– Hi. Today I have the opportunity to chat with Jesse Stommel, who’s the author of Undoing The Grade, Why We Grade, and How to Stop. And Jesse’s a faculty member at the University of Denver in the Writing Programme. And, appreciate you joining us today, Jesse.

– Yeah, it’s good to be with you.

– Thank you. Hey, today we get to chat a little bit about generative AI. For those of you who have not heard about generative AI, just kidding, there’s nobody who hasn’t heard about this thing. So we’re gonna chat a little bit about generative AI a little bit in alternative assessments. And so I think we get started with this one. Just a little quick question for you, Jesse. Generative AI, people are, some of them are pulling their hair out and they’re all screaming about it, other people saying, “This is the best thing that’s ever happened.” So generally, just kind of get a tone here, do you see this thing as a potential being very good for education, or do you think we’ve got a big problem on our hands?

– I mean, I guess I would say that I think that the thing that’s good is the potential conversations that generative AI will raise about education. It’s one of those moments that forced us to ask ourselves, “What is education for? Why do I make assignments? What is it? What is student writing? What does it mean to own our own intellectual property?” So those kinds of questions, I think are really important. And generative AI is forcing us to ask those questions. I think that that where it could be bad for education is if we allow it to sort of put its tendrils into our work without having those critical conversations.

– That’s perfect. I do agree with that 100%. There is no doubt this is gonna impact assessment in a huge way. There’s already people scrambling about how to catch students who are cheating using GenAI. And for those of you listening, please be very, very careful about using any kind of programme doing that. But the bigger underlining question is just the whole idea of assessment and how generative AI may be changing how we assess. So, I guess, that’s the question is how do you think this is gonna impact our assessment or kind of alternative assessments?

– Yeah, I think, I mean, oftentimes what I’m talking about, this intersection between the conversation about GenAI and the conversation about assessment and grades, what really sits in the midst of that conversation is this notion of cheating or this notion of how do students do the work for our classes right? And what pitfalls, what sort of traps are set up for them? And I think that ultimately when you put these two conversations next to one another, it forces us to look at what cheating is, what plagiarism is in a very different way. Because if you take traditional notions of cheating, plagiarism, standardised assessment, and you mash them up with GenAI, what you end up with is a really sticky situation, let’s just say. And so I think better for us to have those hard conversations, rather than trying to imagine our definition of plagiarism, for example, can neat and tidily apply to what generative AI is spewing.

And, I think, we have a moment right now. You know, and this isn’t the first time we’ve had this conversation, we have this conversation when the radio lecture was invented, when the correspondence course was invented, when the learning management system was invented. We have this conversation over and over again. So, the conversation about a generative AI isn’t necessarily all that new. And at each of these moments we have, we have sort of a small moment where we can really probe and reimagine how we’re doing this work and ultimately how we’re gonna have relationships of trust with students. Because, what’s at the core of the conversation about cheating, sort of the best way to reduce cheating in your class, and I’ll say this anecdotally, but there’s also research around this is to develop trust and a relationship of care between teacher and student. Students are less likely to trust, to cheat when they believe that their teacher cares about them as human beings. And, so ultimately, how can we express that care? And I don’t mean to express that care necessarily as a sort of crossing of boundaries or a mixing of the professional and private, but ultimately, how can we express that care through our pedagogical approaches?

– Yeah, that’s fantastic. Well, and I also believe that the reason cheats, obviously the relationship is huge. The other reason that students will tend to cheat on things is they feel like they’re backed into a corner. If I’m gonna flunk a class, then you know, potentially I might plagiarise a paper or buy a paper because it’ll keep me from flunking, and maybe having these problems Students don’t seem to grasp, and we’ve done the same things in our own lives I’m sure, the consequences of doing something might be bigger than that. But, the reason I bring that up is because to me, I mean, you’ve written the book on ungrading, if you start removing some of the grades, it feels like you remove some of the pressure. And if you remove some of the pressure, then there’s not the reason to cheat. So it seems to me like ungrading is a perfect kind of combination with all of just generative AI. Am I off on that one?

– Yeah. Well, I mean, ultimately, if the stakes are so high that it creates a lot of acute anxiety, which is what happens if the stakes are too high, you create acute anxiety. And then what happens when you have that acute anxiety is you tend to perform less well on standardised assessments. And, if you’re in that sort of perfect storm of anxiety, you are more likely to look for some sort of help. And if you’re not getting that help and support from your teacher or from your institution, you might look to something like generative AI to solve the problem of, but what is the problem that you’re having right there? The problem isn’t, “I have to get this assignment done.” The problem is a sense of anxiety or a sense of stakes related to that assignment. And in some ways it’s also because the assignment doesn’t fundamentally matter to you.

And so, if you have high anxiety and the assignment doesn’t give you any intrinsic sense of satisfaction doing it, then you’re more likely to cheat. And I think I actually kind of question whether we can even call what people are doing with generative AI cheating. To me, it’s only cheating if we’ve talked really frankly about what the nature of the work is and how students are meant to do that work. If you just put a ban on generative AI, what are you actually doing? I don’t think that’s a sophisticated enough approach to something like generative AI, which is ultimately extremely complicated. Students have been using generative AI for years without us banning it in the classroom. Grammarly is a perfect example. Students are using tools like this. The thing that’s happened with something like ChatGPT or other recent generative AI is that it’s become impossible for us to ignore.

