Tips for supervising TAs?

In our June "how can we help you?" thread, firstTimer writes:

I will be the primary instructor for a class with graduate student TAs for the first time this fall. The TAs will be responsible for grading but there will be no recitation sections. I was wondering if readers could share tips about working with TAs -- how often do you meet? Do you try to calibrate grading standards? How? Should TA supervision be seen as an opportunity to teach grad students about pedagogy as well? Any advice for a first timer would be much appreciated!

It's been a long time since I've supervised TA's, so I'm probably not the best person to ask for advice--but, for what it is worth, I remember a pretty wide variance for how grad faculty approached this. On the one hand, some did very little like anon says here:

I did almost nothing re supervise the TA. I let them grade by their own standards, and had them come talk to the class about how they would grade. But maybe that's just me!

Others, however, were pretty overbearing, requiring TA's to meet to compare samples of graded work for every major assignment (exams, term-papers, etc.). I'm not sure what the right balance is. Anyone out there with experience here have any tips or insights to share?

Comments

[once a TA]

I think we owe it to grad student TAs to mentor them. I work with TAs - they have tutorials to run as well - and we meet every second week. It is very useful, and provides another means to get information about how the teaching and learning is going. Why wait to end to find out. When I work with TAs, I do what I experienced when I was a TA. I convey to the class that me and the TAs are a team, working on the course together.

[Jake Wright]

As someone who teaches in teams with other faculty and who supervises TAs now, I think that having clear, agreed-upon expectations of everyone is probably the most important thing you can do. There are defensible arguments for hands-on and hands-off approaches (and everything in between), but in my experience, the biggest problems arise when it's not clear who should be doing something or why. I give everyone a document that outlines the expectations everyone should have of everyone else, and then we discuss it and I incorporate feedback from the TAs into a final version that all agree to.

That being said, I favor something closer to a hands-on approach. The best experience I had as a TA was with a professor who brought us together to compare three or four exam essays so we could normalize our grades (he told us all to bring an A, a B, and a C). It was especially useful for those of us who hadn't really had a lot of grading experience before. He also used the course as a chance to mentor us as teachers, giving us all the opportunity to present a lecture to the large class. Keep in mind that most, if not all, of your TAs will get jobs at teaching-focused institutions unless you're at an elite R1, so an opportunity to make them better teachers is vital for their professional development. I like to think I'm a pretty good teacher, and a big part of that was the support I got from faculty when I was a TA, whether it was the formal stuff mentioned above or informal discussions walking to and from class.

[anon]

I used to do long calibration meetings, but more and more I've been going in the direction of having everyone meet to discuss a detailed rubric that keeps them on the same page.

I find that giving new TAs detailed instructions on how to leave written feedback is important for preventing overwork. They tend to write too much - comments that most students won't read, and comments that they spend too much time writing.

One way you can have a grading-only TA make more of a connection with the class is to have them do a guest lecture. (And you can give them some feedback on the lecture, and be in a stronger position to recommend them after seeing them lecture). Of course, all this should be done with an awareness of how many hours it will take them to prepare the lecture, and adjusting things accordingly.

[Sebastian Lutz]

I'm also in favor of calibration meetings when it comes to grading. But I also often provide a list with the main points that I want the answers to contain.

Regarding teaching, I usually also have a list of points that I want the TAs to cover during their seminars.

[Help if you want it]

My approach has been to meet with TAs at the beginning of the semester/quarter, mostly to achieve what Jake Wright discusses (turn around times for grading, setting of other basic expectations, etc.). I also use the meeting as an opportunity to indicate that I am eager to be as collaborative as any given TA would like (e.g. I offer to do observations, either for people who would like a teaching letter or for people who are new to teaching, look over some graded assignments before they're handed back, triaging problems as they pop up, etc.) but that each TA is free to do their own thing (insofar as this doesn't impact their ability to meet the basic expectations we set up) if they'd prefer.

[Michel]

I can only echo what's been said. But I want to emphasize this: don't let them overwork. Even if they want to.

Partly, that means allocating realistic hours to the tasks you're having them perform, and checking in periodically to ensure they're not exceeding those hours. But partly, it also means modelling efficiency, and helping to mentor them with respect to that.

[Amanda]

I find this interesting just because I spent 5 years as a TA, and never once did I have an instructor that was anywhere close to as "hands-on" as the above posters describe. I never once had an instructor talk about unifying grading standards, nor did I ever have a TA meeting with the instructor.

My thoughts are that are that TAs need more instruction and guidance on teaching, and probably grading strategies, but that is not really what I see described above.

Anyway, I'm curious if my experience was out of the norm. For other persons who were TA's, did instructors do the things described above? Did I just have an especially unusual experience over my 20 courses or so of TAing?

[Current PhD student]

The best thing you can do for a TA is ensure they have no grading or regular teaching responsibilities. If a TA is used as a relief to your workload, you are doing it wrong.

You and the TA can both offer comments on a sample set of some papers and compare your results. A TA should teach no more than one week of class, and the lead instructor should be present for both of those sessions to provide feedback on each session. This is the model for my current assistantship, and I have been learning so much from the experience. I also feel valued by my department, which was not the case when I was a TA in my last department.

If the experience is not entirely about mentoring the TA, then the TA won't gain anything, and the TAship isn't really worth that person's time.

[anonymous person]

"If a TA is used as a relief to your workload, you are doing it wrong. "

Aren't TAs employed in order to do some of the workload of a course?

