Multivariate Permutation Tests: With Applications in Biostatistics

ISBN-13: 9780471496700 / Angielski / Twarda / 2001 / 440 str.

Fortunato Pesarin;Pesarin

Opis

Complex multivariate problems are frequently encountered in many scientific disciplines and it can be very difficult to obtain meaningful results. Permutation and nonparametric combination methods provide flexible solutions to complex problems by reducing the problem down to a set of simpler sub-problems. The author presents a novel but well tested approach using real examples taken from biomedical research. Statistical analyses are performed in a nonparametric setting, so that no assumptions need be made about the underlying distribution and the dependence relations between variables.

* Provides a clear exposition of the use of multivariate permutation testing, with emphasis on the use of nonparametric combination methodology.
* Growing area of research with many practical applications, notably in biostatistics.
* Numerous case studies and examples help to illustrate the theory.
* Provides solutions to multi-aspect problems, to problems with missing data, analysis of factorial designs and repeated measures.
* Explains the analysis of categorical, ordered categorical, binary, continuous, and mixed variables in both an experimental and an observational context.
* NPC-Test(c) software (demo copy), SAS macros, S-Plus code and datasets are available on the Web at
http: //www.stat.unipd.it/ pesarin/
For researchers and practitioners in a number of scientific disciplines, particularly biostatistics, the vast collection of techniques, examples and case studies will be an invaluable resource. Graduate students of applied statistics and nonparametric methods will find the book provides an accessible introduction to multivariate permutation testing.

Recenzje

"the book is well written. It cand be useful and recommended for researchers and practitioners in a number of scientific disciplines...and for graduate students..." (Zentralblatt MATH, Vol.972, No.12, 2001)

"...carefully presents a concise and mathematically rigorous treatment of permutation testing...could be used for a mathematically oriented graduate class...will form a source of recent reference material for research workers..." (Short Book Reviews, Vol. 22, No. 1, April 2002)

"This book may herald a new era in biostatistics..." (Psychotherpay and Psychosomatics, September/October 2002) 

"...graduate and post–graduate students in some areas of physics and chemistry can benefit greatly from reading and using this book..." (The Statistician, Date Unknown)

Spis treści

Preface.

Notation and Abbreviations.

Introduction.

Discussion of a Simple Testing Problem.

Theory of Permutation Tests for One–Sample Problems.

Examples of Univariate Multi–Sample Problems.

Theory of Permutation Tests for Multi–Sample Problems.

Nonparametric Combination Methodology.

Examples of Nonparametric Combination.

Permutation Analysis in Factorial Designs.

Permutation Testing with Missing Data.

The Behrens––Fisher Permutation Problem.

Permutation Testing for Repeated Measurements.

Further Applications.

References.

Index.

Blurb

Complex multivariate problems are frequently encountered in many scientific disciplines and it can be very difficult to obtain meaningful results. Permutation and nonparametric combination methods provide flexible solutions to complex problems by reducing the problem down to a set of simpler sub–problems. The author presents a novel but well tested approach using real examples taken from biomedical research. Statistical analyses are performed in a nonparametric setting, so that no assumptions need be made about the underlying distribution and the dependence relations between variables.
∗ Provides a clear exposition of the use of multivariate permutation testing, with emphasis on the use of nonparametric combination methodology.
∗ Explores a growing area of research with many practical applications, notably in biostatistics.
∗ Numerous case studies and examples help to illustrate the theory.
∗ Provides solutions to multi–aspect problems, to problems with missing data, analysis of factorial designs and repeated measures.
∗ Explains the analysis of categorical, ordered categorical, binary, continuous, and mixed variables in both an experimental and an observational context.
∗ NPC–Test software, SAS macros, S–Plus code, and datasets are available on the Web.
For researchers and practitioners in a number of scientific disciplines, particularly biostatistics, the vast collection of techniques, examples and case studies will be an invaluable resource. Graduate students of applied statistics and nonparametric methods will find the book provides an accessible introduction to multivariate permutation testing.