– Yeah, I totally agree with that. And I do think that people have missed that for quite a while. Things like Grammarly, you’ve been able to buy, students have been able to buy paper from paper mills for years. Students have had other students write papers for them. I mean, we’ve had all of those types of things happening. So, this concept that all of a sudden we have a problem, I agree, is not the issue. And I also happened to be old enough to remember when calculators showed up, and I bought a calculator that took four AA batteries and the faculty members were freaking out because I was using a calculator to do my homework. And they said, “You’ll never learn math.” We found out later that, you know, the actual computational part is not all that important. It’s how you think about it. So, I think in many respects, that’s gonna happen with AI as well. So, I love that. I am curious, it feels to me, and I guess I’m going in this direction for just a second, that there’s some inequities that have the opportunity. I don’t know why I use the word opportunity. I think we have to be very cautious about some inequities that might be exacerbated by AI. I don’t know, have you given that much thought at all?

– Well, I mean this is sort of a simple way of thinking about it, which I think is more of a metaphor for all of the possibilities for inequities. If you think about that calculator, you could afford to buy that calculator, what happens to someone who can’t afford to buy that calculator? When I look at tool, when I look at generative AI tools, I mean increasingly, sure there’s a free version, but the paid version gives you access to so many other tools. And what’s also interesting is that when you’re using the free version of some of these tools, I think it’s more likely to be flagged as the inappropriate use of AI. Suddenly you go to some of the paid tools and they work better. And it’s more difficult to notice that generative AI is being used. So, it’s about access to the tools and the technologies, but also who’s most likely to get caught using these technologies.

And, also, the thing that’s interesting is that, I mean, when we’re talking about something like equity, when you read a student’s writing for example, what are the things that flag it as AI? And, could you, when reading a student’s writing flag second language learners, dyslexic students, lots students who are using language in a way that feels slightly odd to a sort of a neurotypical reader, they might flag those as potential AI infringement, when they aren’t. And so, ultimately, thinking about, lots of students write in different ways using different tools. And what we need to do is talk to students about, “Hey, what tools did you use to write this? What are you needing to do to be successful when trying to do writing for a class?” And, also, thinking about our assignments and creating assignments that are flexible. And again, allow different students to use different voices and different tools in order to achieve what our assignment is asking of them.

– I think that’s great. Well, it’s been for years. You know, I keep hoping we talk more about equity, one of my favourite topics. So, I think that we’re being pushed in that direction a little bit. The grading part has always bothered me a little bit. It just seems like there’s a great confluence that might be coming together with UDL, ungrading, AI. It kind of all comes together to where we have some really interesting opportunities right now.

– Yeah, I would say that the other conversation that we need to be having alongside of this is, I do research on food and housing insecurity among college students. And so, not just thinking about how do we change our policies around food and housing insecurity, but how do we change our pedagogies? If, for example, as an amazing researcher, Sarah Goldrich-Rab has found, one in two of our students have experienced food and housing insecurity in the last 30 days. And that number went up in the pandemic. And it’s particularly our students who are from marginalised populations who are experiencing food and housing insecurity at higher rates. If we know this about our students, how does it transform and change our pedagogy when we’re writing an assignment? Sure.

Think about the ways that we can kind of design an assignment so that not only is it not necessary to use generative AI, but also so that generative AI is sort of irrelevant. So that, cheating isn’t something that a student is likely to do on an assignment because it actually feels like it asks them to do something meaningful for themselves in their own lives. But also when we’re writing assignments, we’re thinking about who our students actually are. And the fact that our students do come from marginalised populations. We have second language learners, we have students experiencing food and housing insecurity. And so when we’re thinking about, “How do I design assignments so that it tricks students into not cheating and using generative AI?” No. How do we design assignments so that they’re actually designed for the students who we actually have in our classes?

– I think that’s fantastic. Well, I was gonna ask you the question of how do you get started if you’re a faculty member interested in this, but you, I like to call this plagiarism. You stole my idea before I had it.

– Yeah, I mean, the short version of that is read the syllabus as we’ve already written and read them from the perspective of our actual students and look at the words that we’re writing, the things that we’re constructing. I think that’s the place to start. And then what we’ll find is we’ll find that we’ll hit a sentence and we’ll go, “Whoa, look at the assumptions I’ve made about my students in that sentence. When we read our cheating, and so we’re thinking about cheating and plagiarism. Read your cheating and plagiarism policy. What assumptions are you making about students? What assumptions are you making about education? Ask yourself the question, “What is education for?” And then read your cheating and plagiarism policy. Sit those two things next to one another and you’ll fundamentally change who you are in that juxtaposition, I think.

– I think that’s fantastic. One of the things I like to point out to faculty members is pretend like there’s a thermometer as you’re reading the syllabus. And sometimes switch with somebody else is good too, but like, what’s the temperature in the room as you read the syllabus? And if it’s below about 72 degrees, you might want to rewrite some of those policies. So love that answer. We’ve run right up out of time, so appreciate you coming and chatting with us today a little bit about generative AI and some alternative assessments. Appreciate that, Jesse, and hopefully for those of you out there who haven’t read the book yet, grab a copy of the book. Undoing the Grade. It’s great stuff. But thank you for joining us, Jesse.

– Thank you so much.

– [Todd] Thanks. Bye.