(Ebook) Linear Mixed Models by BRADY T.

WEST,
Kathleen B. Welch, Andrzej T Galecki ISBN
9781032019321, 1032019328 download

https://ebooknice.com/product/linear-mixed-models-51158844

Explore and download more ebooks at ebooknice.com

 We believe these products will be a great fit for you. Click
the link to download now, or visit ebooknice.com
to discover even more!

(Ebook) Linear Mixed Models A Practical Guide Using Statistical

Software by Brady West, Kathleen B. Welch, Andrzej T Galecki
ISBN 9781584884804, 1584884800

https://ebooknice.com/product/linear-mixed-models-a-practical-guide-using-
statistical-software-1209198

(Ebook) Linear Mixed Models: A Practical Guide Using Statistical

Software, Second Edition by Brady T. West, Kathleen B. Welch,
Andrzej T Galecki ISBN 9781466560994, 1466560991

https://ebooknice.com/product/linear-mixed-models-a-practical-guide-using-
statistical-software-second-edition-4739314

(Ebook) Biota Grow 2C gather 2C cook by Loucas, Jason; Viles,

James ISBN 9781459699816, 9781743365571, 9781925268492,
1459699815, 1743365578, 1925268497

https://ebooknice.com/product/biota-grow-2c-gather-2c-cook-6661374

(Ebook) Matematik 5000+ Kurs 2c Lärobok by Lena Alfredsson, Hans

Heikne, Sanna Bodemyr ISBN 9789127456600, 9127456609

https://ebooknice.com/product/matematik-5000-kurs-2c-larobok-23848312
(Ebook) SAT II Success MATH 1C and 2C 2002 (Peterson's SAT II
Success) by Peterson's ISBN 9780768906677, 0768906679

https://ebooknice.com/product/sat-ii-success-math-1c-and-2c-2002-peterson-
s-sat-ii-success-1722018

(Ebook) Master SAT II Math 1c and 2c 4th ed (Arco Master the SAT
Subject Test: Math Levels 1 & 2) by Arco ISBN 9780768923049,
0768923042

https://ebooknice.com/product/master-sat-ii-math-1c-and-2c-4th-ed-arco-
master-the-sat-subject-test-math-levels-1-2-2326094

(Ebook) Cambridge IGCSE and O Level History Workbook 2C - Depth

Study: the United States, 1919-41 2nd Edition by Benjamin
Harrison ISBN 9781398375147, 9781398375048, 1398375144,
1398375047
https://ebooknice.com/product/cambridge-igcse-and-o-level-history-
workbook-2c-depth-study-the-united-states-1919-41-2nd-edition-53538044

(Ebook) Linear and Generalized Linear Mixed Models and Their

Applications by Jiming Jiang, Thuan Nguyen ISBN 9781071612811,
9781071612828, 1071612816, 1071612824

https://ebooknice.com/product/linear-and-generalized-linear-mixed-models-
and-their-applications-23848400

(Ebook) Women and Addiction: A Comprehensive Handbook by

Kathleen T. Brady (editor), Sudie E. Back (editor), Shelly F.
Greenfield (editor) ISBN 9781606231074, 1606231073

https://ebooknice.com/product/women-and-addiction-a-comprehensive-
handbook-42827820
Linear Mixed Models
ISTUDY
 Linear Mixed Models
A Practical Guide Using Statistical
Software
Third Edition

Brady T. West
Kathleen B. Welch
Andrzej T. Gałecki
with contributions from Brenda W. Gillespie

ISTUDY
 Third edition published 2022
by CRC Press
6000 Broken Sound Parkway NW, Suite 300, Boca Raton, FL 33487-2742

and by CRC Press

4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN

© 2022 Brady T. West, Kathleen B. Welch, Andrzej T. Gałecki

First edition published by CRC Press 2006

Second edition published by CRC Press 2015

CRC Press is an imprint of Taylor & Francis Group, LLC

Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have
attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders
if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please
write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized
in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying,
microfilming, and recording, or in any information storage or retrieval system, without written permission from the
publishers.

For permission to photocopy or use material electronically from this work, access www.copyright.com or contact the
Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. For works that are
not available on CCC please contact [email protected]

Trademark notice: Product or corporate names may be trademarks or registered trademarks and are used only for
identification and explanation without intent to infringe.

LCCN no. 2022014298

ISBN: 978-1-032-01932-1 (hbk)

ISBN: 978-1-032-01933-8 (pbk)
ISBN: 978-1-003-18106-4 (ebk)

DOI: 10.1201/9781003181064

Typeset in CMR10 font

by KnowledgeWorks Global Ltd.

Publisher’s note: This book has been prepared from camera-ready copy provided by the authors

ISTUDY
 Dedication

To Laura, Carter, and Everleigh

To all of my mentors, advisors, and teachers, especially my parents and grandparents
—B.T.W.

To Jim, Tracy, and Brian

To our grandchildren, Lu and Sam
In memory of my parents, Fremont and June
—K.B.W.

To Viola, my children and grandchildren

To my teachers and mentors
In memory of my parents
—A.T.G.

ISTUDY
ISTUDY
 Contents

Preface to the Third Edition xv

Preface to the Second Edition xvii

Preface xix

The Authors xxi

Acknowledgments xxiii

List of Tables xxv

List of Figures xxvii

1 Introduction 1
1.1 What are Linear Mixed Models (LMMs)? . . . . . . . . . . . . . . . . . . . 1
1.1.1 Models with Random Effects for Clustered Data . . . . . . . . . . . 2
1.1.2 Models for Longitudinal or Repeated-Measures Data . . . . . . . . 2
1.1.3 The Purpose of This Book . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.4 Outline of Book Contents . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 A Brief History of LMMs . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.1 Key Theoretical Developments . . . . . . . . . . . . . . . . . . . . . 5
1.2.2 Key Software Developments . . . . . . . . . . . . . . . . . . . . . . . 7

2 Linear Mixed Models: An Overview 9

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.1 Types and Structures of Data Sets . . . . . . . . . . . . . . . . . . . 9
2.1.1.1 Clustered Data vs. Repeated-Measures and Longitudinal
Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.1.2 Levels of Data . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.2 Types of Factors and Their Related Effects in an LMM . . . . . . . 12
2.1.2.1 Fixed Factors . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.2.2 Random Factors . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.2.3 Fixed Factors vs. Random Factors . . . . . . . . . . . . . 13
2.1.2.4 Fixed Effects vs. Random Effects . . . . . . . . . . . . . . 13
2.1.2.5 Nested vs. Crossed Factors and Their Corresponding Effects 14
2.2 Specification of LMMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.1 General Specification for an Individual Observation . . . . . . . . . . 16
2.2.2 General Matrix Specification . . . . . . . . . . . . . . . . . . . . . . 17
2.2.2.1 Covariance Structures for the D Matrix . . . . . . . . . . . 19
2.2.2.2 Covariance Structures for the Ri Matrix . . . . . . . . . . 20
2.2.2.3 Group-Specific Covariance Parameter Values for the D and
Ri Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . 21

vii

ISTUDY
 viii Contents

2.2.3 Alternative Matrix Specification for All Subjects . . . . . . . . . . . 22

2.2.4 Hierarchical Linear Model (HLM) Specification of the LMM . . . . 22
2.3 The Marginal Linear Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3.1 Specification of the Marginal Model . . . . . . . . . . . . . . . . . . 23
2.3.2 The Marginal Model Implied by an LMM . . . . . . . . . . . . . . . 23
2.4 Estimation in LMMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.1 Maximum Likelihood (ML) Estimation . . . . . . . . . . . . . . . . 25
2.4.1.1 Special Case: Assume θ Is Known . . . . . . . . . . . . . . 26
2.4.1.2 General Case: Assume θ Is Unknown . . . . . . . . . . . . 27
2.4.2 REML Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.4.3 REML vs. ML Estimation . . . . . . . . . . . . . . . . . . . . . . . . 28
2.5 Computational Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5.1 Algorithms for Likelihood Function Optimization . . . . . . . . . . . 29
2.5.2 Computational Problems with Estimation of Covariance Parameters 31
2.6 Tools for Model Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.6.1 Basic Concepts in Model Selection . . . . . . . . . . . . . . . . . . . 34
2.6.1.1 Nested Models . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.6.1.2 Hypotheses: Specification and Testing . . . . . . . . . . . . 34
2.6.2 Likelihood Ratio Tests (LRTs) . . . . . . . . . . . . . . . . . . . . . 35
2.6.2.1 Likelihood Ratio Tests for Fixed-Effect Parameters . . . . 35
2.6.2.2 Likelihood Ratio Tests for Covariance Parameters . . . . . 35
2.6.3 Alternative Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.6.3.1 Alternative Tests for Fixed-Effect Parameters . . . . . . . . 37
2.6.3.2 Alternative Tests for Covariance Parameters . . . . . . . . 38
2.6.4 Information Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.7 Model-Building Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.7.1 The Top-Down Strategy . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.7.2 The Step-Up Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.8 Checking Model Assumptions (Diagnostics) . . . . . . . . . . . . . . . . . . 41
2.8.1 Residual Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.8.1.1 Raw Residuals . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.8.1.2 Standardized and Studentized Residuals . . . . . . . . . . . 42
2.8.2 Influence Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
2.8.3 Diagnostics for Random Effects . . . . . . . . . . . . . . . . . . . . . 43
2.9 Other Aspects of LMMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.9.1 Predicting Random Effects: Best Linear Unbiased Predictors . . . . 46
2.9.2 Intraclass Correlation Coefficients (ICCs) . . . . . . . . . . . . . . . 47
2.9.3 Problems with Model Specification (Aliasing) . . . . . . . . . . . . . 47
2.9.4 Missing Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
2.9.5 Centering Covariates . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
2.9.6 Fitting Linear Mixed Models to Complex Sample Survey Data . . . 50
2.9.6.1 Purely Model-Based Approaches . . . . . . . . . . . . . . 51
2.9.6.2 Hybrid Design- and Model-Based Approaches . . . . . . . 53
2.9.7 Bayesian Analysis of Linear Mixed Models . . . . . . . . . . . . . . 55
2.10 Chapter Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3 Two-Level Models for Clustered Data: The Rat Pup Example 59

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.2 The Rat Pup Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.2.1 Study Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.2.2 Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

