S3xE3: Undergraduate Teaching Assistants with Michael Ball

April 4, 2022
S3xE3: Undergraduate Teaching Assistants with Michael Ball

In this episode, we talk with Michael Ball from the University of California, Berkeley. Our focus is on undergraduate teaching assistants (TAs), which Berkeley has a long history of. Michael goes into detail about what Berkeley TAs do, the TA hierarchy, and TA training. We learn about a small core group of students that make an undergraduate career out of being TA. Another question we focus on is advice to an institution that is just starting its TA program. Finally, throughout our talk, we discuss how things are different from before the pandemic and now.

You can also download this episode directly.


Kristin [00:09] Hello and welcome to the CS-Ed Podcast, a podcast where we talk about teaching computer science with computer science educators. We’ve decided to start something new. Rather than a season with a clear number of episodes, we’ve picked a theme and we’ll run with it until we run out. This season’s theme is “What’s next?” Where we focus on how we’ve rethought our teaching since COVID-19 came and upended everything. I’m your host, Kristin Stephens-Martinez, an assistant professor of the practice at Duke University. And joining me today is Michael Ball from UC Berkeley, who is a lecturer. And today we’re going to talk about TA management. So Michael, how about you introduce yourself a little bit and then we can get right into our topic?

Michael [00:48] Hello, everyone. Thanks for inviting me on the show. I’m Michael. I’m in my third year lecturing at Berkeley, where I was previously at Berkeley as a student. And currently in some of my roles, aside from just teaching classes, I also am, well, one of those class responsibilities is sharing the pedagogy course, which we’ll talk about. I also spent a fair amount of time working with our group of head GSIs as we call them, graduate students instructors, who are mostly undergraduates. But I work with my own teaching staff, and then throughout the semester, anywhere from 20 to 60 TAs across courses, collaboration, helping solve common problems, being a person for support and so on. So that’s one of the roles. And in that role, I help facilitate their learning and some of the practices that we share across all of our courses. So we’ll get into some of those benefits and challenges of having many, many GSIs.

Kristin [01:48] Yes, yes. So the seed idea for this episode was when I watched your TA training panel at SIGCSE 2021, and it kind of, I think in some ways did open my eyes a little bit to not everyone has undergrad TAs or undergrad GSIs or whatever words or acronyms that you want to apply to them. And I did grad school at Berkeley, and so I saw the like largest possible way of doing undergrad TAs and now here I am at Duke, which is, we do use undergrad TAs a lot, but it’s not at the same scale. And I think there are actually some benefits to scale, but I don’t know if it’s worth the benefits? It’s more like, it’s a nice silver lining, given that you have to deal with the scale that you have. Rather than like, gee, I wish we could do this kind of scale, it’s more like, OK, at least we have this given this is the scale we have to deal with. So one thing I did want to mention is that for the panel, you all did share slides and we will link to those slides in the podcast episode transcription.

Kristin [03:00] So, I think given the audience for this podcast, it would be useful to give them context of where you’re coming from and potentially where I’m coming from when it comes to TAs. So how about you go first, Michael?

Michael [03:18] Yeah. So, I do think Berkeley computer science is, well, there is a time where it was perhaps a little more unique, but it’s by no means sort of the only example. You know, undergraduate TAs have been used in various contexts. I believe some of the first documented work at SIGCSE goes back to Stanford in the 90s. It turns out we have the saying the only thing that scales with undergrads is undergrads. And one of the reasons for this is a function purely of enrollment. As our course enrollment grows, you need someone to be able to teach those sections.

At Berkeley, between our two programs right now, computer science, a B.A. degree, and EECS, a B.S. degree, and a little bit on the side of data science right now, we’re graduating more than 1,300 students a year. It’s somewhere between 12 and 18 percent of the Berkeley undergraduate population each year. And what this means is that our undergrad program vastly exceeds the size of not just the Ph.D. program, but the Ph.D. and master’s degree programs. And so we have a large number of undergrads. And the thing that I think across campus, as Berkeley has grown in capacity over the years, our major demands have shifted, we’re finding other departments on campus are using undergrad TAs as a either primary or secondary source of their teaching task force.

