“…The B matrix, which represents the structure induced by T on data, has row margins b t: Camillo and D'Attoma (2010) and D'Attoma (2009), a default analysis that deals with the factorial decomposition of the inertia related to the juxtaposition of the K and T table, in the presence of categorical variables, is Multiple Correspondences Analysis (MCA) (Benzécri, 1973;Greenacre, 1984;Greenacre & Blasius, 2006;Jobson, 1992;Lebart, Morineau, & Tabard, 1977;Saporta, 1990) that aims at studying the marginal links between pairs of categorical variables in a given table and studying the structure induced by these variables on the units (Estadella et al, 2005). MCA (Benzécri, 1973) is a popular explorative multivariate technique for the analysis of any kind of matrix with nonnegative entries, but it especially involves table of frequency or counts with more than two dimensions in which makes sense the sum by rows or by columns.…”