Episode 7: Incorporating GenAI tools in a capstone course in health informatics with Simon Minshall

Download: 07-Minshall-Transcript

JIEXING

Welcome to the Teach Anywhere Podcast: Ed Tech Stories from real educators, where we interview faculty and instructors about how they use educational technology in their courses at the University of Victoria. My name is Jiexing Hu, and my pronouns are she/her.

THIAGO

I’m your co-host, Thiago Hinkel, and my pronouns are he and him. Before we begin, we would like to acknowledge and respect the Lək̓ʷəŋən (Songhees and Esquimalt) Peoples on whose territory the university stands, and the Lək̓ʷəŋən and W̱SÁNEĆ Peoples whose historical relationships with the land continue to this day.

JIEXING

With us today is Simon Minshall. Simon is an assistant teaching professor at the School of Health Information Science at UVic. With over 20 years of professional experience with imaging technologies and software development in the healthcare and entertainment industries, Simon brings a wide range of skills and knowledge to the school. Hello, Simon. Welcome to the podcast.

SIMON

Hello. Thank you.

THIAGO

Thank you for joining us today, Simon. Before we start chatting about your course, we were hoping that you could introduce yourself. Perhaps you could start by telling us about your journey here at UVic.

SIMON

My journey at UVic started about 10 years ago. This is not my first career; it’s actually my second. I used to work in the UK at Kodak as an imaging engineer, and I was responsible for looking after high-tech machinery that is used in the visual effects industry for films and television. And I ended up returning to Canada and I ended up back here at the university to try to upgrade my skills and carve out for myself a second career. And I ended up at the School of Health Information Science, where I now teach, and I was persuaded to join and do a master’s degree and learn a little bit more about something that I think is more important than the work I was doing before. I also felt a sense of duty to, you know, I’ve made my money at the other career and now it’s time for me to give back. So, I ended up doing a master’s degree, and I thought for myself, that will be fun. I’ll do this degree and I’ll take my time and I’ll relax into it. Not have to work. That lasted four weeks before I was hired and got embroiled in lots of other tech projects. So, I completed my master’s degree. While I was doing it, I taught a couple of labs. I don’t believe I was a TA at all. But then I taught a course in computer programming, which is my core skill. It was necessary for the school as we had students who were not doing well with the existing situation. So that blossomed into me teaching more than one course. And then came an announcement from the director that career plans were had for me. And then I threw my hat into the ring when a job opening came up to be an assistant teaching professor. And I’ve been doing that for about 2.5 years. I teach mostly technology courses and the capstone course, which we’re talking about today as well.

THIAGO

Nice. So, you mentioned you worked in Kodak, like with imaging, with an image company and image software, and when transitioning into academia or back into academia, what are some of your research interests today?

SIMON

Well, I still maintain a strong interest in imaging science itself. This is what I did my undergraduate degree in, and still it’s taking a while. I’m trying to close the loop to kind of resume where I left off in academia at the undergraduate level. I got my undergrad at the University of Westminster in Central London in the UK. So, I’m not doing a whole lot of imaging-based research, but it’s always there. And this concept of the visual image used to tell a story or used to explain something for scientific communication, for example, is always at the forefront. So, if you look at any of my syllabuses and lots of my course materials, it’s often full of imagery, whether I generate graphics or I’m using pictures and things like that. So, it’s kind of just simmering in the background in terms of research. Seemingly, I’m looking at visualization and I think I’m going to be specializing more in public health when it comes to visualization as well. It’s something that’s close to my heart.

THIAGO

Nice. That’s great. Thank you, Simon.

JIEXING

So, Simon, a couple of months ago, you reached out to LTSI to discuss incorporating ChatGPT-like tools into one of your fourth-year courses. You also kindly invited our team to observe the student presentations, and we were all so impressed by how well it turned out. So, therefore, we thought it would be great to have you here to share the stories behind the design as well as your experience with our audience. So, could you tell us a bit more about the course? What’s the course name? And what is the course about?

SIMON

The course name, it’s known by its code HINF 470. Its title is Trends in Health Informatics. Really, it’s a capstone course, and the subject matter within the course changes every year. In practice, it’s a blend between a seminar course and a directed studies course. Now, I say it’s not strictly a directed studies course because students end up working in groups, whereas in a directed study it’s typically one on one. So, it’s typically one of the last courses in the health information science student’s undergraduate degree.

JIEXING

That sounds like fun, having a different topic every year. So, could you remind us, is this the first time you taught this course?

