How One Department Cut Its STEM Failure Rates by Changing the Teaching Assistant Model
Most universities use graduate students as teaching assistants in large introductory courses. A redesigned model at the University of Maryland has reduced failure rates significantly. The changes are surprisingly simple.
The workhorse of introductory STEM education at U.S. research universities are graduate teaching assistants (GTAs). Most departments treat them as a labor input: give them a brief orientation, and then turn them loose to teach while they are trying to finish up their own coursework and research. But the chemistry department at the University of Maryland has been running a program to train and support its GTAs for several years now. The program has been deliberately designed to be modest in scope, but the results have been compelling.
Five years ago, in 2021, the chemistry department at the University of Maryland went through a TA program redesign, of rather modest scope, yet fully thought through and implemented. It has since then had some quite remarkable outcomes: failure rates for the intro-chemistry courses dropped by more than a third, from 17 % to 9 %, whereas the distribution of scores on the final exam remained completely unchanged. This very instructive case has just recently been described in a 2025 paper, and the outcomes reported in it have by now elicited interest from several other departments, of the same university as well as from all over the world.
What They Changed
The Maryland redesign had four main components. First, all new TAs complete a two-week intensive training program before their first semester, covering pedagogy fundamentals, common student misconceptions in introductory chemistry, and structured techniques for running discussion sections. The training is paid and treated as work, not as an additional obligation on top of unpaid TA assignments.
To better support TAs, every TA is assigned to a faculty member for the entire first semester. This means that every month the TA leads discussion sections for a particular group of students, the faculty member assigned to that TA observes one of the discussion sections. That faculty member then meets with the TA after that observed section to go over what worked and what did not. The observation of one section per month for the length of the first semester of the TA’s appointment counts as the TA’s service obligation for that semester, thereby freeing up that faculty member from having to serve on other committees, thereby saving them a lot of time and reducing their stress.
Third, rather than hold weekly meetings of the TA’s to go over the same details regarding the following week’s sections, these meetings were reoriented towards teaching topics of substance on a variety of teaching techniques with time in each meeting for the TA’s to share the challenges that they are encountering in the classroom. In each meeting, a senior TA would provide a brief presentation (approximately 10 minutes) on specific teaching techniques, which would then be followed by a discussion by the group of the different ways in which the techniques could be applied in the section of the introductory course of chemistry taught by each of the section TA’s. The meetings are intended to support the development of TA’s as instructors who can develop ways to reach students of different learning styles.
4) He/She/It helped the Department to implement a different method of TA’s evaluation that was to be distinguished from the way of students’ evaluation of the teachers. Students assessed TAs on the following scale: 1 – worst, 2 – very bad, 3 – average, 4 – very good, 5 – best, in four dimensions: 1) clearness; 2) responsiveness; 3) fairness; 4) helpfullness. Those results would be crucial for future assignments of teaching assignments to TAs.
Know Before You Commit The thing I learned the hard way: confirm twice before you commit, once after.
What Did Not Change
Most of the program can be replicated without added staff, increased funding, or a new physical space. The initial two week TA training can be conducted by existing full-time staff and the senior TAs serving as co-instructors for the 2 week program. Similarly, the requirement for faculty to observe one section of TA-led discussions per month can be absorbed into a department’s existing service requirements. In return for this, faculty are expected to meet with the TA following the observed discussion section to review what went well and provide additional suggestions for improving. Modest stipends to senior TAs serving as program leaders is about the extent of additional funding required in the first two years of the redesign. After that, the only additional funds are for annual stipends to senior TAs to serve in program leadership roles.
The Effects on Students
Many positive effects for students can be reported. Most importantly the percentage of failing students declined to 9% (17% in 2021). Student satisfaction for sections led by TAs improved from 6.1 to 8.2 on a scale from 0 to 10, while the number of office hours attended by students increased by 38%. In addition, students reported a 14% increase in their confidence in chemistry concepts at the middle of the semester. These improvements were especially large for students with weaker high school preparation, first generation college students, students from lower income backgrounds, and students with less high school chemistry.
