Multiple correspondence analysis decomposes a matrix of indicator variables formed from all the variables that you enter. Unlike simple correspondence analysis, where all row classes are from one categorical variable and all column classes are from another categorical variable, multiple correspondence analysis has only column contributors for all categorical variable classes.
Multiple correspondence analysis performs a weighted principal-components analysis of the matrix of indicator variables. If the number of categories in the j categorical columns are c1, c2, ... , cj, the number of underlying dimensions is the sum of (ci - 1), where i = 1, 2, ... , j.
As with simple correspondence analysis, multiple correspondence analysis partitions the Pearson χ2 statistic. Unlike simple correspondence analysis, which includes both row and column profiles, multiple correspondence provides only column profiles. Because there are no rows, multiple correspondence analysis offers only one graph, a column plot, which plots the column coordinates.