For primary data usually there are too many similar variables which are available for estimation, for this PCA and FA method are used to arrange the variables into groups which are most relevant to each other in order to prioritize the selection of variables. Following is the tutorial of this method in STATA