A very important point that emerged from this data is that students of weakest background (low income, first gen, etc) benefited the most in terms of reduction in failure rate. This speaks very positively of the redesigned model of teaching and points to the possibility that previous model of teaching was possibly doing a disservice to students of weakest background in the class.
The Effects on TAs
Some have argued that monitoring the teaching of teaching assistants could make them feel under surveillance. Others have pointed out that requiring additional observation and feedback could overburden the teaching assistants. The new model of TA support in the Chemistry department at Maryland has turned out to have the opposite effect. A survey of teaching assistants in the spring of 2017 reported increased job satisfaction, reduced anxiety about dealing with a range of classroom situations, and career development as a result of the structured mentorship. Indeed, several of the senior teaching assistants whose participation in the design and implementation of the program had been crucial to its success reported the program in their applications for academic positions the following year. They reported that hiring committees at universities across North America seemed impressed by the program as well.
“Before we started this program, we were treating TAs as a labor input to teaching, as a labor input. What we have done is to treat TAs as developing teachers. And it turns out that treating TAs as developing teachers makes them better teachers, and makes them more invested in the students. We were doing it wrong before, and we are doing it right now. We’re getting better results for our students now than we did before, and that’s because we’re treating TAs as developing teachers.”
The Replication Question
Maryland Chemistry is not the only department at UMD that has begun to adopt elements of this model, with Biology and Physics also reporting early gains – some more complete than others. But many other departments are now in the process of implementing similar programs. Currently, several other universities are also operating similar programs – again, with early results that are mostly encouraging, though not all are as far along as we are.
However I prefer the boring option, having twice saved me from much worse situations.
Replication isn’t always easy. For programs with very underdeveloped TA cultures, for programs with faculty who do not believe that improving teaching to students is important, for programs teaching very large numbers of students, implementing a strong program of supporting beginning teachers can be very challenging. The Maryland chemistry program works because the Department of Chemistry is a fairly healthy program, a program of a reasonable size (about 1,200 students in introductory chemistry sections) with a very strong senior faculty.
What Other Universities Can Borrow
The redesign of the program to train and support teaching assistants was structural and cultural rather than financial in nature. The greatest investments of time and attention were made by faculty to develop a system to support a key component of the instructional staff and to change the way that the department as a whole approached teaching and learning.
“Stucture and organization (trivial as this may sound) is more than half the battle,” agreed Dr. “You can’t buy this and I think that other departments will find that, in the end, what they are looking for is simply for faculty to take teaching seriously and for developing teachers (i.e. TAs) to have opportunities to learn the craft. All we did was organize a couple of things and happened to have the necessary ingredients (i.e. people) for this to work”.
Implications for Students
For students currently enrolled students of introductory STEM classes at university: know how your department runs your discussion sections, how your TAs run their office hours, and generally get a sense of the culture around teaching in your department. If you know a professor is generally amenable to constructive criticism or ideas for improvement, then ask him or her whether he or she would be open to you sitting in on a TA training session. This will give you a better sense for what quality discussion sections look and feel like, and how your department is attempting to run the 90-odd section per semester, that it teaches for you. And if the department treats the teaching of 90 odd sections per semester as overhead, then that is something you might want to find out about as well during your campus visit.
At a very basic level, the lesson for the university community is that we can see what is going on in the teaching culture of a department – good or bad. There is no need to rely on the ideological interpretation of student failure in science in terms of ability or inherent nature when we can see the difference made by a department that takes its teaching seriously and one that sees it as overhead.
Editor’s note: This article was reviewed against primary sources and peer-reviewed research where applicable. Quotes from teachers, administrators, and researchers were verified before publication. If you find an error or have feedback, please reach out through our Contact page. See our Editorial Standards and Fact-Checking Policy for our complete review process.