ISTUDY
 Contents ix

3.3 Overview of the Rat Pup Data Analysis . . . . . . . . . . . . . . . . . . . 65

3.3.1 Analysis Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.3.2 Model Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
3.3.2.1 General Model Specification . . . . . . . . . . . . . . . . . 67
3.3.2.2 Hierarchical Model Specification . . . . . . . . . . . . . . . 70
3.3.3 Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.4 Analysis Steps in the Software Procedures . . . . . . . . . . . . . . . . . . 73
3.4.1 SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.4.2 SPSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.4.3 R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3.4.3.1 Analysis Using the lme() Function . . . . . . . . . . . . . 92
3.4.3.2 Analysis Using the lmer() Function . . . . . . . . . . . . 96
3.4.4 Stata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.4.5 HLM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
3.4.5.1 Data Set Preparation . . . . . . . . . . . . . . . . . . . . . 104
3.4.5.2 Preparing the Multivariate Data Matrix (MDM) File . . . 105
3.5 Results of Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 109
3.5.1 Likelihood Ratio Tests for Random Effects . . . . . . . . . . . . . . 109
3.5.2 Likelihood Ratio Tests for Residual Error Variance . . . . . . . . . 110
3.5.3 F -Tests and Likelihood Ratio Tests for Fixed Effects . . . . . . . . 111
3.6 Comparing Results across the Software Procedures . . . . . . . . . . . . . 112
3.6.1 Comparing Model 3.1 Results . . . . . . . . . . . . . . . . . . . . . 112
3.6.2 Comparing Model 3.2B Results . . . . . . . . . . . . . . . . . . . . 115
3.6.3 Comparing Model 3.3 Results . . . . . . . . . . . . . . . . . . . . . 115
3.7 Interpreting Parameter Estimates in the Final Model . . . . . . . . . . . . 115
3.7.1 Fixed-Effect Parameter Estimates . . . . . . . . . . . . . . . . . . . 115
3.7.2 Covariance Parameter Estimates . . . . . . . . . . . . . . . . . . . . 120
3.8 Estimating the Intraclass Correlation Coefficients (ICCs) . . . . . . . . . . 121
3.9 Calculating Predicted Values . . . . . . . . . . . . . . . . . . . . . . . . . 124
3.9.1 Litter-Specific (Conditional) Predicted Values . . . . . . . . . . . . 124
3.9.2 Population-Averaged (Unconditional) Predicted Values . . . . . . . 125
3.10 Diagnostics for the Final Model . . . . . . . . . . . . . . . . . . . . . . . . 125
3.10.1 Residual Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . 125
3.10.1.1 Conditional Residuals . . . . . . . . . . . . . . . . . . . . 125
3.10.1.2 Conditional Studentized Residuals . . . . . . . . . . . . . 127
3.10.2 Distribution of BLUPs . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.10.3 Influence Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . 130
3.10.3.1 Influence on Covariance Parameters . . . . . . . . . . . . . 133
3.10.3.2 Influence on Fixed Effects . . . . . . . . . . . . . . . . . . 134
3.11 Software Notes and Recommendations . . . . . . . . . . . . . . . . . . . . 134
3.11.1 Data Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
3.11.2 Syntax vs. Menus . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.11.3 Heterogeneous Residual Error Variances for Level 2 Groups . . . . . 135
3.11.4 Display of the Marginal Covariance and Correlation Matrices . . . . 135
3.11.5 Differences in Model Fit Criteria . . . . . . . . . . . . . . . . . . . . 135
3.11.6 Differences in Tests for Fixed Effects . . . . . . . . . . . . . . . . . 136
3.11.7 Post-Hoc Comparisons of LS Means (Estimated Marginal Means) . 138
3.11.8 Calculation of Studentized Residuals and Influence Statistics . . . . 138
3.11.9 Calculation of EBLUPs . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.11.10 Tests for Covariance Parameters . . . . . . . . . . . . . . . . . . . . 138
3.11.11 Reference Categories for Fixed Factors . . . . . . . . . . . . . . . . 139

ISTUDY
 x Contents

4 Three-Level Models for Clustered Data: The Classroom Example 141

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
4.2 The Classroom Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
4.2.1 Study Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
4.2.2 Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
4.2.2.1 Data Set Preparation . . . . . . . . . . . . . . . . . . . . . 145
4.2.2.2 Preparing the Multivariate Data Matrix (MDM) File . . . 145
4.3 Overview of the Classroom Data Analysis . . . . . . . . . . . . . . . . . . 149
4.3.1 Analysis Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
4.3.2 Model Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
4.3.2.1 General Model Specification . . . . . . . . . . . . . . . . . 151
4.3.2.2 Hierarchical Model Specification . . . . . . . . . . . . . . . 152
4.3.3 Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
4.4 Analysis Steps in the Software Procedures . . . . . . . . . . . . . . . . . . 157
4.4.1 SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
4.4.2 SPSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
4.4.3 R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
4.4.3.1 Analysis Using the lme() Function . . . . . . . . . . . . . 169
4.4.3.2 Analysis Using the lmer() Function . . . . . . . . . . . . 172
4.4.4 Stata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
4.4.5 HLM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
4.5 Results of Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 186
4.5.1 Likelihood Ratio Tests for Random Effects . . . . . . . . . . . . . . 186
4.5.2 Likelihood Ratio Tests and t-Tests for Fixed Effects . . . . . . . . . 187
4.6 Comparing Results Across the Software Procedures . . . . . . . . . . . . . 188
4.6.1 Comparing Model 4.1 Results . . . . . . . . . . . . . . . . . . . . . 188
4.6.2 Comparing Model 4.2 Results . . . . . . . . . . . . . . . . . . . . . 190
4.6.3 Comparing Model 4.3 Results . . . . . . . . . . . . . . . . . . . . . 190
4.6.4 Comparing Model 4.4 Results . . . . . . . . . . . . . . . . . . . . . 190
4.7 Interpreting Parameter Estimates in the Final Model . . . . . . . . . . . . 195
4.7.1 Fixed-Effect Parameter Estimates . . . . . . . . . . . . . . . . . . . 195
4.7.2 Covariance Parameter Estimates . . . . . . . . . . . . . . . . . . . . 196
4.8 Estimating the Intraclass Correlation Coefficients (ICCs) . . . . . . . . . . 197
4.9 Calculating Predicted Values . . . . . . . . . . . . . . . . . . . . . . . . . 199
4.9.1 Conditional and Marginal Predicted Values . . . . . . . . . . . . . . 199
4.9.2 Plotting Predicted Values Using HLM . . . . . . . . . . . . . . . . . 200
4.10 Diagnostics for the Final Model . . . . . . . . . . . . . . . . . . . . . . . . 201
4.10.1 Plots of the EBLUPs . . . . . . . . . . . . . . . . . . . . . . . . . . 201
4.10.2 Residual Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . 202
4.11 Software Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
4.11.1 REML vs. ML Estimation . . . . . . . . . . . . . . . . . . . . . . . 205
4.11.2 Setting up Three-Level Models in HLM . . . . . . . . . . . . . . . . 205
4.11.3 Calculation of Degrees of Freedom for t-Tests in HLM . . . . . . . . 205
4.11.4 Analyzing Cases with Complete Data . . . . . . . . . . . . . . . . . 206
4.11.5 Miscellaneous Differences . . . . . . . . . . . . . . . . . . . . . . . . 207
4.12 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207

ISTUDY
 Contents xi

5 Models for Repeated-Measures Data: The Rat Brain Example 209

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
5.2 The Rat Brain Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
5.2.1 Study Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
5.2.2 Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
5.3 Overview of the Rat Brain Data Analysis . . . . . . . . . . . . . . . . . . 214
5.3.1 Analysis Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
5.3.2 Model Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
5.3.2.1 General Model Specification . . . . . . . . . . . . . . . . . 215
5.3.2.2 Hierarchical Model Specification . . . . . . . . . . . . . . . 219
5.3.3 Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
5.4 Analysis Steps in the Software Procedures . . . . . . . . . . . . . . . . . . 222
5.4.1 SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
5.4.2 SPSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
5.4.3 R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
5.4.3.1 Analysis Using the lme() Function . . . . . . . . . . . . . 230
5.4.3.2 Analysis Using the lmer() Function . . . . . . . . . . . . 232
5.4.4 Stata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
5.4.5 HLM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
5.4.5.1 Data Set Preparation . . . . . . . . . . . . . . . . . . . . . 239
5.4.5.2 Preparing the MDM File . . . . . . . . . . . . . . . . . . . 240
5.5 Results of Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 244
5.5.1 Likelihood Ratio Tests for Random Effects . . . . . . . . . . . . . . 244
5.5.2 Likelihood Ratio Tests for Residual Error Variance . . . . . . . . . 244
5.5.3 F -Tests for Fixed Effects . . . . . . . . . . . . . . . . . . . . . . . . 245
5.6 Comparing Results across the Software Procedures . . . . . . . . . . . . . 245
5.6.1 Comparing Model 5.1 Results . . . . . . . . . . . . . . . . . . . . . 246
5.6.2 Comparing Model 5.2 Results . . . . . . . . . . . . . . . . . . . . . 246
5.7 Interpreting Parameter Estimates in the Final Model . . . . . . . . . . . . 246
5.7.1 Fixed-Effect Parameter Estimates . . . . . . . . . . . . . . . . . . . 246
5.7.2 Covariance Parameter Estimates . . . . . . . . . . . . . . . . . . . . 252
5.8 The Implied Marginal Covariance Matrix for the Final Model . . . . . . . 253
5.9 Diagnostics for the Final Model . . . . . . . . . . . . . . . . . . . . . . . . 254
5.10 Software Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
5.10.1 Heterogeneous Residual Error Variances for Level 1 Groups . . . . . 258
5.10.2 EBLUPs for Multiple Random Effects . . . . . . . . . . . . . . . . . 258
5.11 Other Analytic Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . 258
5.11.1 Kronecker Product for More Flexible Residual Error Covariance
Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
5.11.2 Fitting the Marginal Model . . . . . . . . . . . . . . . . . . . . . . . 260
5.11.3 Repeated-Measures ANOVA . . . . . . . . . . . . . . . . . . . . . . 261
5.12 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

6 Random Coefficient Models for Longitudinal Data: The Autism

Example 263
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
6.2 The Autism Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
6.2.1 Study Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
6.2.2 Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
6.3 Overview of the Autism Data Analysis . . . . . . . . . . . . . . . . . . . . 269
6.3.1 Analysis Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269

