4th Edition

# Applied Nonparametric Statistical Methods

542 Pages 35 B/W Illustrations
by Chapman & Hall

544 Pages
by Chapman & Hall

Also available as eBook on:

While preserving the clear, accessible style of previous editions, Applied Nonparametric Statistical Methods, Fourth Edition reflects the latest developments in computer-intensive methods that deal with intractable analytical problems and unwieldy data sets.

Reorganized and with additional material, this edition begins with a brief summary of some relevant general statistical concepts and an introduction to basic ideas of nonparametric or distribution-free methods. Designed experiments, including those with factorial treatment structures, are now the focus of an entire chapter. The text also expands coverage on the analysis of survival data and the bootstrap method. The new final chapter focuses on important modern developments, such as large sample methods and computer-intensive applications.

Keeping mathematics to a minimum, this text introduces nonparametric methods to undergraduate students who are taking either mainstream statistics courses or statistics courses within other disciplines. By giving the proper attention to data collection and the interpretation of analyses, it provides a full introduction to nonparametric methods.

PREFACE

SOME BASIC CONCEPTS
Basic Statistics
Populations and Samples
Hypothesis Testing
Estimation
Ethical Issues

FUNDAMENTALS OF NONPARAMETRIC METHODS
A Permutation Test
Binomial Tests
Order Statistics and Ranks
Exploring Data
Efficiency of Nonparametric Procedures
Computers and Nonparametric Methods

LOCATION INFERENCE FOR SINGLE SAMPLES
Layout of Examples
Continuous Data Samples
Inferences about Medians Based on Ranks
The Sign Test
Use of Alternative Scores
Comparing Tests and Robustness
Fields of Application
Summary

OTHER SINGLE-SAMPLE INFERENCES
Other Data Characteristics
Matching Samples to Distributions
Inferences for Dichotomous Data
Tests Related to the Sign Test
A Runs Test for Randomness
Angular Data
Fields of Application
Summary

METHODS FOR PAIRED SAMPLES
Comparisons in Pairs
A Less Obvious Use of the Sign Test
Power and Sample Size
Fields of Application
Summary

METHODS FOR TWO INDEPENDENT SAMPLES
Centrality Tests and Estimates
The Median Test
Normal Scores
Tests for Equality of Variance
Tests for a Common Distribution
Power and Sample Size
Fields of Application
Summary

BASIC TESTS FOR THREE OR MORE SAMPLES
Comparisons with Parametric Methods
Centrality Tests for Independent Samples
The Friedman Quade and Page Tests
Binary Responses
Tests for Heterogeneity of Variance
Some Miscellaneous Considerations
Fields of Application
Summary

ANALYSIS OF STRUCTURED DATA
Factorial Treatment Structures
Balanced 2 × 2 Factorial Structures
The Nature of Interactions
Alternative Approaches to Interactions
Cross-Over Experiments
Specific and Multiple Comparison Tests
Fields of Application
Summary
Exercises

ANALYSIS OF SURVIVAL DATA
Special Features of Survival Data
Modified Wilcoxon Tests
Savage Scores and the Log-Rank Transformation
Median Tests for Sequential Data
Choice of Tests
Fields of Application
Summary

CORRELATION AND CONCORDANCE
Correlation in Bivariate Data
Ranked Data for Several Variables
Agreement
Fields of Application
Summary

BIVARIATE LINEAR REGRESSION
Fitting Straight Lines
Fields of Application
Summary

CATEGORICAL DATA
Categories and Counts
Nominal Attribute Categories
Ordered Categorical Data
Goodness-of-fit Tests for Discrete Data
Extension of McNemar's Test
Fields of application
Summary

ASSOCIATION IN CATEGORICAL DATA
The Analysis of Association
Some Models for Contingency Tables
Combining and Partitioning of Tables
A Legal Dilemma
Power
Fields of Application
Summary

ROBUST ESTIMATION
When Assumptions Break Down
Outliers and Influence
The Bootstrap
M-estimators and Other Robust Estimators
Fields of Application
Summary

MODERN NONPARAMETRICS
A Change in Emphasis
Density Estimation
Regression
Logistic Regression
Multivariate Data
New Methods for Large Data Sets
Correlations within Clusters
Summary

Exercises appear in each chapter.

APPENDIX 1
APPENDIX 2
REFERENCES
INDEX

### Biography

Peter Sprent, Nigel C. Smeeton

… The greatest strength of this book is that it is written at a level that is perfectly understandable by readers with only a course or two of introductory-level statistics. As such, it is appropriate for use as either a textbook for a first course in nonparametric methods for undergraduate statistics majors or as a reference for practitioners in other fields. It is also quite suitable as a supplementary statistics textbook for graduate students … . Key concepts are taught using worked-out examples from a variety of fields. … a worthwhile choice for either an introductory-level textbook or a self-study reference for nonspecialists. The writing is very accessible and not weighted down by any mathematics beyond the grasp of the intended audience. …
Psychometrika, Vol. 75, No. 3, September 2010

… this book has an effective organization and covers a wider scope of non-parametric methods than former editions. Therefore, I believe that this book can serve its intended audience.
Journal of the Royal Statistical Society, Series A, Vol. 173, Issue 1, January 2010

Most fourth editions look surprisingly similar to the third editions. Applied Nonparametric Statistical Methods is an exception. Sprent and Smeeton have taken an accessible and well-regarded work and expanded, reorganized, and improved on it. … Sprent and Smeeton offer a strong connection with respect to the how and why of the techniques. … The book’s major strength is its prioritization of coverage. The authors take painstaking care to inculcate an understanding of the appropriate use of nonparametric methods, as well as an appreciation for their application over a wide range of fields. The examples are well chosen, and the variety should ensure that every reader finds at least some of the problems interesting. … As a competitor to the texts by Conover (1999), Gibbons and Chakraborti (2004), Higgins (2004), and Wasserman (2006), Applied Nonparametric Statistical Methods more than holds its own. The combination of clear writing and comprehensive coverage make it an excellent introductory text. …
Technometrics, Vol. 51, No. 2, May 2009

…The chapters have been substantially reorganized, and new material is provided on methods related to factorial designs and time-to-event data. An entirely new chapter, ‘Modern Nonparametrics,’ closes the text with a variety of topics … the worked examples are thoroughly and meticulously done … constant mention is made of the available software (e.g., StatXact, R, Minitab, SPSS) to conduct specific procedures. … solutions to selected end-of-chapter exercises are annotated and quite helpful. Overall, this is a solid choice for a first course in nonparametric statistics for undergraduates.
Journal of the American Statistical Association, Vol. 104, No. 487, September 2009

… expands coverage on the analysis of survival data and the bootstrap method. … the new edition also focuses on some modern developments. The formal testing procedures are illustrated in a nice way with realistic examples leading to final conclusions, comments, and a discussion… The book has a clear style with well-organized material. The book works well as a reference book for users of nonparametric methods in different research areas. It is also a good textbook for undergraduate courses in statistics as well as courses for students majoring in other disciplines.
—Hannu Oja, International Statistical Review, Vol. 27, No. 1, 2008

Praise for the Third Edition
Strengths of this text certainly include its organization and writing style. Applied Nonparametric Statistical Methods provides a very clear exposition of modern nonparametric methods. Many students and practitioners will find it an excellent resource and reference for nonparametric statistics.
—Technometrics, 2003

… extremely valuable for statisticians as well as for researchers in applied fields. … This well-written book is highly recommended for those readers who want to get a feeling for the nonparametric methods which they apply when analysing their data.
Statistics in Medicine, 2004