Rank And Pseudo-Rank Procedures For Independent Observations In Factorial Designs

This book explains how to analyze independent data from factorial designs without having to make restrictive assumptions, such as normality of the data, or equal variances. The general approach also allows for ordinal and even dichotomous data. The underlying effect size is the nonparametric relative effect, which has a simple and intuitive probability interpretation. The data analysis is presented as comprehensively as possible, including appropriate descriptive statistics which follow a nonparametric paradigm, as well as corresponding inferential methods using hypothesis tests and confidence intervals based on pseudo-ranks. Offering clear explanations, an overview of the modern rank- and pseudo-rank-based inference methodology and numerous illustrations with real data examples, as well as the necessary R/SAS code to run the statistical analyses, this book is a valuable resource for statisticians and practitioners alike.