ISTUDY
 xii Contents

6.3.2 Model Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

6.3.2.1 General Model Specification . . . . . . . . . . . . . . . . . 271
6.3.2.2 Hierarchical Model Specification . . . . . . . . . . . . . . . 274
6.3.3 Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
6.4 Analysis Steps in the Software Procedures . . . . . . . . . . . . . . . . . . 276
6.4.1 SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
6.4.2 SPSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
6.4.3 R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
6.4.3.1 Analysis Using the lme() Function . . . . . . . . . . . . . 286
6.4.3.2 Analysis Using the lmer() Function . . . . . . . . . . . . 288
6.4.4 Stata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
6.4.5 HLM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
6.4.5.1 Data Set Preparation . . . . . . . . . . . . . . . . . . . . . 295
6.4.5.2 Preparing the MDM File . . . . . . . . . . . . . . . . . . . 296
6.5 Results of Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 300
6.5.1 Likelihood Ratio Test for Random Effects . . . . . . . . . . . . . . . 300
6.5.2 Likelihood Ratio Tests for Fixed Effects . . . . . . . . . . . . . . . . 301
6.6 Comparing Results across the Software Procedures . . . . . . . . . . . . . 302
6.6.1 Comparing Model 6.1 Results . . . . . . . . . . . . . . . . . . . . . 302
6.6.2 Comparing Model 6.2 Results . . . . . . . . . . . . . . . . . . . . . 302
6.6.3 Comparing Model 6.3 Results . . . . . . . . . . . . . . . . . . . . . 306
6.7 Interpreting Parameter Estimates in the Final Model . . . . . . . . . . . . 306
6.7.1 Fixed-Effect Parameter Estimates . . . . . . . . . . . . . . . . . . . 306
6.7.2 Covariance Parameter Estimates . . . . . . . . . . . . . . . . . . . . 308
6.8 Calculating Predicted Values . . . . . . . . . . . . . . . . . . . . . . . . . 309
6.8.1 Marginal Predicted Values . . . . . . . . . . . . . . . . . . . . . . . 309
6.8.2 Conditional Predicted Values . . . . . . . . . . . . . . . . . . . . . . 311
6.9 Diagnostics for the Final Model . . . . . . . . . . . . . . . . . . . . . . . . 313
6.9.1 Residual Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . 313
6.9.2 Diagnostics for the Random Effects . . . . . . . . . . . . . . . . . . 316
6.9.3 Observed and Predicted Values . . . . . . . . . . . . . . . . . . . . 317
6.10 Software Note: Computational Problems with the D Matrix . . . . . . . . 318
6.10.1 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
6.11 An Alternative Approach: Fitting the Marginal Model with an Unstructured
Covariance Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
6.11.1 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 322

7 Models for Clustered Longitudinal Data: The Dental Veneer Example 323
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
7.2 The Dental Veneer Study . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
7.2.1 Study Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
7.2.2 Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
7.3 Overview of the Dental Veneer Data Analysis . . . . . . . . . . . . . . . . 328
7.3.1 Analysis Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
7.3.2 Model Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
7.3.2.1 General Model Specification . . . . . . . . . . . . . . . . . 331
7.3.2.2 Hierarchical Model Specification . . . . . . . . . . . . . . . 333
7.3.3 Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
7.4 Analysis Steps in the Software Procedures . . . . . . . . . . . . . . . . . . 338
7.4.1 SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
7.4.2 SPSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345

ISTUDY
 Contents xiii

7.4.3 R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
7.4.3.1 Analysis Using the lme() Function . . . . . . . . . . . . . 349
7.4.3.2 Analysis Using the lmer() Function . . . . . . . . . . . . 353
7.4.4 Stata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
7.4.5 HLM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
7.4.5.1 Data Set Preparation . . . . . . . . . . . . . . . . . . . . . 360
7.4.5.2 Preparing the Multivariate Data Matrix (MDM) File . . . 361
7.5 Results of Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 365
7.5.1 Likelihood Ratio Tests for Random Effects . . . . . . . . . . . . . . 365
7.5.2 Likelihood Ratio Tests for Residual Error Variance . . . . . . . . . 366
7.5.3 Likelihood Ratio Tests for Fixed Effects . . . . . . . . . . . . . . . . 366
7.6 Comparing Results across the Software Procedures . . . . . . . . . . . . . 366
7.6.1 Comparing Model 7.1 Results . . . . . . . . . . . . . . . . . . . . . 366
7.6.2 Comparing Results for Models 7.2A, 7.2B, and 7.2C . . . . . . . . . 369
7.6.3 Comparing Model 7.3 Results . . . . . . . . . . . . . . . . . . . . . 372
7.7 Interpreting Parameter Estimates in the Final Model . . . . . . . . . . . . 372
7.7.1 Fixed-Effect Parameter Estimates . . . . . . . . . . . . . . . . . . . 372
7.7.2 Covariance Parameter Estimates . . . . . . . . . . . . . . . . . . . . 374
7.8 The Implied Marginal Covariance Matrix for the Final Model . . . . . . . 375
7.9 Diagnostics for the Final Model . . . . . . . . . . . . . . . . . . . . . . . . 377
7.9.1 Residual Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . 378
7.9.2 Diagnostics for the Random Effects . . . . . . . . . . . . . . . . . . 379
7.10 Software Notes and Recommendations . . . . . . . . . . . . . . . . . . . . 382
7.10.1 ML vs. REML Estimation . . . . . . . . . . . . . . . . . . . . . . . 382
7.10.2 The Ability to Remove Random Effects from a Model . . . . . . . . 382
7.10.3 Considering Alternative Residual Error Covariance Structures . . . 382
7.10.4 Aliasing of Covariance Parameters . . . . . . . . . . . . . . . . . . . 383
7.10.5 Displaying the Marginal Covariance and Correlation Matrices . . . 384
7.10.6 Miscellaneous Software Notes . . . . . . . . . . . . . . . . . . . . . . 384
7.11 Other Analytic Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . 385
7.11.1 Modeling the Covariance Structure . . . . . . . . . . . . . . . . . . 385
7.11.2 The Step-Up vs. Step-Down Approach to Model Building . . . . . . 386
7.11.3 Alternative Uses of Baseline Values for the Dependent Variable . . 386

8 Models for Data with Crossed Random Factors: The SAT Score Example 389
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
8.2 The SAT Score Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
8.2.1 Study Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
8.2.2 Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
8.3 Overview of the SAT Score Data Analysis . . . . . . . . . . . . . . . . . . 394
8.3.1 Model Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . 394
8.3.1.1 General Model Specification . . . . . . . . . . . . . . . . . 394
8.3.1.2 Hierarchical Model Specification . . . . . . . . . . . . . . . 395
8.3.2 Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
8.4 Analysis Steps in the Software Procedures . . . . . . . . . . . . . . . . . . 396
8.4.1 SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
8.4.2 SPSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
8.4.3 R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403
8.4.4 Stata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
8.4.5 HLM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
8.4.5.1 Data Set Preparation . . . . . . . . . . . . . . . . . . . . . 407

ISTUDY
 xiv Contents

8.4.5.2 Preparing the MDM File . . . . . . . . . . . . . . . . . . . 408

8.4.5.3 Model Fitting . . . . . . . . . . . . . . . . . . . . . . . . . 409
8.5 Results of Hypothesis Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 410
8.5.1 Likelihood Ratio Tests for Random Effects . . . . . . . . . . . . . . 410
8.5.2 Testing the Fixed Year Effect . . . . . . . . . . . . . . . . . . . . . 411
8.6 Comparing Results across the Software Procedures . . . . . . . . . . . . . 411
8.7 Interpreting Parameter Estimates in the Final Model . . . . . . . . . . . . 411
8.7.1 Fixed-Effect Parameter Estimates . . . . . . . . . . . . . . . . . . . 412
8.7.2 Covariance Parameter Estimates . . . . . . . . . . . . . . . . . . . . 412
8.8 The Implied Marginal Covariance Matrix for the Final Model . . . . . . . 415
8.9 Recommended Diagnostics for the Final Model . . . . . . . . . . . . . . . 416
8.10 Software Notes and Additional Recommendations . . . . . . . . . . . . . . 417

9 Power Analysis and Sample Size Calculations for Linear Mixed Models 419
9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
9.2 Direct Power Computations . . . . . . . . . . . . . . . . . . . . . . . . . . 419
9.2.1 Software for Direct Power Computations . . . . . . . . . . . . . . . 420
9.2.2 Examples of Direct Power Computations . . . . . . . . . . . . . . . 420
9.3 Examining Power via Simulation . . . . . . . . . . . . . . . . . . . . . . . 429
9.3.1 Examples of Simulation-Based Approaches . . . . . . . . . . . . . . 430

A Statistical Software Resources 437

A.1 Descriptions/Availability of Software Packages . . . . . . . . . . . . . . . . 437
A.1.1 SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
A.1.2 IBM SPSS Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
A.1.3 R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
A.1.4 Stata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
A.1.5 HLM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438
A.2 Useful Internet Links . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

B Calculation of the Marginal Covariance Matrix 441

C Acronyms / Abbreviations 443

Bibliography 445

Index 453

ISTUDY
 Preface to the Third Edition

In late 2020, the authors were contacted by Rob Calver at Chapman & Hall / CRC Press
and asked about the possibility of working on a third edition of Linear Mixed Models: A
Practical Guide Using Statistical Software. As was noted in the second edition, much can
change over a six-year period in terms of the power and availability of software for statistical
analysis, and there have been rapid developments in software and methods for mixed-effects
modeling specifically since the publication of the second edition. When we decided to write
this third edition, we sought to provide a comprehensive update of the available tools for
fitting linear mixed-effects models in the newest versions of SAS, SPSS, R, Stata, and HLM,
giving all of the examples a facelift and focusing on new tools for visualization of results and
interpretation. We also saw the need for the addition of a unique chapter on power analysis
for mixed-effects models, given additional software developments in this area and frequent
questions from colleagues about how exactly to approach this important study design issue.
Finally, new conceptual and theoretical developments in mixed-effects modeling have also
been included.
As was the case with the first two editions of this book, we will continue to provide
readers with updates regarding methodological and software developments on the book’s
website (http://websites.umich.edu/~bwest/almmussp.html). It has always been our
objective to provide readers with an easy-to-read text that will provide clear and pragmatic
guidance on data analysis problems involving linear mixed models, and we hope that this
third edition continues to serve that purpose. Please do not hesitate to contact the authors
with any corrections or suggested additions.

xv

ISTUDY
ISTUDY
 Preface to the Second Edition

Books attempting to serve as practical guides on the use of statistical software are always
at risk of becoming outdated as the software continues to develop, especially in an area of
statistics and data analysis that has received as much research attention as linear mixed
models. In fact, much has changed since the first publication of this book in early 2007, and
while we tried to keep pace with these changes on the website for this book, the demand
for a second edition quickly became clear. There were also a number of topics that were
only briefly referenced in the first edition, and we wanted to provide more comprehensive
discussions of those topics in a new edition. This second edition of Linear Mixed Models: A
Practical Guide Using Statistical Software aims to update the case studies presented in the
first edition using the newest versions of the various software procedures, provide coverage
of additional topics in the application of linear mixed models that we believe valuable for
data analysts from all fields, and also provide up-to-date information on the options and
features of the software procedures currently available for fitting linear mixed models in
SAS, SPSS, Stata, R/S-plus, and HLM.
Based on feedback from readers of the first edition, we have included coverage of the
following topics in this second edition:

• Models with crossed random effects, and software procedures capable of fitting these
models (see Chapter 8 for a new case study);
• Power analysis methods for longitudinal and clustered study designs, including software
options for power analyses and suggested approaches to writing simulations;
• Use of the lmer() function in the lme4 package in R;
• Fitting linear mixed models to complex sample survey data;
• Bayesian approaches to making inferences based on linear mixed models; and
• Updated graphical procedures in the various software packages.

