Is there any documentation for the output from the multiple correspondence analysis program? (In the Dimension reduction tab).
I think this may be doing what I want, namely to choose a representative subset of highly correlated variables, but not as explanatory variables, rather more in a PCA/FA descriptive sense, and to enable a smaller set of variables to be collected/analysed for future analyses.
What exactly is being done? This looks rather like PCA but what is going on?
e.g. what exactly are the transformed variables with the new transformed correlation matrix ?
what are the 'discrimination measures'?
what are the quantifications in the quantifications/categories plots?
More generally, documentation desperately needed.
I cant find any documentation on the web.