I am investigating the relationship between latent classes of student experiences (also aggregated on the school level) and student achievement based on the PISA data (cross-sectional, continuous ‘distal outcomes’ and categorical ordinal indicators of classes). I intend to use Multilevel Latent Class Analysis, so that I can model the relationships on both the student level and school level. However, I am worried because the student data is nested not only in schools, but also in countries (there are more than 60 countries). As I am not interested in country-level effects, but only want to account for the clustering within countries, is there a way for me make such an adjustment? I could use either Mplus or R. Do you have any advice, or would you recommend any resources where someone used MLCA with such 3-level clustering to predict distal outcomes?
