If you want to see how I recreated the OP's graph, just fork my GitHub repo. It has R code and the data sets, too:

https://github.com/tmoertel/nyva-cursory

Anyway, here are the variables I used to identify comparable observations:

    ## these variables identify comparable teacher observations
    id_vars <- c("subject", "grade", "dbn",
                 "teacher_name_first_1", "teacher_name_last_1")

So I used subject, grade, school (identified by dbn) and the teacher's first and last names.