Hi all
I am trying to make a multiple correspondence analysis in Stata using the Burt approach. My aim is to create a financial inclusion index from a set of dichotomous variables. I have one problem.
When predicting an index, can the index's mean value be negative?