We hope that readers will find the updated coverage of these topics helpful for their
research activities.
We have substantially revised the subject index for the book to enable more efficient
reading and easier location of material on selected topics or software options. We have
also added more practical recommendations based on our experiences using the software
throughout each of the chapters presenting analysis examples. New sections discussing over-
all recommendations can be found at the end of each of these chapters. Finally, we have
created an R package named WWGbook that contains all of the data sets used in the example
chapters.
We will once again strive to keep readers updated on the website for the book, and also
continue to provide working, up-to-date versions of the software code used for all of the
analysis examples on the website. Readers can find the website at the following address:
http://www.umich.edu/~bwest/almmussp.html

xvii

ISTUDY
ISTUDY
 Preface

The development of software for fitting linear mixed models was propelled by advances in
statistical methodology and computing power in the late 20th century. These developments,
while providing applied researchers with new tools, have produced a sometimes confusing
array of software choices. At the same time, parallel development of the methodology in
different fields has resulted in different names for these models, including mixed models,
multilevel models, and hierarchical linear models. This book provides a reference on the use
of procedures for fitting linear mixed models available in five popular statistical software
packages (SAS, SPSS, Stata, R/S-plus, and HLM). The intended audience includes applied
statisticians and researchers who want a basic introduction to the topic and an easy-to-
navigate software reference.
Several existing texts provide excellent theoretical treatment of linear mixed models
and the analysis of variance components (e.g., McCulloch and Searle, 2001; Searle, Casella,
and McCulloch, 1992; Verbeke and Molenberghs, 2000); this book is not intended to be
one of them. Rather, we present the primary concepts and notation, and then focus on
the software implementation and model interpretation. This book is intended to be a refer-
ence for practicing statisticians and applied researchers, and could be used in an advanced
undergraduate or introductory graduate course on linear models.
Given the ongoing development and rapid improvements in software for fitting linear
mixed models, the specific syntax and available options will likely change as newer versions
of the software are released. The most up-to-date versions of selected portions of the syn-
tax associated with the examples in this book, in addition to many of the data sets used
in the examples, are available at the following website: http://www.umich.edu/~bwest/
almmussp.html

xix

ISTUDY
Random documents with unrelated
content Scribd suggests to you:
 The remaining quarter of a column was occupied by a
circumstantial and highly colloquial account of the incidents recorded
in the Book of Genesis, and it very plainly suggested that the work
had been published by Messrs. Hoskins as a satire upon the success
of the Hebrew race in the New England States. The reviewer even
made an attempt to identify Joseph with a prominent Republican
politician, and Potiphar’s wife with the Democratic party, who were
alleged to be making overtures to the same gentleman.
But I really did once meet with a sub-editor who had reviewed
“The Swiss Family Robinson” as a new work. He commenced by
telling the readers of the newspaper that the book was a wholesome
story of a worthy Swiss pastor, and so forth.
I also knew a musical critic who, on being entrusted with the duty
of writing a notice of Il Don Giovanni, as performed by the Carl Rosa
Company, began as follows: “Don Giovanni, the gentleman from
whom the opera takes its name, was a licentious Spanish nobleman
of the past century.” The notice gave some account of the affaires of
this newly-discovered reprobate, glossing over the Zerlina business
rather more than Mozart thought necessary to do, but being very
bitter against Leporello, “his valet and confidant,” and finally
expressing the opinion somewhat dogmatically that “few of the
public would be disposed to say that the fate which overtook this
callous scoundrel was not well earned by his persistence in a course
of unjustifiable vice. The music is tuneful and was much encored.”
Upon the occasion of this particular representation I recollect that
I wrote, “An Italian version of a Spanish story, set to music by a
German, conducted by a Frenchman, and interpreted by a Belgian, a
Swiss, an Irishman and a Canadian—this is what is meant by English
Opera.”
My notice gave great offence; but the other was considered
excellent.
 The moral tone that pervaded it was most praiseworthy, the
people said.
And so it was.
I have got about five hundred musical jottings which, if provoked,
I may one day publish; but, meantime, I cannot refrain from giving
one illustration of the way in which musical notices were managed
long ago.
Madame Adelina Patti had made her first (and farewell)
appearance in the town where I was located. I was engaged about
two o’clock in the morning putting what I considered to be the
finishing touches to the column which I had written about the diva’s
concert, when the reporter of the leading paper burst into the room
in which I was writing. He was in rather a dishevelled condition, and
he approached me and whispered that he wanted to ask me a
question outside—there were others in the room. I went through the
door with him and inquired what I could do for him.
“I was marked for that blessed concert, and I went too, and now
I’m writing the notice,” said he. “But what I want to know is this—Is
Patti a soprano or a contralto?”

I have just now discovered that it would be unwise for me to

continue very much farther these reminiscences of editors and sub-
editors, the fact being that I have some jottings about every one of
the race whom I have ever met, and when one gets into a desultory
vein of anecdotage like that in which I now find myself for the first
time in my life, one is liable to exhaust a reader’s forbearance before
one’s legitimate subject has become exhausted. I think it may be
prudent to make a diversion at this period from the sub-editors of
the past to the suppers of the newspaper office. Gastronomy as a
science is not drawn out to its finest point within these precincts.
There is still something left to be desired by such persons as are
fastidious. I have for long thought that it would be by no means
extravagant to expect every newspaper office to be supplied with a
kitchen, properly furnished, and with the “good plain cook,” who so
constantly figures in the columns (advertising), at hand to turn out
the suppers for all departments engaged in the production of the
paper.
It is inconvenient for an editor to be compelled to cook his own
supper at his gas stove, while the flimsies of the speech upon which
he is writing are being laid on his desk by the sub-editor, and the
foreman’s messenger is asking for them almost before they have
ceased to flutter in the cooling draught created by opening the door.
Equally inconvenient is it for the sub-editor and the reporters to get
something to prevent them from succumbing to starvation. The
compositors in some offices have lately instituted a rule by which
they “knock off” for supper at half-past ten; but what sort of a meal
do they get to sustain them until four in the morning? I have no
hesitation in pronouncing it to be almost as indifferent as that upon
which the editor is forced to subsist for, perhaps, the same period. I
have seen the compositors—some of them earning £5 a week—
crouching under their cases, munching hunches (the onomatopæia
is Homeric) of bread, while their cans of tea—that abomination of
cold tea warmed up—were stewing over their gas burners.
In the sub-editors’ room, and the reporters’ room, tea was also
being cooked, or bottles of stout drunk, the accompanying,
comestibles being bread or biscuits. After swallowing tea that has
been stewing on its leaves for half-an-hour, and eating a slab of
office bread out of one hand while the other holds the pen, the
editor writes an article on the grievances of shopmen who are only
allowed an hour for dinner and half-an-hour for tea; or, upon the
slavery of a barmaid; or, perhaps, composes a nice chatty half-
column on the progress of dyspepsia and the necessity for attending
carefully to one’s diet.
Now, I affirm that no newspaper office should be without a
kitchen. The compositors should be given a chance of obtaining all
the comforts of home at a lesser cost than they could be provided at
home; and later on in the night the reporters, sub-editors, and editor
should be able to send up messages as to the hour they mean to
take supper, and the dish which they would like to have. Here is an
opportunity for the Institute of Journalists. Let them take sweet
counsel together on the great kitchen question, and pass a
resolution “that in the opinion of the Institute a kitchen in complete
working order should form part of every morning newspaper office;
and that a cook, holding a certificate from South Kensington, or,
better still, Mrs. Marshall, should be regarded as essential to the
working staff as the editor.”
I do not say that a box of Partagas, or Carolinas, should be
provided by the management for every room occupied by the literary
staff; though undoubtedly a move in the right direction, yet I fear
that public feeling has not yet been sufficiently aroused by the bitter
cry of the journalist, to make the cigar-box and the club chair
probable; but I do say that since journalism has become a
profession, those who practise it should be treated as if they were as
deserving of consideration as the salesmen in drapers’ shops. Surely,
as we have sent the bitter cry into all the ends of the earth on behalf
of others, we might be permitted the luxury of a little bitter cry on
our own account.

