Reconstructing Chi-squared Intra-table Distances in the Analysis of a Set of Contingency Tables

Abstract

To study similarities among the set of rows -and columns- of a contingency table, Correspondence Analysis uses chi-squared distances between row profiles -and column profiles- of that table. This article presents a factor analysis for the study of a set of contingency tables in which, unlike more classical methods such as the analysis of juxtaposition and Intra Analysis; the existing relationships within each table -defined across the chi-squared distances-, are not altered.