Applied nonparametric statistical methods. (Q2756709)

This third edition of the monograph on statistical methodology again concentrates on by now classical issues in nonparametrics. It consists of eleven chapters. After a brief introduction, chapter two deals with elementary properties of signs and ranks in the context of a one-sample location problem. Chapter three offers some extensions, including, e.g., runs, the Kolmogorov-Smirnov goodness-of-fit test and a discussion of angular data. In chapter four, these issues are applied to analyze paired data. Chapter five deals with rank statistics for two-sample location and scale families. Three and more samples are investigated in chapter six (Kruskal-Wallis and Friedman tests). Problems related to correlation and regression are briefly discussed in chapters seven and eight, while chapters nine and ten are devoted to the analysis of categorical data. Finally, in chapter eleven, some issues on M-estimation and the bootstrap are addressed.NEWLINENEWLINENEWLINEThe text more or less contains no proofs. Rather the material is presented in an informal way, discussing fundamental issues like, e.g., small sample approximations, P-values and power of tests through many data examples. A possible reader would also benefit from working through the many exercises attached to each chapter.NEWLINENEWLINENEWLINESummarizing, the monograph offers a traditional approach to nonparametric statistics. More modern concepts based on stochastic processes are not mentioned. Mathematical details are kept to a minimum so that the text is also accessible to applied researchers without a strong mathematical background.