This brings me down to the recollections I retain of the strange

ideas that some of the staff of journals with which I have been
connected, possessed as to the most appropriate menu for supper.
One of these gentlemen, for instance, was accustomed to make
oatmeal porridge in a saucepan for himself about two o’clock in the
morning. When accused of being a Scotchman, he indignantly
denied that he was one. He admitted, however, that he was an
Ulsterman, and this was considered even worse by his accusers. He
invariably alluded to the porridge in the plural, calling it “them.” I
asked him one night why the thing was entitled to a plural, and he
said it was because no one but a blue-pencilled fool would allude to
it as otherwise. I had the curiosity to inquire farther how much
porridge was necessary to be in the saucepan before it became
entitled to a plural; if, for instance, there was only a spoonful, surely
it would be rather absurd to still speak of it as “them.” He replied,
after some thought, that though he had never considered the matter
in all its bearings, yet his impression was that even a spoonful was
entitled to a plural.
“Did you ever hear any one allude to brose as ‘it’?” he asked.
I admitted that I never had.
“Then if you call brose ‘them,’ why shouldn’t you call stirabout
‘them’?” he asked, triumphantly.
“I must confess that I never had the matter brought so forcibly
before me,” said I.
As he was going to “sup them,” as he termed the operation of
ladling the contents of the saucepan into his mouth, I hastily left the
room. I have eaten tiffin within easy reach of a dozen lepers on
Robben Island in Table Bay, I have taken a hearty supper in a tent
through which a camel every now and again thrust its nose, I have
enjoyed a biltong sandwich on the seat of an African bullock waggon
with a Kaffir beside me, I have even eaten a sausage snatched by
the proprietor from the seething panful in the window of a shop in
the Euston Road—I did so to celebrate the success of a play of mine
at the Grand Theatre—but I could not remain in the room while that
literary gentleman partook of that simple supper of his.
On my return when he had finished I never failed to allow in the
most cordial way the right of the preparation to a plural. It was to be
found in every part of the room; the table, the chairs, the floor, the
fireplace, the walls, the ceiling—all bore token to the fact that it was
not one but many.
In the hands of a true Ulsterman stirabout “are” a terrible weapon.
As a mural decorative medium “they” leave much to be desired.
 Only one man connected with the Press did
I ever know addicted to the bloater as a supper dish. The man
came among us like a shadow and disappeared as such, after a
week of incompetence; but he left a memory behind him that not all
the perfumes of Arabia can neutralise. It was about one o’clock in
the morning—he had come on duty that night—that there floated
through the newspaper office a dense blue smoke and a smell—such
a smell! It was of about the same density as an ironclad. One felt
oneself struggling through it as though it were a mass of chilled
steel plates, backed with soft iron. On the upper floor we were built
in by it, so to speak. It arose on every side of us like the wall of a
prison, and we kept groping around it for a hole large enough to
allow of our crawling through. Two of us, after battering at that
smell for a quarter of an hour, at last discovered a narrow passage in
it made by a current of air from an open window, and having
squeezed ourselves through, we ran downstairs to the sub-editors’
room.
Through the crawling blue smoke we could just make out the
figure of a man standing in his shirt sleeves in front of the fire using
a large two-pronged iron fork as a toothpick. On a plate on the table
lay the dislocated backbone of a red herring (harengus rufus).
The man was perfectly self-possessed. We questioned him closely
about the origin of the smoke and the smell, and he replied that,
without going so far as to pronounce a dogmatic opinion on the
subject, and while he was quite ready to accept any reasonable
suggestion on the matter from either of us, he, for his part, would
not be at all surprised if it were found on investigation that both
smoke and smell were due to his having openly cooked a rather
bloated specimen of the Yarmouth bloater. He always had one for his
supper, he said; critically, when not too pungent—he disliked them
too pungent—he considered that a full-grown bloater, well preserved
for its years and considering the knocking about that it must have
had, was fully equal to a beefsteak. There was much more practical
eating in it, he should say, speaking as man to man. And it was so
very simple—that was its great charm.
 For himself, he never could bear made-up dishes; they were, he
thought, usually rich, and he had a poor-enough digestion, so that
he could not afford to trifle with it.
Just then the foreman loomed through the dense smoke, and,
being confronted with the hydra-headed smell, he boldly grappled
with it, and after a fierce contest, he succeeded in strangling one of
the heads and then set his foot on it. He hurriedly explained to the
subeditor that all the hands who had lifted the copy that had been
sent out were setting it up with bowls of water beside them to save
themselves the trouble of going to the water-tap for a drink.
The next day the clerks in the mercantile department were
working with bottles of carbolic under their noses, and every now
and again a note would be brought in from a subscriber ordering his
paper to be stopped until a new consignment of printers’ ink should
arrive, in which the chief ingredient was not so pungent.
At the end of a week the sub-editor was given a month’s salary
and an excellent testimonial, and was dismissed. The proprietor of
the journal had the sub-editors’ room freshly painted and papered,
and made the assistant-editor a present of two pounds to buy a new
coat to replace the one which, having hung in the room for an entire
night, had to be burnt, no cleaner being found who would accept the
risk of purifying it. The cleaners all said that they would not run the
chance of having all the contents of their vats left on their hands.
They weren’t as a rule squeamish in the matter of smells; they only
drew the line at creosote, and the coat was a long way on the other
side.
Seven years have passed since that sub-editor partook of that
simple supper, and yet I hear that every night drag-hounds howl at
the door of the room, and strangers on entering sniff, saying,—
“Whew! there’s a barrel of red herrings somewhere about.”
 CHAPTER IX.—ON THE HUMAN
IMAGINATION.
Mr. Henry Irving and the Stag’s Head—The sense of smell—A
personal recollection—Caught “tripping”—The German band—In the
pre-Wagnerian days—Another illustration of a too-sensitive
imagination—The doctor’s letter—Its effects—A sudden recovery—
The burial service is postponed indefinitely.

I
T might be as well, I fancy, to accept with caution the statement
made in the last lines of the foregoing chapter. At any rate, I
may frankly confess that I have always done so, knowing how
apt one is to be carried away by one’s imagination in some matters.
Mr. Henry Irving told me several years ago a curious story on this
very point, and in regard also to the way in which the imagination
may be affected through the sense of smell.
When he was very young he was living at a town in the west of
England, and in one of the streets there was a hostelry which bore a
swinging sign with a stag’s head painted upon it, with a sufficient
degree of legibility to enable casual passers-by to know what it was
meant to simulate. But every time he saw this sign, he had a feeling
of nausea that he could overcome only by hurrying on down the
street. Mr. Irving explained to me that it did not appear to him that
this nausea was the result of an offended artistic perception owing
to any indifferent draughtsmanship or defective technique in the
production of the sign. It actually seemed to him that the painted
stag possesses some influence akin to the evil eye, and it was
altogether very distressing to him. After a short time he left the
town, and did not revisit it until he had attained maturity; and then,
remembering the stag’s head and the curious way in which it had
affected him long before, he thought he would look up the old place,
if it still existed, and try if the evil charm of the sign had ceased to
retain its potency upon him. He walked down the street; there the
sign was swinging as of old, and the moment he saw it he had a
feeling of nausea. Now, however, he had become so impregnated
with the investigating spirit of the time, that he determined to search
out the origin of the malign influence of the neighbourhood; and
then he discovered that the second house from the hostelry was a
soap and candle factory, on a sufficiently extensive scale to make a
daily “boiling” necessary. It was the odour arising from this
enterprise that induced the disagreeable sensation which he had
experienced years before, and from which few persons are free
when in the neighbourhood of tallow in a molten state.
I do not think that this story has been published. But even if it has
appeared elsewhere it scarcely requires an apology.

Though wandering even more widely than usual from my text—

after all, my texts are only pretexts for unlimited ramblings—I will
give another curious but perfectly authentic case of the force of
imagination. In this case the imagination was reached through the
sense of hearing.
At one time I lived in a town at the extremity of a very fine bay, at
the entrance to which there was a small village with a little bay of its
own and a long stretch of sand, the joy of the “tripper.” I was a
“tripper” of six in those days, and during the summer months an
excursion by steamer on the bay was one of the most joyous of
experiences. But the steamer was a very small one, and apt to yield
rather more than is consistent with modern ideas of marine stability
to the pressure of the waves, which in a north-easterly wind—the
prevailing one—were pretty high in our bay. The effect of this
instability was invariably disastrous to a maiden aunt who was
supposed to share with me the enjoyment of being caught
“tripping.” With the pertinacity of a man of six carrying a model of a
cutter close to his bosom, I refused to “go below” under the
circumstances, with my groaning but otherwise august relative, and
she was usually extremely unwell. It so happened, however, that the
proprietors of the steamboat were sufficiently enterprising to engage
—perhaps I should say, to permit—a German band to drown the
groans of the sufferers in the strains of the beautiful “Blue Danube,”
or whatever the waltz of the period may have been—the “Blue
Danube” is the oldest that I can remember. Now, when the “season”
was over, and the steamer was laid up for the winter, the Germans
were accustomed to give open-air performances in the town; so that
during the winter months we usually had a repetition on land of the
summer’s répertoire at sea. The first bray that was given by the
trombone in the region of the square where we lived was, however,
quite enough to make my aunt give distinct evidence of feeling “a
little squeamish”; by the time the oboe had joined hands, so to
speak, with the parent of all evil, the trombone, she had taken out
her handkerchief and was making wry faces beneath her palpably
false scalpet. But when the wry-necked fife, and the serpent—the
sea-serpent it was to her—were doing their worst in league with, but
slightly indifferent to, the cornet and the Saxe-horn, my aunt retired
from the apartment amid the derisive yells of the young demons in
the schoolroom, and we saw her no more until the master of the
music had pulled the bell of the hall-door, and we had insulted him
in his own language by shouting through the blinds “schlechte
musik!—sehr schlechte musik!” We were ready enough to learn a
language for insulting purposes, just as a parrot which declines to
acquire the few refined words of its mistress, will, if left within the
hearing of a groom, repeat quite glibly and joyously, phrases which
make it utterly useless as a drawing-room bird in a house where a
clergyman makes an occasional call. For years my aunt could never
hear a German band without emotion, since the crazy little steamer
had danced to their strains. In this case, it must also be remarked,
the feeling was not the result of a highly-developed artistic
temperament. The blemishes of the musical performances were in
no way accountable for my relative’s emotions, though I believe that
the average German band frequenting what theatrical-touring
companies call “B. towns,” might reasonably be regarded as
sufficient to precipitate an incipient disorder. No, it was the force of
imagination that brought about my aunt’s disaster, which, I regret to
say, I occasionally purchased, when I felt that I owed myself a treat,
for a penny, for this was the lowest sum that the impresario would
take to come round our square and make my aunt sick. The sum
was so absurdly low, considering the extent of the results produced,
I am now aware that no really cultured musician, no impresario with
any self-respect, would have accepted it to bring his band round the
corner; but when one reflects that the sum on the original scrittura
was invariably doubled—for my aunt sent a penny out when her
sufferings became intense, to induce the band to go away—the
transaction assumes another aspect.
We hear of the enormous increase in the salaries paid to musical
artists nowadays, and as an instance of this I may mention that a
friend of mine a few months ago, having occasion for the services of
a German band—not for medicinal purposes but for a philological
reason—was forced to pay two shillings before he could effect his
object! Truly the conditions under which art is pursued have
undergone a marvellous change within a quarter of a century. I
could have made my aunt sick twenty-four times for the sum
demanded for a single performance nowadays. And in the sixties, it
must also be remembered, Wagner had not become a power.

