Underlining these college completion rates are prominent racial gaps. In 2009, over 50% of Asian adults aged 25 and older held a bachelor’s degree or higher, compared to fewer than 20% of African-Americans and Hispanics.
A natural question to ask is, once students enter college, what factors determine the likelihood they succeed and graduate?
A study I conducted with coauthors Doug Campbell from the New Economic School in Moscow and Scott Carrell from University of California, Davis, shows the race of teaching assistants (TAs) play can play an important role in academic outcomes.
Several prior studies have presented evidence on various factors that could influence undergraduate success, including professor quality, gender and race, coaching and advising, and academic probation.
However, one glaring omission has been teaching assistants (TAs), who account for nearly 15% of the total employment of post secondary teachers in the US annually.
TAs are graduate students employed by a university who perform various duties in the course while under the supervision of a professor or lecturer. Many of these duties, such as hosting small weekly discussion sections, holding office hours, tutoring and grading assignments, are likely to impact student success in the course.
TA-student relationships are unique in that they are more likely to be a peer-based interaction, since the typical age gap between undergraduates and TAs is relatively small. Additionally, with class sizes and student-professor ratios increasing in the US, TAs are likely to play an increasingly important role in the US post-secondary education system.
Undergraduate and graduate programs in the US have been experiencing a dramatic shift in student racial composition for the past 40 years: In 1976, 82% of students enrolled in undergraduate programs in the US were white, compared to only 57% in 2013. A similar pattern can be observed in post-baccalaureate programs, where the fraction of nonwhite students grew by 180% from 1976 to 2013.
To investigate the importance of TA race, we collected student administrative data from a racially diverse public university in California, coupled with TA assignment data from the university’s Department of Economics.
We found that students received better grades in classes taken with TAs who were of a similar race. Students were also more likely to attend their TAs' optional discussion sections and office hours when the TA was of a similar race, providing direct evidence of students responding to similarly raced TAs.
The results are more prominent in classes where TAs had been given a copy of the exam prior to the exam date. We consider this to be evidence of “teaching to the exam,” where TAs adjust their lesson plans to divulge information that is more relevant to the exam. Consequently, students who visit their TAs are better prepared for the exam.
We also found racial interaction effects to be the strongest in classes which had no multiple choice on the exams. This result could stem from several possible explanations:
First, critical thinking is typically a key component to success on essay-based questions. Critical thinking skills may be fostered in settings where students discuss and ask questions about the course material, such as in TA discussion sections and office hours.
The second explanation suggests that TAs are (subconsciously) responding to students of similar race through grading. Classes with no multiple choice exams are classes where TAs have to exercise more subjective judgments when grading, and students of specific races may be more likely to answer non-multiple choice questions in a manner that TAs of similar race favor.
Research suggests that equally skilled students of different races may perform differently due to the students' self-belief about their ability to succeed. We believe these gaps may be muted (or exacerbated) by the TA’s race.
Another way in which the TA’s race could exercise influence is through a “match quality effect.“ TAs of different races may have, on average, particular teaching styles or capabilities that could be better suited to students of similar race.
This channel includes language matching, where, all else equal, a student learns more if particular material is taught in the student’s native language. Students who share the same race as their TAs are more likely to share the same native language.
Finally, TAs may exhibit bias with respect to how they treat, consciously or subconsciously, students of a similar race.
Overall, students perform better when taking classes with TAs with similar race. Many factors, such as differences in student attendance or grading, could be driving these results.