I use many categorical variables and I want to use these variables to generate latent classes by LCA, how to select the most discriminating variables ?

Are there any techniques or methods on statistical software to choose variables for latent class analysis (LCA) ?

Similar questions and discussions