Strong-minded persons, such as the first Lord Brougham, may

take a sardonic delight in reading their own obituary notices, and
such persons would probably scoff at the suggestion made in an
earlier chapter, that the shock of reading the record of his death in a
newspaper might have a disastrous effect upon a man, but there is
surely no lack of evidence to prove the converse of “mentem
mortalia tangunt.”
I heard when in India a story which seemed to me to be, as an
illustration of the effects of imagination, quite as curious as the well-
known case of the sailor who became cured of scurvy through
fancying that the clinical thermometer with which the surgeon took
his temperature was a drastic remedy. A young civil servant at
Colombo felt rather fagged after an unusually long stretch of work,
and made up his mind to consult the best doctor in the place. He did
so, and the doctor went through the usual probings and
stethoscopings, and then looked grave and went over half the
surface again. He said he thought that on the whole he had better
write his opinion of the “case” in all its particulars and send it to the
patient.
The next morning the patient received the following letter:—
“My dear Sir,—I think it only due to the confidence which you have
placed in me to let you know in the plainest words what is the result
of my diagnosis of your condition. Your left lung is almost gone, but
with care you might survive its disappearance. Unhappily, however,
the cardiac complications which I suspected are such as preclude the
possibility of your recovery. In brief, I consider it to be my duty to
advise you to lose no time in carrying out any business
arrangements that demand your personal attention. You may of
course live for some weeks; but I think you would do wisely to count
only on days.
“Meantime, I would suggest no material change in your diet,
except the reduction of your brandy pegs to seven per diem.”
This letter was put into the hands of the unfortunate man when
he returned from his early ride the next morning. Its effect was to
diminish to an appreciable degree his appetite for breakfast. He sat
motionless on his chair out on the verandah and stared at the letter
—it was his death-warrant. After an hour he felt a difficulty in
breathing. He remembered now that he had always been uneasy
about his lungs—his left in particular. He put his hand over the place
where he supposed his heart to lie concealed. How could he have
lived so many years in the world without becoming aware of the fact
that as an every-day sort of an organ—leaving the higher emotions
out of the question altogether—his heart was a miserable failure?
Sympathy, friendship, love, emotion,—he would not have minded if
his heart were incapable of these, if it only did its business as a
blood pump; but it was perfectly plain from the manner in which it
throbbed beneath his hand, that it was deserving of all the
reprobation the doctor had heaped upon it.
His difficulty of respiration increased, and with this difficulty he
became conscious of an acute pain under his ribs. He found when he
attempted to rise that he could only do so with an effort. He
managed to totter into his bedroom, and when he threw himself on
his bed, it was with the feeling that he should never rise from it
again.
His faithful Khânsâmah more than once inquired respectfully if the
Preserver of the Poor would like to have the Doctor Sahib sent for,
and if the Joy of the Whole World would in the meantime drink a
peg. But the Preserver of the Poor had barely strength to express
the hope that the disappearance of the Doctor Sahib might be
effected by a supernatural agency, and the Joy of the Whole World
could only groan at the suggestion of a peg. The pain under his ribs
was increasing, and he had a general nightmare feeling upon him.
Toward evening he sank into a lethargy, and at this point the
Khânsâmah made up his mind that the time for action had come; he
went for the doctor himself, and was fortunate enough to meet him
going out in his buggy to dine.
“What on earth have you been doing with yourself?” he inquired,
when he had felt the pulse of the patient. “Why, you’ve no pulse to
speak of, and your skin—What the mischief have you been doing
since yesterday?”
“How can you expect a chap’s pulse to be anything particular
when he has no heart worth speaking of?” gasped the patient.
“Who has no heart worth speaking of?”
The patient looked piteously up at him.
“That’s kicking a man when he’s down,” he murmured.
“What’s the matter with you anyway?” said the doctor. “Your
heart’s all right, I know—at least, it was all right yesterday. Is it your
liver? Let me have a look at your eyes.”
He certainly did let the doctor have a look at his eyes. He lay
staring at the good physician for some minutes.
 “No, your liver is no worse than it was yesterday,” said the doctor,
“Do you mean to say that your letter was only a joke?” said the
patient, still staring.
“A joke? Don’t be a fool. Do you fancy that I play jokes upon my
patients? I wrote to you what was the exact truth. I flatter myself I
always tell the truth even to my patients.”
“Oh,” groaned the patient. “And after telling me that I hadn’t more
than a few days to live you now say my heart’s all right.”
“You’re mad, my good fellow, mad! I said that you must go
without the delay of a day for a change—a sea voyage if possible—
and that in a week you’d be as well as you ever were. Where’s the
letter?”
It was lying on the side of the bed. The patient had read it again
after he had thrown himself down.
“My God!” cried the doctor, when he had brought it over to the
lamp. “An awful thing has happened. This is the letter that I wrote to
Lois Perez, the diamond merchant, who visited me yesterday just
before you came. My assistant must have put the letter that was
meant for Perez into the envelope addressed to you, and your letter
into the other cover. Great heavens!”
The patient was sitting up in the bed.
“You mean to say that—that—I’m all right?” he gasped.
“Of course you’re all right. You told me you wanted a sea voyage,
and naturally I prescribed one for you to give you a chance of
getting your leave without any trouble.”
The patient stared at the doctor for another minute and then fell
back upon his pillow, turned his face to the wall, and wept.
Only for a few minutes, however; then he suddenly sprang from
the bed, caught the doctor by the collar of his coat, looked around
for a weapon of percussion, picked up the pillow and forthwith
began to belabour the physician with such vehemence that the
Khânsâmah, who hurried into the room hearing the noise of the
scuffle, fled from the compound, being certain that the Joy of the
Whole World had become a maniac.
After the lapse of about a minute the doctor was lying on the floor
with the tears of laughter streaming down his cheeks and on to his
disordered shirt-front, while the patient sat limp on a chair yelling
with laughter—a trifle hysterically, perhaps. At the end of five
minutes both were sitting over a bottle of champagne—not too dry—
discussing the extraordinary effect of the imagination upon the
human frame.
“But, by Jingo! I mustn’t forget poor Lois Perez,” cried the doctor,
starting up. “You may guess what a condition he is in when you
know that the letter you read was meant for him.”
“By heavens, I can make a good guess as to his condition,” said
the patient. “I was within measurable distance of that condition half
an hour ago. But I’m hanged if you are going to make any other
poor devil as miserable as you made me. Let the chap die in peace.”
“There’s something in what you say,” said the doctor. “I believe
that I’ll take your advice; only I must rescue your letter from him. If
it were found among his effects after his death next week, I’d be set
down as little better than a fool for writing that he was generally
sound but in need of a long sea voyage.”
He drove off to the house of the Portuguese dealer in precious
stones, and on inquiring for him, learned that he had left in the
afternoon by the mail steamer to take the voyage that the doctor
had recommended. He meant to call at the Andamans, and then go
on to Rangoon, the man in charge of the house said.
“There’ll be an impressive burial service aboard that steamer
before it arrives at the Andaman Islands,” said the doctor to his wife
as he told her what had occurred. The doctor was in a very anxious
state lest the letter which the Portuguese had received should be
found among his papers. His wife, however, took a more optimistic
view of the situation. And she was right; for Lois Perez returned in
due course from Rangoon with a very fine collection of rubies; and
five years afterwards he had still sufficient strength left to get the
better of me in the sale of a cat’s-eye to which he perceived I had
taken a fancy that was not to be controlled.
CHAPTER X—THE VEGETARIAN AND
OTHERS.
“Benjamin’s mess”—An alluring name—Scarcely accurate—A frugal
supper—Why the sub-editor felt rather unwell—“A man should stick
to plain homely fare”—Two Sybarites—The stewed lemon as a
comestible—The midnight apple—The roasted crabs—The Zenana
mission—The pibroch as a musical instrument—A curious blunder—
The river Deccan—Frankenstein as the monster—The outside critics
—A critical position—The curate as critic—A liberal-minded clergyman
—Bound to be a bishop—The joy-bells.

T
O return to the sub-editors and their suppers, I may say that I
never met but one vegetarian pressman. He was particularly
fond of a supper dish to which the alluring name of Benjamin’s
Mess was given by the artful inventor. I do not know if the editor of
this compilation had any authority—Biblical or secular—for assuming
that its ingredients were identical with those with which Joseph, with
the best of intentions, no doubt, but with very questionable
prudence, heaped upon the dish of his youngest brother. I am not a
profound Egyptologist, but I have a distinct recollection of hearing
something about the fleshpots of Egypt, and the longing that the
mere remembrance of these receptacles created in the hearts of the
descendants of Joseph and his Brethren, when undergoing a course
of enforced vegetarianism, though somewhat different in character
from that to which, at a later period, Nebuchadnezzar—the most
distinguished vegetarian that the world has ever known—was
subjected. Therefore, I think it is only scriptural to assume that the
original mess of Benjamin was something like a glorified Irish stew,
or perhaps what yachtsmen call “lobscouce,” and that it contained at
least a neck of mutton and a knuckle of ham—the prohibition did not
exist in those days, and if the stew did not contain either ham or
corned beef it would not be worth eating. But the compilation of
which my friend was accustomed to partake nightly, and to which
the vegetarian cookery book arrogates the patriarchal title, was
wholly devoid of flesh-meat. It consisted, I believe, of some lentils,
parsnips, a turnip, a head of cabbage or so, a dozen of leeks, a
quart of split peas, a few vegetable marrows, a cucumber, a handful
of green gooseberries, and a diseased potato to give the whole a
piquancy that could not be derived from the other simple
ingredients.
I was frequently invited by the sub-editor to join him in his frugal
supper, but invariably declined. I told him that I had no desire to
convert my frame into a costermonger’s barrow.
Upon one occasion the man failed to come down to the office
when he was due. He appeared an hour later, looking very pale. His
features suggested those of an overboiled cauliflower that has not
been sufficiently strained after being removed from the saucepan.
He explained to me the reason of his delay and of his overboiled
appearance.
“The fact is,” said he, “that I did not feel at all well this morning.
For my breakfast I could only eat one covered dishful of
peasepudding, a head or two of celery and a few carrots, with a
tureen of lentil soup and a raw potato salad; so my wife thought she
would tempt me with a delicacy for my dinner. She made me a bran
pie all for myself—thirty-two Spanish onions and four Swedish
turnips, with a beetroot or two for colouring, and a thick paste of
oatmeal and bran—that’s why it’s called a bran pie. Confound the
thing! It’s too fascinating. I can never resist eating it all, and
scraping the stable bucket in which it is cooked. I did so to-day, and
that’s why I’m late. Well, well, perhaps I’ll gain sense late in life. I
don’t feel quite myself even yet. Oh, confound all those dainty
dishes! A man should stick to plain homely fare when he has work to
do.”
But on reflection I think that the most peculiar supper menus of
the sub-editorial staff were those partaken of by two journalists who
occupied the same room for close upon a year—a room to which I
had access occasionally. One of these gentlemen was accustomed to
place in a saucepan on the fire a number of unpeeled lemons with
as much water as just covered them. After four hours’ stewing, this
dainty midnight supper was supposed to be cooked. It certainly was
eaten, and with very few indications, all things considered, of
abhorrence, by the senior occupant of the sub-editor’s room. He told
me once in confidence that he really did not dislike the stewed
lemons very much. He had heard that they were conducive to
longevity, and in order to live long he was prepared to make many
sacrifices. There could be little doubt, he said, that the virtue
attributed to them was real, for he had been partaking of them for
supper for over three years, and he had never suffered from
anything worse than acute dyspepsia. I congratulated him. Nothing
worse than acute dyspepsia!
His stable companion, so to speak, did not believe in heavy hot
suppers such as his colleague indulged in. He said it was his
impression that no more light and salutary supper could be imagined
than a single apple, not quite ripe.
He acted manfully up to his belief, for every night I used to see
him eating his apple shortly after midnight, and without offering the
fruit the indignity of a paring. The spectacle was no more stimulating
than that of the lemon-eater. My mouth invariably became so
puckered up through watching the midnight banquets of these
Sybarites, it was only with difficulty that I could utter a word or two
of weak acquiescence in their views on a question of recognised
difficulty.
It is somewhat remarkable that the apple-eating sub-editor should
be the one who was guilty of the most remarkable error I ever knew
in connection with an attempted display of erudition. He had set out
to write a lively little quarter-of-a-column leaderette on a topic which
was convulsing society in those days—namely, the cruelty of boiling
lobsters alive. I am not quite certain that the question has even yet
been decided to the satisfaction either of the humanitarian who likes
lobster salad, or of the lobster that finds itself potted. Perhaps the
latter may some day come out of its shell and give us its views on
the question.
At any rate, in the year of which I write, the topic was almost a
burning one: the month was September, Parliament had risen, and
as yet the sea-serpent had not appeared on the horizon. The apple-
eating sub-editor was doing duty for the assistant-editor, who was
on his holidays; and as evidence of his light and graceful erudition,
he asserted in his article that, however inhuman modern cooks
might be in their preparation of Crustacea for the fastidious palates
of their patrons, quite as great cruelty—assuming that it was cruelty
—was in the habit of being perpetrated in cookery in the days of
Shakespeare. “Readers of the immortal bard of Avon,” he wrote, “will
recollect how, in one of the charming lyrics to ‘Love’s Labour’s Lost,’
among the homely pleasures of winter it is stated that ‘roasted crabs
hiss in the bowl.’
“This reference to the preparation of crabs for the table makes it
perfectly plain that it was quite common to cook them alive, for were
it otherwise, how could they hiss? That listening to the expression of
the suffering of the crabs should be regarded by Shakespeare as one
of the joys of a household, casts a somewhat lurid light upon the
condition of English Society in the sixteenth century.”