But also what I would say is that throughout this history, one of the things that I think makes Berkeley great is we invest incredibly heavily in training for teaching assistants. So there are campus requirements for all first-time teaching assistants, but also in the CS department itself, we have a culture that has basically allowed us to maintain this pipeline. So for context, what we’re kind of talking about is this fall semester, we had about, I want to say, 430 TA positions hired, and this does not include what we call tutors or readers. So readers are specifically graders, and tutors do one on one small group sessions. They typically don’t have office hours, but they don’t do grading, they don’t lead a section, and I don’t have those numbers offhand. But we hired somewhere around 430 to 450 GSI appointments undergrads, grads across more than 3,400 applications for them, so it’s an enterprise at this point. And we have the advantage of scale, it also means we have the necessity to sort of maintain the scale and the infrastructure to support the enterprise itself.

Kristin [05:53] So to provide more context. Do you have a sense of what the undergrad TA to student ratio is?

Michael [06:02] Yeah. So I would say in a classroom setting, so we typically at Berkeley use the terms lab and discussion, where lab is generally a little more hands-on, right? Sometimes those are actual computer labs, sometimes they’re labs where students bring their laptops, of course, and sometimes like for one of my courses, it’s actually sort of a hybrid lab discussion period. But generally, what I’d say is we sort of start with the assumption that in the room, it’s about one to thirty for most courses. And from there, we kind of work backwards. During COVID, we would play around, of course, online. We found that over Zoom, given attendance patterns or just the dynamics, we could go up to 40 or 45, and it didn’t feel cramped in some of these cases, or you could have different interaction methods. But in a room, we definitely don’t want to be packing people in, COVID or not being present.

Michael [06:54] And so I think that that’s generally the measure that we sort of unofficially shoot for. What happens in practice is that for courses that have multiple lab and discussion sections per week, sometimes those duties might be split across multiple undergraduates. So someone might cover the lab sections and students interact with another person for the discussion section. So in practice, it might actually look like the number of TAs is a little bit lower than 1 to 30, which is great. But what usually happens in that case is that the amount of work per TA in those cases, they are, you know, focusing a little bit more specifically on one piece of the course or a little bit more work. So it’s not like we’re taking these undergraduates, often cases do not have the same sort of appointment time as a typical graduate student works. So sometimes it’s half the length, sometimes it’s a third of the length of what a graduate student might have. And sometimes for our head undergraduate TAs, it’s the same amount of commitment per week. But across the board, it sort of varies by course and sort of logistically how it makes sense to separate duties, right? It’s really hard to divide someone’s time on the scale of sub hours per week, and it’s probably not super effective, even if it were possible. So each course will divide that labor as sort of fits the flow or the model of how that course is run.

Kristin [08:14] So what do TAs do?

Michael [12:56] So what TAs do depends vastly on the course. I will say maybe it’s interesting to talk about it, like the five percent of our undergraduate TAs, who in some ways treat this as like an undergrad TA career. I mean, I definitely put myself in this position, where the top five percent exist in the space where for some of our classes, the instructor could disappear and the class continues. And I don’t think I appreciated quite how unique this was until I went to my first SIGCSE conference as a student and the discussions were, you know, “Oh, well, how did you cover class this week?” or “Oh, we canceled class this week.” At Berkeley, and not every professor is like this and it depends on the class of course. But by and large, when we go to SIGCSE, our classes still run, and our TAs will be the ones who step in and give a guest lecture and, you know, they’ll prep for that. So the top few percent in some ways, they have for that specific course, the background, the knowledge, the understanding of the ethos of the course and the same way that the primary instructor might. And to the point where if my TAs make a decision, we’ll discuss and debate it or whatnot, or, you know, we’ll go through those. But I sort of treat that decision-making as they have the same context that another faculty member would, of like is this a good course policy or something like that? And we can have those discussions.

