Distribution-free Tests
Published in January 1992
Book
Paperback/Softback
448 pages
978-0-415-08418-5 (ISBN)
Description
This text, aimed at undergraduate and postgraduate students, looks at a selection of distribution-free methods and tests. The authors have attempted to explain the techniques carefully and to give numerical examples for each. Each chapter deals with a specific topic. For most of these topics two tests are presented: a relatively sophisticated one with power properties, and a simple one where ease of execution is the main priority. Apart from the chi-squared and sign tests which date from the beginning of the century, the origins of the techniques lie within the period from the late 1930s to the present day. The tests have been chosen on the basis of usefulness and ease of application. Also covered are some of the important advances which have been made since the mid-1950s. Introductory chapters examine the necessary statistical background and general ideas about hypotheses-testing. The book may be of interest not only to statistics students but also those who use statistical techniques covering a wider span.
More details
Edition
New edition
Language
English
Place of publication
London
United Kingdom
Publishing group
Taylor & Francis Ltd
Target group
College/higher education
Edition type
New edition
Illustrations
line illustrations, tables, bibliography, index
Dimensions
Height: 234 mm
Width: 156 mm
Weight
700 gr
ISBN-13
978-0-415-08418-5 (9780415084185)
Copyright in bibliographic data is held by Nielsen Book Services Limited or its licensors: all rights reserved.
Schweitzer Classification
Content
Part 1 Essential background: some statistical ideas; the underlying ideas of statistical tests. Part 2 Tests involving one sample: one-sample location tests - the Wilcoxon signed-rank test; goodness-of-fit tests - chi-squared goodness-of-fit test, Kolmogorov-Smirnov test, Lilliefors' test for normality. Part 3 Tests involving two samples: tests for differences in location - the Mann-Whitney test, Tukey's quick test; tests for differences in spread - the Siegel-Tukey test; tests for general differences - Rosenbaum's test; two related samples. Part 4 Correlation and regression: correlation - Spearman's and Kendall's rank correlation coefficient; regression - Theil's "incomplete" method, Iman-Conover method for monotonic regression; quick methods for correlation and regression - the Olmstead-Tukey "corner" test, Brown and Mood's quick regression method; association between two variables - Yates' correction. Part 5 Tests involving more than two samples: more than two (k) independent samples - Kruskal-Wallis test; more than two (k) related samples - Friedman's test; more than two (k) samples - ordered alternative hypothesis - the Terpstra-Jonckheere test, Page's test; tests for interaction and Latin squares designs. Part 6 Miscellaneous tests: tests based on runs - the Wald-Wolfowitz two-sample runs test; more tests using frequency data - Fisher's exact test, the Mood-Westenberg two-sample dispersion test; confidence intervals.