(I totally agree that we should mentor TAs carefully and well, and also that it is important to take care with how one constructs assignments, how often assignments are due, etc. to not overburden them. But I'm confused by the idea that TAs should not grade. When I have TAs, they teach sections, grade (I share grading with them, which is *not* the norm in my department) and guest lecture if they want to (which they always do, since they have an eye towards improving their teaching and job marketability), and help me with course planning and construction. Some of our TAs don't teach at all, since they are hired for courses without separate discussion sections. Are you saying that these TAs should also not be asked to grade?

I'm totally open to being wrong about all this. But I do think being a TA is a job as well as an opportunity to learn and be mentored. If TAs are hardly teaching and are not grading, then what are they being paid for? Just being mentored?

[Current PhD student (same as above)]

I am saying that no TA should grade, yes. Although, they should practice giving comments on papers (20 papers per assignment is more than enough) and talk through their comments with their professors. But uh, teaching a regular discussion section while in coursework? Not a chance. I've done that. It destroyed me.

I am being polemical, but not just for the sake of it. In my department, when a professor accepts a TA (we actually call them something different), that professor is signing on for an extra workload and receives no relief. The concept is exclusively oriented around mentoring graduate students, and there isn't a whiff of exploiting grad student labor. It has made graduate school such a pleasant and more supportive experience for me. I actually feel like my professors care about me when they take me on.

And yes, they are (I am) paid to be mentored, if you want to describe it that way. I get paid to learn how to teach philosophy. That still involves going to an undergraduate class, prepping for that class, reading/commenting on samples of student work, holding office hours, teaching for one week, talking about activities, creating essay prompts, and all the normal things a TA does (besides be responsible for grading and teaching a weekly discussion section). In almost every way, the experience is the same as others--it's just that I am not used as a relief for a professor, and that makes me feel good.

[Amanda]

Wow I can't imagine what program current PhD student is in. I thought the entire point of TAs was to grade and teach sections. That's what I did for all my TAships. If it destroyed ppl, our program would have no graduates.

How many TAs don't grade or teach sections? I think 90% of them do, if not more.

[anon]

Current PhD, your situation sounds nice, but I think extending it to many other programs would involve radically restructuring how their classes are run. TAs aren't optional "relief" at many institutions; instead, they're a necessary part of dealing with the grading.

I'm talking about situations where one faculty member is (say) teaching two courses, each of which has 200+ students enrolled. So at schools like these, TAs that don't grade aren't really something to aspire for. And grading isn't a part of that faculty member's responsibilities to begin with.

That said, at these institutions, mentoring the TAs can still happen. Several posts in this thread have been about how to manage that. Since this is how TAs work at most institutions, I think it's a productive thing to talk about.

[anonymous person]

Current PhD student: It's nice that you're in that situation, but... if teaching a discussion section during coursework "destroyed" you, then you should probably know that while many graduate students are underpaid and/or overworked, having a tenure track job, or a temporary job where you are teaching something like a 4-4 courseload, is a lot more work than taking some classes, writing some papers, teaching one or two discussion sections a week, and grading some papers.

Also you should recognize that the situation you are in is very abnormal for a lot of reasons. I teach over 350 students some semesters. (I have a 2-2 course load, but I often teach two large lecture courses in the same semester.) I am on the tenure track at an R1 where I am expected to publish a lot, and in top-10 journals, in order to get tenure. If my TAs were not doing a significant portion of my grading... it would actually be physically impossible for even a better, improved, more efficient version of me to do the research (and, realistically, service) components of my job. In order to accomplish what you are describing, I would need my university to hire way more faculty in my department, reduce class sizes, etc.

I think it's important that we mentor grad students well. But it's also important to prepare grad students for the reality of what it is like to be an academic. If you're the kind of person who is "destroyed" by teaching discussion sections while doing coursework, maybe being an academic is not for you.

[Current PhD student (same as above but less snobby this time)]

"If you're the kind of person who is "destroyed" by teaching discussion sections while doing coursework, maybe being an academic is not for you."

Honestly, I learned that lesson during my first program, in a way that was very hard. I wish I could have learned it in a nicer way. But yes, I have no intention of taking a 4-4 course load--I am not particularly torn about leaving academia if I can't get a job where I already live, a good and stable salary, and a 40 hour work week. To me the choice is easy, but I know I'm not the norm. I think it would be fun to adjunct one course a semester on the side, and to be able to pass on courses that are unreasonably large without worrying about paying the rent. So that's my goal.

And I recognize it would be unreasonable to think that programs could possibly lower their class sizes from gargantuan to reasonable. I don't really know how I would handle a large class as a student or instructor.

If I'm being less polemical, I would just say it's helpful to craft assignments that are easy to grade if classes are large. Additionally, I would suggest developing a lesson plan to give your TAs every week if it's the case that they have to teach a regular discussion section every week. (Don't make them create lesson plans every week.) I also suggest cutting them slack whenever they seem to need it, without asking any questions, and maybe without them even asking for slack in the first place. And talk to them about little tricks you've learned. One professor told me she stopped assigning readings on the first day after a break, because no one really did them anyway. For her it was better to embrace that reality and make those days into activity days. Those anecdotal tips help me a lot, I keep a list of them.

[Amanda]

I always taught two discussion sections a week, plus grading and attending the professor's lectures. I don't think the professors would have had any way of knowing if "I looked like I needed a break" as I had no interaction with them throughout the semester. That is one extreme, I guess - although I didn't know it at the time.

I think that grad programs should have courses on teaching, or a standardized mentoring program. I do think it is crazy that most grad students get little to no training in teaching. Out of the faculty at my program, one professor came and watched me teach one section - I did TA for this person two or three times, so I got the feedback two or three times. But this was one professor out of 20. Anyway, I think there needs to be some type of mandated mentoring or teacher training in most programs, because not enough professors will do it if left to their own devices.