Michael [09:47] The majority of our TAs, though I would say, fall into this category where it’s a little more typical. Their primary responsibility is spending, you know, anywhere from one to three hours a week in front of students where that amount of time is dependent on, of course. They’re leading a discussion section, which is a little bit of review worksheets, hopefully pretty interactive. Though, as tends to happen, can tend towards review lecture, although we try to move away from that. It could be leading a lab section, which is hands on at a computer as much as we can, getting students to pair up and work together. And then the other bulk of their week is spent on office hours, so that’s anywhere from again one to three hours, depending on the course model. Typically, that’s in person, obviously with COVID, that was all remote for a while. This fall, it’s very much a hybrid experience. Either individual office hours can be hybrid or I have plenty of TAs right now that do one hour online, one hour in-person. And that’s sort of up to preference student demands and so on, how we schedule those things. And then I think the typical duties that come with TAs, so we use Piazza still, but trending towards Ed, so there’s a couple hours a week, depending on the TAs role of monitoring, answering questions. An hour a week of staff meetings, and at Berkeley, one distinction is that our TAs, when they grade, they primarily grade exams for most courses and our readers are the ones who grade homework assignments and projects and things like that for courses that have manual grading. It also turns out that most of the weekly work we’ve trended towards auto-grading.

Kristin [11:35] I think the piece of the picture that we haven’t super described before I want to go on to the next topic is more on what the top couple percent of your TAs do. Because I know from when I was a grad TA for the CS1 there, some of them very much specialized and basically seemed like the, you know, they were the top software engineer in the TA team. And often were also part of like, the level, as you said, where some of them like the faculty could walk away and they would be able to deliver the lecture.

Michael [12:12] Yeah. So I think, and after this we should talk about how TAs get there, because we sort of call this a pipeline and I think that’s very much critical to this process. So, yeah, our top few TAs in our biggest courses, they will sort of break up the roles and spend some of their time managing other aspects. So for context, when you have a 1,700 person course, and you’re trying to hit that ratio of 1 to 30, let’s say. What this means is you have somewhere around 50 TAs, you have another couple dozen readers, perhaps a couple dozen tutors. The largest semester that I’ve seen, the course that we’re talking about most prominently is CS61A, which is the largest course on campus. When it hits, typically, the enrollment will stabilize around 1,700 to 1,800 at its final semesters. But the total academic student staff hits about 90 or so. And so what this means is that aside from folks like John DeNero, who are incredible at making the machine run, you also have a ton of student support who have been with the course for a while.

Michael [13:28] And so typically what this looks like is John or the instructor and a few of the head TAs, the ones who set the goals for the week, right? They’ll perhaps do meeting agendas together, sort of get into a flow with that. You usually have one person who does software, and this sometimes will move around between courses. We have a large group of academic interns, which are sort of more commonly known as lab assistants. And this has actually been one of the changes that when we talk about pre and post-COVID models that I think is really both a challenge and opportunity. But it’s often one of the TA’s responsibilities is to manage the flow of academic interns, where it’s assigning them to lab sections to help out with, but also perhaps doing things like a weekly one-hour training with our academic interns to say, you know, here’s what the lab topics are, here are the things to watch out for with helping students. We call them AIs, which makes them sound like computer programs on their own, but they’re most certainly not. But one of their responsibilities will be making sure that our team of AIs, because they’re earning credits, that they’re showing up to labs, that the TAs who need help in their labs, that if someone needs to move students around and so on, that we manage that flow. So that’s a pretty big responsibility that happens primarily in our largest lower-division courses.

Michael [14:54] And then we’ll typically divvy up things, like discussion section review, homework review, any of the regularly occurring topics in our biggest courses will have one TA who understands either the infrastructure is particularly good, right? We always get TAs who someone gets the idea of coming up with creative homework problems or is really detail-oriented on reviewing the writing in labs. And so someone will lead a smaller team of TAs trading off responsibilities each week of just going through review, making sure things are up to date. And when you have anywhere from a couple dozen to many dozens of students, it’s helpful to have someone who says, you know, hey, this is my responsibility. And they’re like, in some ways, a good manager, right? They’re not doing all that work every week, but they are helping their other TAs do the work each week, whether that’s just keeping people on task or in the great cases also working with those TAs to say: Hey, when you review a lab, here’s how we want to go through this content. Here’s how we want to do our dry runs. Here are the things that we should note if there are issues. And that really becomes its own part of the continual training that TAs receive as well.

Kristin [16:09] So it sounds like to me that you have your top couple of percent of undergrad TAs that kind of are this very nice flow of they’re passing responsibilities back and forth to each other, each of them is specialized in certain ways, but they’re able to pick up and help out with each kind of thing. And that’s partially just because I’m assuming that they’ve TAed the class multiple semesters before they even got into that position.

