An analysis and synthesis of multiple correspondence analysis, optimal scaling, dual scaling, homogeneity analysis and other methods for quantifying categorical multivariate data (Q1070717)

It is shown that all the four approaches, viz. reciprocal averages, analysis of variance, principal component analysis and canonical analysis are equivalent for multiple correspondence analysis for scaling of categorical data. The most interesting part of the paper, however, is the introduction of the 'duality diagram' and the role it plays in the understanding of the geometry of multiple correspondence analysis and its synthesis. The duality diagram has indeed played a pivotal role in the development of this area of data analysis in the French literature.