The number of innovative applications of randomization tests in various fields and recent developments in experimental design, significance testing, computing facilities, and randomization test algorithms have necessitated a new edition of Randomization Tests.
Updated, reorganized, and revised, the text emphasizes the irrelevance and implausibility of the random sampling assumption for the typical experiment in three completely rewritten chapters. It also discusses factorial designs and interactions and combines repeated-measures and randomized block designs in one chapter. The authors focus more attention on the practicality of N-of-1 randomization tests and the availability of user-friendly software to perform them. In addition, they provide an overview of free and commercial computer programs for all of the tests presented in the book.
Building on the previous editions that have served as standard textbooks for more than twenty-five years, Randomization Tests, Fourth Edition includes a CD-ROM of up-to-date randomization test programs that facilitate application of the tests to experimental data. This CD-ROM enables students to work out problems that have been added to the chapters and helps professors teach the basics of randomization tests and devise tasks for assignments and examinations.
Table of Contents
Statistical Tests That Do Not Require Random Sampling. Randomized Experiments. Calculating P-Values. Between-Subjects Designs. Factorial Designs. Repeated-Measures and Randomized Block Designs. Multivariate Designs. Correlation. Trend Tests. Matching and Proximity Experiments. N-of-1 Designs. Tests of Quantitative Laws. Tests of Direction and Magnitude of Effect. Fundamentals of Validity. General Guidelines and Software Availability.
Eugene Edgington, Patrick Onghena
“…Overall, this is an interesting and well-written book that provides a useful discussion of the theory, design, and application of randomization tests, illustrated with appropriate examples using experimental data. The end-of-chapter questions and exercises make it useful also as a textbook for college students. It should be of interest for every experimenter who is interested in randomization or permutation tests or is skeptical about the reliability of the assumptions of parametric tests.”
—Andreas Karlsson (Uppsala University), Journal of the Royal Statistical Society