I am trying to fit data from 3 closely related experiments. Basically the same experiment taken under slightly different running conditions. The PDFs from the three running conditions will be different. Should I just do a simultaneous fit? Or is this a situation where rooCategories might be fruitfully used?