In its most general form, tabular analysis includes any analysis that uses tables, in other words, almost any form of quantitative analysis. In this article, however, it refers only to the analysis of categorical variables (both nominal and ordered) when that analysis relies on cross-classified tables in the form of frequencies, probabilities, or conditional probabilities (percentages). In general, the use of such cross-tabulated data is practical only with variables that have a limited number of categories. Therefore, this article deals with some of the analytic problems of categorical data analysis. Although it sometimes is difficult to separate analysis from methods of data presentation, the emphasis here is decidedly on analysis (see Davis and Jacobs 1968).

Tabular analysis can take many different forms, but two methods deserve special attention. The first is known as subgroup analysis. The underlying logic of this type of analysis was codified under the...