Contingency Tables

A very common nonparametric statistical problem is the analysis of a table of data in which each cell in the table is a count in a particular category. The question is to determine whether the groups are statistically different from one another. As an example, consider the question of men versus women and their political affiliation, Republican versus Democrat, in which a 2 × 2 table is generated. If additional political groups are added, say Green and Undeclared, we then have a 2 × 4 table. If we wished to analyze these data to determine whether men and women differ in political affiliation or whether Democrats versus Greens differ in gender distribution, we would apply nonparametric statistical methods appropriate to a contingency table.