Analyzing Large Multiway Contingency Tables : a Simple Method for Selecting Variables

EVEN in today's era of rapid advances in data analysis sophistication, the preparation and analysis of cross-classified frequency data is stiH an important activity for many marketing analysts. Often cross-tabulations are useful in summarizing complex marketing relationships. It is difficult to imagine any study in which the data are not displayed in frequency form at some point in the analysis. The simple two-way contingency table remains the most common type of cross-tabulation. However, restricting attention to only the pairwise relationships between variables may be naive, or even inappropriate, in many situations; rather, the marketing analyst faced with large numbers of variables may wish to focus on the structure of simultaneous relationships among three or more variables. When three or more variables are cross-tabulated, the table formed by combinations of these variables is called a multidimensional contingency table. Due in large part to the work of Green and his