Modeling Categorical Data with Chi Square Automatic Interaction Detection and Correspondence Analysis

Abstract

For the past decade we have been concerned with the application of categorical data techniques to mobility and housing choice. In a previous issue of this journal we reviewed a number of these modeling techniques. We now extend that earlier work by introducing correspondence analysis as a geometric interpretative technique that fits in an intermediate position between CHAID and nested logit analysis. The method is illustrated on the same housing and tenure choice data as was used in the previous paper. 1991 The Ohio State University