Nonparametric Statistics : An Introduction

Through the use of actual research investigations that have appeared in recent social science journals, Gibbons shows the reader the specific methodology and logical rationale for many of the best-known and most frequently used nonparametric methods that are applicable to most small and large sample sizes. The methods are organized according to the type of sample structure that produced the data to be analyzed, and the inference types covered are limited to location tests, such as the sign test, the Mann-Whitney-Wilcoxon test, the Kruskal-Wallis test and Friedman′s test. The formal introduction of each test is followed by a data example, calculated first by hand and then by computer.

"Nonparametric Statistics is a short and sweet introduction to the five most familiar nonparametric location tests and associated confidence intervals and multiple comparisons. . . . This book is extremely limited in coverage, but none the worse for that. You won′t find anything on the Fisher Exact Test, or measures of association for the cases covered. What you will find is good, accurate, comprehensible accounts of location tests for two or more treatments. . . . The book is well-written throughout, with fewer than expected mistakes. . . . All in all, a good basic introduction to nonparametric tests, with few frills, as you would expect. The computer package comparisons sound very useful warning bells and are a welcome frill."

--British Journal of Mathematical and Statistical Psychology

Using actual research investigations that have appeared in recent social science journals, Jean Dickinson Gibbons shows you the specific methodology and logical rationale for many of the best-known and most frequently used nonparametric methods (that are applicable for most small and large sample sizes). The methods are organized according to the type of sample structure that produced the data, and the inference types covered are limited to location tests, such as the sign test, the Mann-Whitney-Wilcoxon test, the Kruskal-Wallis test, and Friedman′s test. Formal introductions to each test are followed by a data example, calculated first by hand and then by computer.