ABSTRACT
The concept of partial category is of great importance for analysing qualitative variables. The balance between both types of variable must be ensured both within mixed groups and between the groups, as usual in multiple factor analysis (MFA). Simultaneously processing quantitative and qualitative variables by factorial analysis was addressed for factorial analysis of mixed data. The quantitative variables are represented with the help of their correlation coefficients with the factors. The qualitative variables essentially appear through their categories, represented by the barycentre of the individuals that possess them. The quantity maximised by the factors of the MFA can be expressed in terms of a canonical analysis, with these factors as the general variables. The advantage of studying quantitative variables by coding them as qualitative and then conducting a multiple correspondence analysis (MCA) has been confirmed. The linear relationships detected by the principal component analysis are also identified by the MCA.
