I have a dataframe that contains similar values across three columns of school, lesson, grade and teacher_id. The column enumerator is not unique because the data collected was observation data collected by two different people on the same school, class, lesson and grade. Obs1, Obs2 and Obs3 columns are scores given by the two enumerators(pairs) based on the 3 entities (Obs1, Obs2, Obs3) being observed in the class.