It was the lemon-eating sub-editor who, on being requested by

the editor to write something about the Zenana Mission, pointing out
the great good that it was achieving, and the necessity there was for
maintaining it in an efficient condition, produced a neat little article
on the subject. He assured the readers of the paper that, among the
many scenes of missionary labour, none had of late attracted more
attention than the Zenana mission, and assuredly none was more
deserving of this attention. Comparatively few years had passed
since Zenana had been opened up to British trade, but already,
owing to the devotion of a handful of men and women, the nature of
the inhabitants had been almost entirely changed. The Zenanese,
from being a savage people, had become, in a wonderfully short
space of time, practically civilised; and recent travellers to Zenana
had returned with the most glowing accounts of the continued
progress of the good work in that country. The writer of the article
then branched off into the “labourer-worthy-of-his-hire” side of this
great evangelisation question—in most questions of missionary
enterprise this side has a special interest attached to it—and the
question was aptly asked if the devoted labourers in that remote
vineyard were not deserving of support. Were civilisation and
Christianity to be snatched from the Zenanese just when both were
within their grasp? So on for nearly half a column the writer
meandered in the most orthodox style, just as he had done scores of
times before when advocating certain missions.
I found him the next day running his finger down the letter Z, in
the index to the Handy Atlas, with a puzzled look upon his face. I
knew then that he had received a letter from the editor, advising him
to look out Zenana in the Atlas before writing anything further about
so ticklish a region.

I also knew a sub-editor who fancied that the pibroch was a

musical instrument widely circulated in the Highlands.
But who can blame a humble provincial journalist for making an
odd blunder occasionally, when a leading London newspaper, in
announcing the death, some years ago, of Captain Wallace, son of
Sir Richard Wallace, stated that the sad event had occurred while he
was “playing at bagatelle in the Bois de Boulogne”? It might
reasonably have been expected, I think, that the sub-editor of the
foreign news should know of the existence of the historic mansion
Bagatelle, which the Marquis of Hertford left to Sir Richard Wallace
with the store of art treasures that it contained.
What excuse, one may also ask, can be made for the Dublin
Professor who referred in print “to those populous districts of
Hindostan, watered by the Ganges and the Deccan”?

In alluding to Frankenstein as the monster, and not merely the

maker of the monster, the mistakes made by provincial journalists of
the old school may certainly also be condoned, when we find the
same ridiculous hallucination maintained by one of the most highly
representative of modern journalists, as-well as by the editor of a
weekly paper of large circulation, who enshrined it in the preface to
a book for which he was responsible. In this case the writer could
not have been pressed for time. But the marvel is, not that so many
errors are run into by provincial journalists, but that so few can be
laid to their charge. With telegrams pouring in by private wire, as
well as by the P.A. and C.N., to say nothing of Baron Reuter’s and
Messrs, Dalziel’s special services; with the foreman printer, too,
appearing like a silent spectre and departing like one that is not
silent, leaving the impression behind him that no newspaper, except
that composed by a hated rival, can possibly be produced the next
morning;—with all these drags upon the chariot wheels of
composition, how can it be reasonably expected that an editor or a
sub-editor will become Academic in his erudition? When, however, it
is discovered the next day by some tenth-rate curate, who probably
gets a free copy of the paper, that the quotation “O tempora! O
mores!” is attributed to Virgil instead of Cicero, in a leading article a
column in length, written upon a speech of seven columns, the
writer is at once referred to as an ignorant boor, and an invitation is
given to all that curate’s friends to point the finger of scorn at the
journalist.
A long experience has convinced me that the curate who gets a
free copy of the paper, and who is most velvet-gloved in approaching
any member of the staff when he wants a favour, such as a
leaderette on the Zenana Mission, in which several of his lady friends
are deeply interested, or a paragraph regarding a forthcoming
bazaar, or the insertion of a letter signed “Churchman,” calling
attention to some imaginary reform which he himself has instituted
—this very curate is the person who sends the marked copies of the
paper to the proprietor with a gigantic Sic opposite every mistake,
even though it be only a turned letter.
I put a stop to the tricks of one of the race who had annoyed me
excessively. I simply inserted verbatim a long letter that he wrote on
some subject. It was full of mistakes, and to these the next day, in a
letter which he meant to be humorous, he referred as “printer’s
errors.” I took the liberty of appending an editorial note to this
communication, mentioning that the mistakes existed in the original
letter, and adding that I trusted the writer would not think it
necessary to attribute to the printer the further blunders which
appeared in the humorous communication to which my note was
appended.
The fellow sought an interview with me the next day, and found it.
He was furiously indignant at the course which I had adopted, and
said I had taken advantage of the haste in which he had written
both letters. I brought out of my desk forthwith a paper which he
had taken the trouble to re-edit with red ink for the benefit of the
proprietor, who had, naturally, handed it to me. I recognised the
handwriting of the red-ink editor the moment I received the first of
his letters.
“Did you make any allowance for the haste of the writers of these
passages that you took the trouble to mark and send to the
proprietor?” I inquired blandly.
He said he did not know what it was that I referred to; and added
that it was a gratuitous assumption on my part to say that he had
marked and sent the paper.
“Very well,” said I. “I’ll assume that you deny having done so. May
I do so?”
“Certainly you may,” he replied. “I have something else to do
beside pointing out the blunders of your staff.”
“Then I ask your pardon for having assumed that you marked the
paper,” said I. “I was too hasty.”
 “You were—quite too hasty,” said he, going to the door.
“I’ve acknowledged it,” said I. “And therefore I’ll not go to your
rector until to-morrow evening to prove to him that his curate is a
sneak and a liar as well as an extremely ignorant person.”
He returned as I sat down.
“What paper is it that you allude to?” he asked.
“I showed it to you,” said I. “It was the paper that you re-edited in
red ink and posted anonymously to the proprietor.”
“Oh, that?” said he. “Why on earth didn’t you say so at once? Of
course I sent that paper. My dear fellow, it was only my little joke. I
meant to have a little chaff with you about the mistakes.”
“Go away—go away,” said I. “Go away, Stiggins.”
And he went away.

I need scarcely say that such clergymen are not to be interviewed

every day. Equally exceptional, I think, was the clergyman who was
good enough to pay me a visit a few months after I had joined the
editorial staff of a daily paper. Although I had never exactly been the
leader of the coughers in church, yet on the other hand I had never
been a leader of the scoffers outside it; and somehow the parson
had come to miss me. I had an uneasy feeling when he entered my
room that he had come on business—that he might possibly have
fancied I was afflicted with doubts on, say, the right of unbaptised
infants to burial in consecrated ground, and that he had come
prepared to lift the burden from my soul; but he never so much as
spoke of business until he had picked up his hat and gloves, and had
said a cheerful farewell. Only then he remarked, as if the thing had
occurred to him quite suddenly,—
“Oh, by the way, I don’t think I noticed you in church during the
past few Sundays. I was afraid that you were indisposed.”
 “Oh, no,” said I. “I was all right; but the fact is, you see, that I’ve
become a sort of editor, and as I can never get to bed before three
or four in the morning, it would be impossible for me to rise before
eleven. To be sure I’m not on duty on Saturday nights, but the force
of habit is so great that, though I may go to bed in decent time on
that night, I cannot sleep until my usual hour.”
“Oh, I see, I see,” said he, beginning to draw on his gloves. “Well,
perhaps on the whole—all things considered—the—ah—” here he
was seized with a fit of coughing, and when he recovered he said he
had always been an admirer of old Worcester, and he rather thought
that some cups which I had on a shelf were, on the whole, the most
characteristic as regards shape that he had ever seen.
Then he went away, and I perceived from the appearance that his
back presented to me, that he would one day become a bishop. A
clergyman with such tact as he exhibited can no more avoid being
made a bishop than the young seal can avoid taking to the water.
Before five years had passed he was, sure enough, raised to the
Bench, and every one is delighted with him. The celery from the
Palace garden invariably takes the first prize at the local shows; his
lordship smiles when you congratulate him on his repeated
successes with celery, but when you talk about chrysanthemums he
becomes grave and shakes his head.
This is his tact.

The church of which he was rector was situated in a fashionable

suburb of the town, and it possessed one of the noisiest peals of
bells possible to imagine. They were the terror of the
neighbourhood.
Upon one occasion an elderly gentleman living close to the church
contracted some malady which necessitated, the doctor said, the
observance of the strictest quiet, even on Sundays. A message was
sent to the chief of the bellringers to this effect, the invalid’s wife
expressing the hope that for a Sunday or two the bells might be
permitted to remain silent. Of course her very reasonable wish was
granted. The chief of the ringers thoughtfully called every Sunday
morning to inquire after the sufferer’s condition, and for three weeks
he learned that it was unchanged, and the bells consequently
remained silent. On the fourth Sunday, he was told that the man had
died during the night. He immediately hastened off to the other
seven bellringers, worse than the first, and telling them that their
prohibition was removed, they climbed the belfry and rang forth the
most joyous peal that had ever annoyed the neighbourhood.
“Ah,” said the lady with whom I lodged, “there are the joy bells
once more. Poor Mr. Jenkins must be dead at last.”
 CHAPTER XI.—ON SOME FORMS OF
SPORT.
An invitation to shoot rooks—The sub-editors gun—A quotation
from “The Rivals”—The rook in repose—How the gun came to be
smashed—Recollections of the Spanish Main—A greatly overrated
sport—The story of Jack Burnaby’s dogs—A fastidious man—His
keeper’s remonstrance—The Australian visitor—-A kind offer—Over-
willing dogs—The story of a muzzle-loader—How Mr. Egan came to
be alive—Why Patsy Muldoon smiled—The moral—Degrees of
dampness—Below the surface—The chameleon blackberry—A
superlative degree of thirst.