Michael [16:35] Yeah. And at Berkeley, most TAs have about eight hours a week of assigned responsibilities, our head staff typically have between 15 and 20. So they’re giving, you know, quite a lot of their time to the course. It is, for many of them, the way that they fund their college experience. So our undergrads at that level get tuition remission in the same way that most grad students do. And also, of course, comes with an actual salary. So that is their college job and they do a good job at it, but they do sort of get to that position by spending multiple semesters, either as a sort of just starting out as a regular TA. In our upper-division courses, where you might take the course as a junior, you don’t have time to build that multiyear pipeline necessarily. And this actually becomes one of the bigger challenges is, they either were really early in their career where they took that upper-division courses as their first upper-division course late in their sophomore year or early in their junior year. But what often happens is they spend their time being a TA for a lower-division course and instead of just progressing within that course, they move on to the upper-division course. And so we have that sort of pipeline of training and experience transferring up the course stack in a little bit of a different way. But by the end of it, we have TAs who have been there, a few undergrads have the opportunity to do this for three straight years in a row, essentially beginning of sophomore year through the end of senior year. And by the end, you know, they knew that course inside and out. And the thing that we really have to watch out for is when they graduate, that they don’t take too much of the knowledge with them.

Kristin [18:13] Oh, the knowledge bleed.

Michael [18:16] Because that can be a real challenge. For the most part, if we have a continual stream of TAs, we have a few people who you know, they do it for a year, they find another job, they move to another course. And I think that’s pretty common. But we do have a core group that once they start, they see the course through until their graduation.

Kristin [18:36] One of the things that I do with my head stuff, if they’re graduating, is like, your job is to download your entire brain to your next person in line before you graduate, because senioritis is a thing.

Michael [18:49] Yeah, yeah, absolutely.

Kristin [18:52] The next person in line is supposed to be doing your job for the last month of the semester.

Kristin [18:57] OK, so transitioning to more about what happened when the pandemic hit. What did you change? What do you want to go back to as we slowly exit the pandemic? Probably, you know, at the speed of a glacier, but we’re still slowly making our way out. And what changes do you plan to keep?

Michael [19:17] On the TA front, one of the things that has come out of COVID that I think is really positive is, I think, the renewed awareness that students have lives and that there are things going on beyond our classes, right? We always knew this. We always told them that they should do things more than classes. But you know, everything from late policies to assignment deadlines to just how we remind students that it’s OK to ask for help and extensions, I think we had a renewed focus on, you know, it is more obvious right now, at least that many more of you are struggling. For whatever reason, right. And one of the things I think has been really great is, and again, both sort of the perils, but the benefits of scalability and a team to sort of meet that is across a bunch of our courses, beefing up the tooling that we have for addressing extensions, for addressing accommodations.

Michael [20:15] So for my courses, I like slip days as a policy. I think that they’re a good way of just giving students the ability if they need a few extra days on assignment to just take the time without sort of, you know, needing approval. And that already cut down a bunch of work. And one of the policies I did was I just gave people a bunch more slip days throughout the semester. And going forward, there’s no reason not to do that. It turns out that like, I’m sure there’s an optimal number, but if I set it high enough, students won’t feel the need to sort of hoard them like they sometimes did before or save them for the most important assignments and not just actually use them. And so far, I haven’t seen that students get any more behind by being generous on those things. So slip days are one sort of policy aspect of and that just involves setting up the tools that we have in a clever way. But we start doing things like tying some Google forms into some of our core systems. But also just setting up better defaults, right? Telling core staff anything under seven days, even if it’s sort of beyond the slip date deadline or whatever it might be for your course, whether you have them or not, if you ask for an extension of up to a week, we’re just automatically going to grant it, put it in. Reduce the barriers of asking for approval of do I need to like, you know, every bit of decision fatigue that we can remove is super helpful.

Michael [21:38] And so one of the things that we’ve been really good, I think about is think about how do we keep improving these processes going forward, such that any time something happens to some student, whether that’s just stress or something more serious, we have a clear way of planning and helping, you know, helping them out by saying, OK, here’s a quick, immediate step. A few courses now when you go beyond a couple of assignments, their process is you have to set up a Google form to schedule a 10-minute check-in with us, and at that 10-minute check-in, you set new assignment deadlines. And just realizing that we can set up our tooling and our course structure to make these things easy to respond to. And that our goal should be to, especially if it’s something formative like homework, right? Get them to do it. Because doing it is always better than not doing it.

