haven't I?"

"Five," said Thomas. "What I am wondering," he added, "is why you

should have been afraid of using any club in your small island off the coast
of Scotland."

CHAPTER IV

Twenty Strokes After

"The green, the green," I shouted joyfully, in the manner of the ancient
Greeks, though I was only on the edge of it.

"Go on," said Thomas.

I took a careful aim and put the white down.

"You see," I said carelessly, leaning on my putter.

STUMP CRICKET

April will soon be here," said Miss Middleton, with a sigh of happiness.

"Bless it," I agreed. "My favourite month. Twelve," I added

conversationally, "is my lucky number, and Thursday the day of the week
on which I do least work. When next the twelfth of April falls on a
Thursday, which may not be for centuries, look out. Something terrific will
happen."
 "It's about now that one begins to wonder if one is in form, or likely to
be."

"Just about now," I agreed. "I always say that when the draw is
announced for the semi-finals of the English Cup, in which, of course, I
take not the slightest interest whatever, and in fact hardly know what teams
are left in for it, though I must say I hope Southampton wins this year,
because, after all, Fry did play for them once, but they'll have a bit of a job
to beat the Wolves you know—and then there's Newcastle and Fulham after
that, and of course, you can't be ..."

"I'm tired of that sentence," said Miss Middleton.

"So was I. I only wanted to make it clear that I have no use for these
spectacular gladiatorial combats. Give me cricket, the game of——"

Miss Middleton did not appear to be listening.

"Do you bowl as fast and as good a length as you talk?" she asked
thoughtfully.

"No. More swerve perhaps. And I bowl with my head a good deal."

"I see. Quite different. Well, then, will you coach me this spring? Do,
there's a dear."

"I should love to. I know all the things to say."

She got up excitedly.

"Come along then. I've got the rippingest bat. But you must promise not
to bowl too fast."

I had said that I knew all the things to say, but as a matter of fact there is
only one thing to say: "You should have come out to 'er, sir." (Or, I suppose,
in Miss Middleton's case, "You should have come out to him, madam.") It's
a silly remark to make, because it is just what one is always doing. At
school I could come out to anything that was straight and not too high; the
difficulty lay in staying in. Nobody ever told me how to do that.
 Miss Middleton led the way to a walled-in tennis lawn, which lay next
to the broccoli tops and things, and was kept away from it by only six feet
of brick. If it had simply been a question of cabbage I should have said
nothing, but there would be grapes there too.

"I know," said Miss Middleton. "But we must play against a wall. Don't
bowl too much to leg."

I hadn't bowled since October the Fourth. The first post-October ball
was a trifle over-pitched, and a little too much to the right. All the same I
was just saying, "You should have come out to that one," when there was a
crash from the direction of long-on.

"By Jove, I didn't know you were so good. Was that the grapes?"

"How awful! Yes. It simply seemed to fly off the bat. I did ask you not
to bowl there, didn't I?"

She looked so penitent that I had to comfort her.

"It's all right," I said consolingly. "I had a man there. You would have
been out all the way. Besides," I went on, "a little air will do the grapes
good. They stay all the time in one hot room, and then when they go out
into the cold, they don't muffle up, and the natural consequence is—— Or
am I thinking of influenza?"

"Never mind. We must remember not to do it again, that's all. Give me

some to cut."

There are several ways of cutting. For myself, I was taught to cut
"square" with the left leg across and "late" with the right, the consequence
being that I can do neither. W.G. (to work downwards) generally uses the
fore-arm for the stroke, Ranji the wrist. Miss Middleton keeps both feet
together and puts her whole body into it; and the direction in which the ball
travels is towards long-on. There was another crash.

"Golf is your game," I said admiringly. "You lay it dead on the

greenhouse every time."
 "I say, what shall we do? Father will be furious."

I looked at my watch.

"I can just catch the three-twenty-five," I said.

"Oh, don't be a coward, when it's all your fault for bowling so badly."

"Perhaps the glass is insured," I suggested. "It is generally."

"It's insured against hail," she said doubtfully.

I looked at the sky. It was one of the most beautiful blues I have seen.

"No," said Miss Middleton sadly.

"It will be a point for lawyers to argue, I fancy, what is actually meant
by hail. You would probably define it at once as aqueous vapour cooled
down in the atmosphere to the freezing point of water."

"I don't know. Perhaps I should."

"But 'hail' here obviously has a wider significance. I take it to mean

'anything that descends suddenly from the clouds.' I haven't 'Williams on
Real Property' with me, but——"

"Come on," said Miss Middleton, "let's say it does mean that. And could
you, please, keep them a bit more on the off?"

"It's no good my keeping them there if you don't."

The worst of coaching—I speak now as an expert—is that it is so

difficult to know what to say when a lady whirls her bat twice round her
head, gives a little shriek, gets the ball on the knee, and says, "What ought I
to have done then?" The only answer I could think of was "Not that."

"I thought you knew all about coaching," she said scornfully.
 "But, you see, it depends on what you were wanting to do," I said
meekly. "If it was a drive you should have come out to it more, and if it was
a cut you should have come down on it; while if it was a Highland fling you
lacked abandon, and if you were killing a wasp——"

"A good coach would know what was the best thing to do with that
particular ball, wouldn't he? And that's just what he would tell you."

"He wouldn't know," I said modestly. "You don't often meet that sort of
ball in good cricket."

"No, I suppose not. That's why I didn't know what to do, I expect. You
know I generally know exactly what to do, only I can't do it."

"Is that really so?" I cried excitedly. "Why, then, of course, you ought to
coach me!"

