Conceptual Models for Visualizing Contingency Table Data

For some time I have wondered why graphical methods for categorical data are so poorly developed and little used compared with methods for quantitative data. For quantitative data, graphical methods are commonplace adjuncts to all aspects of statistical analysis, from the basic display of data in a scatterplot, to diagnostic methods for assessing assumptions and finding transformations, to the final presentation of results. In contrast, graphical methods for categorical data are still in infancy. There are not many methods, and those that are available in the literature are not accessible in common statistical software; consequently, they are not widely used.