Michael [22:28] And then also, I think just realizing that we can relax some of the tension that we felt about, well, if I give one student an extension, do I have to give every student an extension? It’s like no, it doesn’t quite work that way, right? But people already have individual circumstances. And so I think the big thing is being able to respond to students and provide flexibility. And we are slowly retooling the courses that with a thousand students really do feel like they need to sort of be this machine that is super consistent, sort of just gets in that position where it almost runs itself in a sense, it has consistent deadlines every week or every two weeks, whatever the cadence is. But now saying we can have a good template, right? Because for many students that keeps them on track, it helps pace the course, but we have a way of breaking out of this a little more easily when it’s necessary.

Kristin [23:22] So we’ve talked about what you can do with TAs when your program is fully grown. And also you potentially have the scale where your top five-ish percent of TAs are amazing and stellar and able to have, like, they have all the skills because not only are they super passionate, but they’re also like one of the top students in their class. So what kind of advice would you give to a department that is just starting out? And to provide some structure to this advice. What would you tell them that they need to consider? What they should consider? And what would be good to consider, but not required if they have to like, focus on the needs and the shoulds first?

Michael [24:06] Yeah. So one thing that I would say is, and fortunately, this is something that is free of time and effort, is you need to believe that your undergraduates are capable of getting to a place where they can be this good. And I don’t mean by thinking about admissions systems or GPAs or whatnot, right? Because some of my best TAs are TAs who originally got like an A-minus or B-plus right. They’re not always the A-plus students. Some of them absolutely are. And this comes from talking to other faculty members across campus who initially seemed horrified by this idea of many undergrads, but is to believe that, you know, at some point these students can be capable of learning and being engaged. And the second thing, I think is to trust them with the responsibility. And that doesn’t mean like, hey y’all, go write that assignment, throw it up on the website, we’re good to go for this week, homework’s out, right? You’re still going to be involved in that process. But I do think one of the things that I found empowering as a TA and I know my TAs find empowering is they get to think about a tough problem, right, like pedagogies, not a simple thing, there are no clear answers. Even writing an auto grader and what to assess you’re often like, actually, there’s more solutions than I thought there would be to this question.

Kristin [25:31] Oh, of course, always.

Michael [25:34] I think the really fun thing is when they realize that, you know, this is something that they get to make a decision. And, you know, hopefully, when they’re starting out, it’s not a unilateral decision, they’re in discussion with you. But what I also tell my TAs, especially the head ones, right? It’s like, if you make a decision and I end up disagreeing with it, and we need to maybe recover from it somehow, it’s fine, I don’t hold that against you. Like we can recover from a mistake if we need to. But I know at this point that you made a decision with the best interests of the course and the students in mind that you thought it was going to be helpful. And my goal is always to get, you know, as many TAs as possible to that point of like, if you need to make a decision for whatever reason, that you have a good framework and you have the ability to reason through that. And so, you know, I think that that mindset is really important. The second thing that does require time and investment is, and I don’t know how you do this necessarily all at once, because this is one of those things where there’s a lot of parallel processes running at our scale.

Michael [26:39] But the big thing that helps at Berkeley is we have academic interns. We have readers and tutors. We have, I guess we can sort of just call them, you know, regular TAs. And we have head TAs. And that is very clearly an ordered pipeline. And so having a thing where students can come into the class with little commitment and just help their peers is a great way for them to build interest in teaching, right? Being an academic intern is, I mean, it’s a unit or two. It is a commitment for the semester. But if it doesn’t work out, you don’t like teaching, you don’t like being in front of students. I mean, if you’re feeling challenged, but you want to keep learning, we have a course actually that’s dedicated towards academic interns now who just want to improve their skills.

Michael [27:32] So we have training there. That course is not required, but really what we have is a way for students to sort of gain experience working with students before they become a TA. And I think that that is super helpful. For readers and tutors, that’s typically the first, and those are paid positions. That’s sort of the first step where you’re a member, of course staff, you’re in weekly staff meetings, or at least weekly reader or grader meetings, you’re learning how the course works, and you’re gaining experience with providing feedback. All the aspects of pedagogy I would say, that are not necessarily directly in front of students. So that’s a huge aspect of this pipeline. And then, you know, typically when we draw on hiring new TAs for the semester, some courses actually conduct interviews. But we typically draw on our readers and tutors as our first line.