A
FRIEND of mine once came to my office to invite me to an
afternoon’s rook-shooting. I was not in my room and he found
me in the sub-editor’s. I inquired about the trains to the place
where the slaughter was to be done, and finding that they were
satisfactory, agreed to join him on the following afternoon.
Then he turned to the sub-editor—a pleasant young fellow who
had ideas of going to the bar—and asked him if he would care to
come also. At first the sub-editor said he did not think he would be
able to come, though he would like very much to do so. A little
persuasion was sufficient to make him agree to be one of our party.
He had not a gun of his own, he said, but a friend had frequently
offered to lend him one, so that there would be no difficulty so far as
that matter was concerned.
The next day I managed, as usual, just to catch the train as it
began to move-away from the platform. My colleague on the
newspaper had the door of the compartment open for me, and I
could see the leather of his gun-case under the seat. I put my rook
rifle—it was not in a case—in the network, and we had a delightful
run through the autumn landscape to the station—it seemed miles
from any village—where my friend was awaiting us in his dogcart,
driving tandem. The drive of three miles to the rook-wood was
exhilarating, and as we skirted some lines of old gnarled oaks, I
perceived in a moment that we could easily fill a railway truck with
birds, they were so plentiful. I made a remark to this effect to my
friend, who was driving, and he said that when we arrived at the
shooting ground and gave the birds the chance to which they were
entitled we mightn’t get more than a couple of hundred all told.
The shooting ground was under a straggling tree about fifty yards
from the ruin of an old castle, said to have been built by the Knights
Templar. Here we dismounted from the dogcart, sending it a mile or
two farther along the road in charge of the man, and got ready our
rifles.
“What on earth have you got there?” my friend inquired of the
sub-editor, who was working at the gun-case.
“It’s the gun and cartridges,” replied the young man; “but I’m not
quite certain how to make fast the barrels to the stock.”
“Great heavens!” cried my friend. “You’ve brought a double-
barrelled sporting gun to shoot rooks!”
And so he had.
We tried to explain to him that for any human being to point such
a weapon at a rook would be little short of murder, but he utterly
failed to see the force of our arguments. He very good-humouredly
said that, as we had come out to shoot rooks, he couldn’t see how it
mattered—especially to the rooks—whether they were shot with his
gun or with our rook rifles. He added that he thought the majority of
the birds were like Bob Acres, and would as lief be shot in an
ungentlemanly as a gentlemanly attitude.
Of course it is impossible to argue with such a man. We only said
that he must accept the responsibility for the butchery, and in this he
cheerfully acquiesced, slipping cartridges into both barrels—the
friend from whom he had borrowed the weapon had taught him how
to do this.
 We soon found that at this point the breaking-strain of his
information was reached. He had no more idea of sport than a
butcher, or the Sonttag jager of the Oberlander Blatter.
As the rooks flew from the ruins to the belt of trees my friend and
I brought down one each, and by the time we had reloaded, we
were ready for two more, but I fired too soon, so that only one bird
dropped. I saw the eyes of the man with the shot-gun gleam, “his
heart with lust of slaying strong,” and he forthwith fired first one
barrel and then the other at an old rook that cursed us by his gods,
sitting on a branch of a tree ten yards off.
The bird flapped heavily away, becoming more vituperative every
moment.
“Look here,” I shouted, “you mustn’t shoot at a bird that’s sitting
on a branch.”
“Oh. yes,” said my friend, with a grim smile. “Oh, yes, he may. It’ll
do him no more harm than the birds.”
Not a bird did that young sportsman fire at except such as had
assumed a sitting posture, and, incredible though it may seem, he
only succeeded in killing one. But from the moment that his skill was
rewarded by witnessing the downward flap of this one, the lust for
blood seemed to take possession of him, as it does the young
soldiers when their officers have succeeded in preventing them from
blazing away at the enemy while still a mile off. He continued to load
and fire at birds that were swaying on the trees beside us.
“There’s a chance for you,” said my friend, “sarkastik-like,”
pointing to a rook that had flapped into a branch just above our
heads.
The young man, his face pale and his teeth set, was in no mood
for distinguishing between one tone of voice and another. He simply
took half a dozen steps into the open and, aiming steadily at the
bird, fired both barrels simultaneously. Down came the rook in the
usual way, clawing from branch to branch. It remained, however, for
several seconds on a bough about eight feet from the ground; then
we had a vision of the sportsman clubbing his gun, and making a
wild rush at his prey—and then came a crash and a cheer. The
sportsman held aloft in one hand the tattered rook and in the other
a double-barrelled gun with a broken stock.
He had never fired a shot in his life before this day, and all his
ideas of musketry were derived from the stories of pirates and
buccaneers of the Spanish Main—wherever that may be—which had
come to him for review. He thought that the clubbing of his weapon,
in order to prevent the escape of the rook, quite a brilliant thing to
do.
He had, however, completely smashed the gun, and that, my
friend said, was a step in the right direction. He could not do any
more butchery with it that day.
It cost him four pounds getting that gun repaired, and he
confessed to me that, according to his experience, fowling was a
greatly overrated sport.

It was while we were driving to the train that my friend told me

the story of Jack Burnaby’s dogs—a story which he frankly confessed
he had never yet got any human being to believe, but which was
accurate in all its details, and could be fully verified by affidavit. He
did not succeed in obtaining my credence for it. There are other
forms of falsehood besides those verified by an affidavit, and I could
not have given more implicit disbelief than I did to the story, even if
it had formed the subject of this legal method of embodying a
fiction.
It appeared that never was there a more fastidious man in the
matter of his sporting dogs than one Algy Grafton. Pointers that
called for outbursts of enthusiasm on the part of other men—quite
as good sportsmen as Algy—failed to obtain more than a
complimentary word from him, and even this word of praise was
grudgingly given and invariably tempered by many words which
were certainly not susceptible of a eulogistic meaning.
 Among his friends—such as declined to resent the insults which he
put upon their dogs—there was a consensus of opinion that the
animal which would satisfy him would not be born—allowing a
reasonable time for the various processes of evolution—for at least a
thousand years, and then, taking into consideration the growth of
radical ideas, and the decay of the English sport, there would be
little or no demand for a first-class dog in the British Islands.
Algy Grafton had just acquired the Puttick-Foozler moor, and
almost every post brought him a letter from his head-keeper
describing the condition of the birds and the prospects of the
Twelfth. Though the letters were written on a phonetic principle, the
correctness of which was, of course, proportionate to the accuracy
of a Scotchman’s ear, and though the head-keeper was scarcely an
optimist, still there was no mistaking the general tone of the
information which Algy received through this source from the north:
he gathered that he might reasonably look forward to the finest
shoot on record.
Every letter which he got from the moor, however, contained the
expression of the keeper’s hope that his master would succeed in his
search for a couple of good dogs. The keeper’s hope was shared by
Algy; and he did little else during the month of July except interview
dogs that had been recommended to him. He travelled north and
south, east and west, to interview dogs; but so ridiculously fastidious
was he that at the close of the first week in August he was still
without a dog. He was naturally at his wit’s end by this time, for as
the Twelfth approached there was not a dog in the market. He
telegraphed in all directions in the endeavour to secure some of the
animals which he had rejected during the previous month, but, as
might have been expected, the dogs were no longer to be disposed
of: they had all been sold within a day or two after their rejection by
Mr. Grafton. It was on the seventh of August that he got a letter
from his correspondent on the moor, and in this letter the tone of
mild remonstrance which the keeper had hitherto adopted in
referring to his master’s extravagant ideas on the dog question, was
abandoned in favour of one of stern reprimand; in fact, some
sentences were almost abusive. Mr. Donald MacKilloch professed to
be anxious to know what was the good of his wearing out his life on
the moor if his master did not mean to shoot on it. He hoped he
would not be thought wanting in respect if he doubted the sanity of
the policy of waiting without a dog until it pleased Providence—Mr.
MacKilloch was a very religious man—to turn angels into pointers
and saints into setters, a period which, it seemed to Mr. MacKilloch,
his master was rather oversanguine in anticipating.
It was not surprising that, after receiving this letter from the
Highlands, Algy Grafton was somewhat moody as he strolled about
his grounds on the morning of the eighth, nor was it remarkable
that, when the rectory boy appeared with a letter stating that the
Reverend Septimus Burnaby was anxious for him to run across in
time to lunch at the rectory, to meet Jack Burnaby, who had just
returned from Australia, Algy said that the rector and his brother
Jack and all the squatters in the Australian colonies might be hanged
together. Mrs. Grafton, however, whose life had not been worth a
month’s purchase since the dog problem had presented itself for
solution, insisted on his going to the rectory to lunch, and he went.
It was while smoking a cigar in the rectory garden with Jack
Burnaby, who had spent all his life squatting, but with no apparent
inconvenience to himself, that Algy mentioned that he was broken-
hearted on account of his dogs. He gave a brief summary of his
travels through England in search of trustworthy animals, and
lamented his failure to obtain anything that could be depended on to
do a day’s work.
“By George! you don’t mean to say there’s not a good dog in the
market now?” said Mr. Burnaby, the squatter.
“But that’s just what I do mean to say,” cried Algy, so plaintively
that even the stern and unbending MacKilloch might have pitied him.
“That’s just what I do mean to say. I’d give fifty pounds to-day for a
pair of dogs that I wouldn’t have given ten pounds for a month ago.
I’m heart-broken—that’s what I am!”
“Cheer up!” said Mr. Burnaby. “I have a couple of sporting dogs
that I’ll lend to you until I return to the Colony in February next—the
best dogs I ever worked with, and I’ve had some experience.”
“It was Providence that caused you to come across to me to-day,
Grafton,” said the rector piously, as Algy stood speechless among the
trim rosebeds.
“You’re sure they’re good?” said Algy, his old suspicions returning.
“Good?—am I sure?—oh, you needn’t have them if you don’t like,”
said the Australian.
“I beg your pardon a thousand times,” cried Algy. “Don’t fancy that
I suggest that the dogs are not first rate. Oh, my dear fellow, I don’t
know how to thank you. I am—well, my heart is too full for words.”
“There’s not a man in England except yourself that I’d lend them
to,” said Mr. Burnaby. “I give you my word that I’ve been offered
forty pounds for each of them. Oh, there isn’t a fault between them.
They’re just perfect.”
Algy was delighted, and for the remainder of the evening he kept
assuring his poor wife that he was not quite such a fool as some
people, including the Scotch keeper, seemed to fancy that he was.
He had felt all along, he said, that just such a piece of luck as had
occurred was in store for him, and it was on this account he had
steadily refused to be gulled into buying any of the inferior animals
that had been offered to him.
Oh, yes, he assured her, he knew what he was about, and he’d let
MacKilloch know who it was that he had to deal with.
The Australian’s dogs were in the custody of a man at
Southampton, but he promised to have them sent northward in good
time. It was the evening of the eleventh when they arrived at the
lodge. They were strange wiry brutes, and like no breed that Algy
had ever seen. The head-keeper looked at them critically, and made
some observations regarding them that did not seem grossly
flattering. It was plain that if Mr. MacKilloch had conceived any
sudden admiration for the dogs he contrived to conceal it. Algy said
all that he could say, which was that Mr. Burnaby knew perfectly well
what a dog was, and that a dog should be proved before it was
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebooknice.com