*******

We had a very jolly afternoon. I fancy I shall be in some form this year.
Miss Middleton is one of the best bowlers I have seen, but I brought off
some beautiful shots. I wanted some tea badly afterwards.

"What glorious days we have now," said Miss Middleton's mother, as

she handed me a cup.

"Glorious," said Miss Middleton's father.

"H'm, yes," I said doubtfully. "But you know I'm afraid it won't last. It's
beginning to look rather like—like hail."

"Yes," said Miss Middleton. "We both thought so."

EXPLORING
 Come, gather around, my 'earties, and listen
a while to me,
For I 'ave a yarn to spin you, a yarn of the
Polar Sea;
It's as true as I'm standing here, lads, as true as it
blows a gale,
That I was the first as nearly burst a-finding the
Great Big Nail—
As sworn to by Etukishook, Gaukrodger, J. C. Clegg,
Sir Fortescue Flannery, and the Cardinal
Merry del Val.

It was all of a parky morning that wunnerful fourth

of March,
When I put on a hextry weskit and made for the
Marble Arch;
So I sez good-bye to my country, "Lunnon," I sez,
"adoo!"
And I up and strode down the Edgware Road
athirsting to see it through,
Followed by Etukishook, Gaukrodger, J. C. Clegg,
Sir Fortescue Flannery, and the Cardinal
Merry del Val.

I 'adn't no blooming gum-drops, I 'adn't no polar

bears,
I 'adn't no sextant neither, but I thinks to myself,
"'Oo cares?"
And I waggled my watch-chain jaunty, which was
jewelled in every hole.
"I can always steer by my cumpas 'ere, it's pointing
straight to the Pole."
So it is!" said Etukishook, Gaukrodger, J. C. Clegg,
Sir Fortescue Flannery, and the Cardinal
Merry del Val.
 I walked for the 'ole of that morning, then I sez
to myself, "Old son,
This here is a dash-for-the-Pole like, and it's
darn little dash you've done."
So I enters an 'andy station, and I sez to the
man in the 'utch,
"'Ere, gimme a ticket as goes to Wick—no, a
first-return—'ow much?—
Ah, and five third singles for Etukishook, Gaukrodger,
J. C. Clegg, Sir Fortescue Flannery, and the
Cardinal Merry del Val."

We sailed from Wick to the norrard for 'undreds

of days and nights.
Till we came at last to the ice-floes and followed the
Northern Lights,
The Horroreo-boreo-balis, which it turned us all
'orrible pale,
And I sez to my men, "To-morrow and then we
shall land at the Great Big Nail."
"'Ooray!" said Etukishook, Gaukrodger, J. C. Clegg,
Sir Fortescue Flannery and the Cardinal Merry
del Val.

'Twas the cumpas as went and found it—it seemed

to have turned its head,
It would spin like mad for a minute and then it
would lay like dead;
It took on just like a wild thing, you'd almost 'a'
sworn it cried,
Till at last it shot through the glass and got right up
on its end and died.
"That proves it," cried Etukishook, Gaukrodger,
J. C. Clegg, Sir Fortescue Flannery, and the
Cardinal Merry del Val.
 We gave three cheers for ole England and we up
with the Union Jack,
And we plugged our pipes and we smoked
'em and we thought about getting back;
But a wunnerful pride so filled us as we sat on top
of the Ball,
That innocent tears (the first for years) rolled out
of the eyes of all.
Partikerlarly out of those of Etukishook, Gaukrodger,
J. C. Clegg, Sir Fortescue Flannery, and the
Cardinal Merry del Val.

Then I called for a pen and paper, and I wrote to

the King, "Dear King,
"I've found the Pole, and I'm tying a piece of it up
with string;
I'll send it round in the morning for your Majesty's
grace to see
Just drop me a wire, if you like it, sire, and I'll
collar the lot! Signed: Me.
Witnesses: Etukishook, Gaukrodger, J. C. Clegg, Sir
Fortescue Flannery and the Cardinal Merry
del Val."

So that's how it 'appened, my 'earties, no matter

what others may say.
(Did they see the Pole? They didn't! That
proves I 'ad took it away).
It's as true as I'm standing here, lads, as true as
The Daily Mail,
That I was the first as nearly burst a-finding the
Great Big Nail.
 SHOPPING

"If you should happen to be in Regent Street tomorrow at four" (ran the
assignation), "just where what's the name of the street comes into it, and a
lady in a very pretty new mauve coat and skirt bows to you, raise your hat
and say 'Crisis' and she will let you help her with her shopping."

My guess at the name of the street was successful. I raised my hat and
said "Good-afternoon."

"But you had to say 'Crisis,'" said Miss Middleton. "That's the
password."

"I can't. I've sworn I'll never say it again. I took a most fearful oath.
Several people heard me taking it, and swooned."

"But how do I know you're the right one if you don't say it? Well, I
suppose I shall have to let you come. I've just lost mother; she went in at the
silver department and out at the art fabrics—like people when they can't pay
for their hansoms."

"Yes, that's bad. 'The accused, who appeared to feel her position acutely,
gave a false address.' What are you going to buy?"

"Well, I thought I'd just help you get your presents first."

"I'm not giving any this Christmas. I gave a lot only a year ago."

"Oh, but haven't they paid you any wages since then?"

"Yes, a few trifling sums, only—— Quick, there's your mother!" I

pulled Miss Middleton hastily into the nearest shop and shut the door.

"What fun!" she said breathlessly. "Mother loves hide-and-seek."

Mrs Middleton hurried past, covered with parcels, and dived into
another door.