Kristin [28:26] Part of me is wondering if the only reason that kind of pipeline, semi pyramid works at Berkeley is because Berkeley is so big.

Michael [28:39] Yeah.

Kristin [29:40] And so you have critical mass in terms of the number of students that are willing to be in each of these different groups. While like, here at Duke, I’m definitely hiring undergrads straight into being a lab TA in front of a group of people, though at least for the lower divs, usually they’re working in pairs and we pair them up where there’s a more experienced TA with an inexperienced one. So there is some mentoring and that kind of thing happening there. So I think part of me agrees with you that this could succeed, where they do, you know, a quarter or a half of a class, however it has to be counted at the school, and they kind of are student TAs to the TA, kind of thing. To me, it sounds like that’s a strong thing you should consider if you wanted to start a TA program. Do you think it’s a necessary thing?

Michael [29:43] Yeah. I think what I would say at some level, and I guess the other thing that I should mention that is a campus requirement for at least Berkeley, if not the entire UC system, is all first time teaching assistant appointments, so not readers and tutors, have to go through a semester-long pedagogy course. And I’m sort of in the faculty rotation for teaching this. And that really is, for most of them, their just in time training. It is a course, you know, it is sort of a half-credit course, so it’s not, you know, it’s certainly not as intense as most of our computer science courses. But there is some training, and I guess what I’d say is that at some level, just like no one learns computer science by osmosis, right. It is based on practice. There is a lot of emulation that you see of other TAs. But I do think at some level a place for training is, you know, is really important. It could be just in time. It could be, you know, pre-course.

Michael [30:42] What I do think for smaller programs is the group of lab assistants. You don’t necessarily need a four-stage pipeline. You don’t need lab assistants and a mentoring program, and X, Y, and Z, right? But I do think some way of letting students see and test and getting involved in teaching beforehand is a really helpful thing because that becomes, I mean for us when we have 3,400 applicants applying for course positions, that’s not actually 3,400 students because most people apply for more than one course, but it is sort of 3,400 decisions that need to be made in aggregate at least.

Michael [31:23] Seeing that teaching experience somewhere is a big part of this. But what I would say is if it’s not the course lab assistants, I think that’s great because if students want to understand that course or they want to understand you as an instructor, right, they join that specific thing. But it could be a campus tutoring center where students come and get drop-in help, where students who want to be tutors, right, have a place to go beforehand. And I think that that works out pretty well. I mean, our largest courses have north of 100 academic interns in the fall semester. One thing that I would say is that with COVID, while things moved to online, we did not have a good support system for academic interns. We didn’t have a good way of getting academic interns in Zoom rooms. I mean, they could join, but the thrill of walking around a lab and sitting down by a student and talking through a problem with them just did not exist the same way remotely. I think the main thing is students, I mean there’s lots of reasons that you might love teaching, right, but going into a classroom, sitting with the student, helping them have that moment is something that when you do it enough in your first semester of working with students, that builds up real motivation to keep teaching. And so I think that’s been the thing that is sort of the most powerful, is you hook students pretty early on in their career and they want to keep giving back. So it doesn’t have to be a formal program. It doesn’t necessarily have to be each course does their own thing. But I think some form of low stakes, let’s try it out is really helpful for a lot of students.

Kristin [33:09] OK, so if you were the head of must start TA program, that’s how you’d start?

Michael [33:15] Yeah, and I would start it pretty simply, right? What you need is to start a lab assistant academic intern program is, you know, students are giving their time right, so they should get something in return for that. And so create a course code. But you have some place to enroll. You know, usually Google Form with like, how many hours per week do you want to show up to course sections? Which sections are you available? So someone does have to do the work of mapping and assigning things out.

