Visualizing Categorical Data*

Michael Friendly
York University
Toronto, Ontario

* Published as Chapter 20 (pp. 319-348) in Cognition and Survey Research, edited by M. G. Sirken, D. J. Hermann, S. Schechter, N. Schwartz, J. M. Tanur, and R Tourangeau, 1999, John Wiley and Sons, Inc. ISBN 0-471-24138-5. These copies reflects the submitted draft, but not editorial changes in the final copy. I am grateful to Douglas Hermann for helpful comments on the manuscript.


Abstract

Graphical methods for quantitative data are well-developed, and widely used in both data analysis (e.g., detecting outliers, verifying model assumptions) and data presentation. Graphical methods for categorical data, however, are only now being developed, and are not widely used.

This paper outlines a general framework for data visualization methods in terms of communication goal (analysis vs. presentation), display goal, and the psychological and graphical design principles which graphical methods for different purposes should adhere to.

These ideas are illustrated with a variety of graphical methods for categorical data, some old and some relatively new, with particular emphasis on methods designed for large, multi-way contingency tables. Some methods (sieve diagrams, mosaic displays) are well-suited for detecting and patterns of association in the process of model building; others are useful in model diagnosis, or as graphical summaries for presentation of results.

Complete paper

These versions contain the color versions of figures which do not appear in the printed chapter (alas).