Measures of Association for Ordinal Variables

The study and analysis of association between variables is of fundamental interest in most scientific disciplines. Ordinal variables are variables whose values only can be compared in terms of their order. Sometimes referred to as ordered categorical variables, ordinal variables are common in, for example, medicine and the social sciences. Measures of association for ordinal variables are consequently useful in numerous applications. This book is the second printing of the author's dissertation from April 2009. The book combines rigor with readability, and will serve as a good introduction to the subject from both a theoretical and a practical point of view. The reader will gain an understanding of the complications of measuring association of ordinal variables, as well as the main techniques for circumventing the problems. The book also contains many interesting historical notes. All methods discussed in the book are implemented in the R package JJcorr, which is maintained by the author.

Autorentext

Joakim Ekström is a statistician with a Ph.D. from Uppsala University, Sweden. His research interests are broad and include multivariate analysis, measures of association, analysis of categorical data, as well as the history of statistics. Besides statistics, he enjoys composing and playing music, and is a marathon runner.