Inside Higher Ed reports this morning (Sept. 21) on a new study from the Bureau of Economic Research, “TAs Like Me: Racial Interactions between Graduate Teaching Assistants and Undergraduates.” The study finds that “students earn higher grades at statistically significant levels from teaching assistants of the same race or ethnicity as the students.” The abstract is here and full text (for $5) here.
Inside Higher Ed must have paid the $5 (I have not), since Scott Jaschik’s article includes information not in the abstract. He writes that the authors, Lester Lusher and Scott Carrell of the University of California at Davis and Doug Campbell of the New Economic School (Moscow)
examined grading patterns and student and teaching assistant demographics in economics courses at a large, diverse (and not identified) university in California. To avoid any issues related to self-selection by students of particular teaching assistants, they were able to work with data from sections where TAs are assigned to sections after students have enrolled in them.
Part of the reasons for the higher grades, they found, is that “students were more likely to attend optional discussion sections and to go to office hours when teaching assistants shared their race or ethnicity.” Another big part, however, is TAs’ teaching to the test. As Scott Jaschik writes:
The impact of same-race TAs on student performance is the greatest in classes without multiple choice questions but with tests that are available to teaching assistants in advance of when they are given.
“We interpret this result as evidence of ‘teaching to the exam,’” the authors write, ”where TAs divulge information that is pertinent to the class’ exams if given the opportunity. Students who are more likely to interact with the TAs by attending the TAs’ discussion sections and office hours are the beneficiaries of teaching to the test.”
The authors conclude, according to the abstract, that “TA-student match quality and role model effects are the primary drivers” of the fact that students get better grades from TAs of their own race and ethnicity. But a glaring possibility is largely ignored, both by the authors and IHE. Did students “earn” higher grades from TAs of their own race or ethnicity “in classes without multiple choice questions” — that is, in classes where grades are not determined by tests with right or wrong answers but rather by the judgment of the instructors — because of racially preferential treatment by the TAs?