Michael [33:45] And then what I would say is you do set up some time. It doesn’t necessarily need to be an hour a week, but an hour at the beginning of the semester and an hour maybe every other week or a couple of hours every few weeks to check in with them and say, here’s where this course is going. Do you have questions about just helping students? Here are the things due. And you don’t have to build this all from scratch, right? Colleen Lewis’ CS teaching tips are a fantastic resource to, especially some of the in-lab, poster checklist-style resources that she’s built, that have five or six bullet points on them, like that’s a really great place to just start with. Hey, you’re going to be in a lab section with other students. Here’s some really big do’s and don’ts. Try and keep this in mind if you have questions. You know, here’s how to reach us.

Michael [34:38] But I think in a lot of ways, starting low key and just letting it build up over time has really been this success. We’re now, I want to say at Berkeley, you know, everyone’s enrollments drop post dot com boom, so somewhere in the early 2010s things really start picking up again, that was when we redesigned CS61A. And at that point, with enrollment growth, things just happened over time. Like, we started out with a culture of some lab assistants. And so I do think you can start small, you know, automate only when it becomes necessary. Right?

Kristin [35:11] Yeah.

Michael [35:12] And so that is probably where I would start. But along the way, I would also say the other piece that I would remind everyone, and I think that, I guess I should just not take this for granted, but I do spend in staff meetings an hour a week with my team of TAs. So part of this time is logistics, but a good bit of time is me giving them advice on how to do X, Y, and Z, right? So when you answer a question in lab, how do I approach this. When we’re talking about exams, feedback on their work, when we’re talking about how should a course run. And making time such that within your core staff, they have ways of getting feedback. And after talking to other instructors, I always forget that this is a thing that many of us do, but not everyone does. And it is absolutely critical to having a team of people who are all on board with each other.

Kristin [36:03] Oh yeah. I recently actually started a practice in my TA meeting where a couple of them, not all of them, because it’s too many of them, they share either a win or a “I would like feedback” on how something happened with a TA kind of experience with a student. And usually, it’s a “I would like feedback” on how I could have done this better. I’m hoping soon some TA is going to be like I had a win and this is what happened. But right now it’s very much a they explain what happened and then I often get feedback of like, OK, this is how you could have done it differently, this is how I would have handled a situation like that. But it also kind of varies on your teaching style because I do this element of like humor with some of my style that doesn’t necessarily translate to some people. But let us close out with first, TL; DL too long didn’t listen. What would you say is the most important thing you want our listeners to get out of our conversation?

Michael [36:58] Yeah. So I think there’s a couple big themes and one of them is sort of right what I said then, which is believe in your TAs and give them the responsibility to make decisions. It’s OK if they make mistakes in their learning because they’re learning too. There’s never been a mistake in a course that I couldn’t recover from in some way, right. Some of them were harder to recover from, but most of them are really just about sending an email to students. So I think that’s one thing. Undergrads are incredibly capable. They’re smart, they’re, you know, they want to be involved. The second thing is, I do think training is the most important thing, is to remember that, you know, I haven’t talked so much about expert blindness, but is to remember that we learned to not, to try and address expert blindness in teaching computer science. We sort of train ourselves around those things. It turns out that when we’re training TAs and we’re working with students, we have a lot of expert blindness about what it means to be a good instructor. And so training in some form is key for us. That at Berkeley is in large part, this pathway from academic intern up to TA up to head TA to where they are modeling what other, what their peers are doing as teachers. But it is also in large part, just in time training where we have courses for that. And you don’t need to have everything all at once. But I do think it is important that along the way, I mean, not just undergraduates, but, you know, graduate student instructors, first-time TAs, that they have a space to get feedback, to ask questions, to see what good teaching looks like. And that can take many forms. But training is key. And if you give them training and you empower them to make decisions about either labs or homework or pacing or make slides for discussion section, you know, they’ll find a way of involving themselves in the course. And over time, I think this leads to TA a flywheel of interested TAs inviting more interested future TAs into your courses.

Kristin [39:11] Awesome. Well, thank you so much for joining us, Michael.

Michael [39:14] Yeah, thank you. This was absolutely fun.

Kristin [39:17] And this was the CSED Podcast hosted by me, Kristin Stephens-Martinez, and produced by Amarachi Anakaraonye. And remember, teaching computer science is more than just knowing computer science, and I hope you found something useful for your teaching today.


Be sure to follow us on Twitter and Facebook, and subscribe on Apple Podcasts, Spotify, Amazon Music, or wherever you get your podcasts.