SIMON

It’s not. It’s actually the fourth time I’ve taught it now most recently. It’s the first time I’ve taken quite so much of a risk introducing this topic. The course topic does change every year. It’s something that makes it interesting for me as the instructor because I get a novel experience every time. However, from the student’s point of view, it’s the first time they’ve taken it. So, they don’t get the benefit of seeing that, oh, what are we doing this year? It’s simply what they’re being asked to do. So, I can give you a couple of examples. There’s a problem with medical education from the health informatics perspective in that doctors and nurses, when they’re being trained, don’t necessarily get structured information management. They don’t get IT training, for example. And, you might argue, they don’t need it. But, sometimes, it can be quite useful because the moment they end up in professional practice, they are using very complicated computer systems, such as EHR pack systems and other acronyms that I cover in my first course that I teach, the introduction to the program. So, we did that one year. A couple of years ago, we did a similar topic around simulation in health informatics education. So, the idea is that, in the department, we have three simulation labs, one is a command and control center if you like, a situation room; another has a hospital bed in it full of buttons and switches and sensors and all kinds of uh magical things; and a third one is a mock up of an apartment where we have smart fridges, cameras, pressure sensors, all kinds of things in there. So, we had all these things. And so, we thought, why don’t we take the students and get them to look at all of this simulation equipment and locations that we have and use that in their health informatics education? Why ask them? Because it was their education; they’re in their fourth year. And we got them to imagine, well, what if you did have access to this stuff early on how would you use it? And so, that’s kind of the theme. We’re always building on what they already know. Typically, by the time the students end up in my class, they’ve done at least two, perhaps three professional co- ops. They come in as maybe high schoolers and they leave our program as professionals because they’ve had a year’s worth of on the job, professional experience. So, that’s great. And they make me so proud the students; they come in to my very first class in their first year of the program, and they’re maybe a little shy, perhaps they’re a little wary. They don’t want to ask questions. They don’t want to seem foolish. My goodness! By the time they hit the capstone course, they are on fire. They’re interesting. They throw ideas around, and it’s just great. It’s hard to keep them quiet sometimes, which I love. And it’s very suitable for a course like this because basically in a Capstone course, I’m at asking them to go off and do something. OK, with very little direction from me.

JIEXING

That’s nice. Thank you. My next question is, I wonder how many students do you usually get in this course? And who is a typical student in this course?

SIMON

Tough question to answer. Well, actually, the first part is quite easy because we have numbers. I have at least 25 students in the course. I should mention it’s a required course. So, I have a bit of a captive audience, and whoever’s going through the program and it’s their fourth year, they’re in my class. It peaked one year, and we had 45. This was challenging for, and you two were there for the presentations, when we do the presentations, they are small groups. I say we do group work, but really it’s groups of two or three. Now, who’s a typical student in the course? Well, that’s hard to say. It’s a semi professional. They’ve had a couple of terms worth of co- ops, so they have a little bit of uh industrial experience. They’ve done most of the courses; they know what they like. That’s very useful, particularly for this instance of the course. It’s an emerging, the typical student is an emerging health informatics professional.

JIEXING

That makes sense. Thank you.

THIAGO

So, yeah, you said that this course, this specific course, will change every term. It has like a different topic, or it focuses on something different, right? But this time specifically, and that’s the reason we are having this conversation, one of the reasons, right, is that you decided to use generative AI tools during this course, like to teach this course or for students to explore it, right? So, we wonder, how did you first become interested in integrating generative AI in your course or in your teaching? And what were you hoping to achieve by doing that?

SIMON

What made me interested in generative AI? Well, fundamentally, it’s an emerging technology. I mean, it hasn’t really… So, I should back up a little bit. Generative AI is a lot of things. It can generate the sound of a loved one who’s passed away. And this is useful for um certain scenarios in health care. It can generate silly pictures of mythical beasts. My son is working on one. It can generate soundtracks; it can do all kinds of stuff. So, in this course, I’m really only interested in its written output. OK. So, we, without using the product name, it is text-based chat bots specifically that I’m looking at. One of the things that interest me most about this is that in the previous iteration of this course, and this is really the seed, a group of students came up to me and said, they were in a group of 3, 2 of them were very nervous. They were completely pale on a video call, and that’s challenging in itself. They said, listen, we’ve done the work. My partner and I have done most of the work, but the third member of our group seems to have written everything with a chatbot. It all looks like AI. They were panicking. They thought that if they submitted this report, even though it was group work, that they would be kicked out of the program and they were fearful for their reputation. They knew, they felt deeply that this was wrong. Doesn’t that puzzle me? Because this is what we do, we experiment with new technologies. That’s why they’re taking my course. That’s why we were talking about simulations in that case. So, I came up with a solution for them; two solutions actually. One solution was to just step back, rewrite everything that this person did with AI. If the two of you want to take that on, that’s fine. We’ll deal with discipline later. And the second option I gave them was, why don’t you keep the AI portion? And I mentioned seed. And then write a reflection on what it was like to use that. What method did you do? How did you go about it? You recognized it was wrong, how did that make you feel and so on? And add that to your report. If you like, just for the sake of academic integrity, include, the tool was chat GPT that they happen to use, write the name of the genAI tool you used as one of the authors. So, they thought about it and they went away. And their conscience or perhaps their fear got the better of them, and they decided just to rewrite the part, which was fine. They ended up with a great report and everything ended happily ever after. But, it made me think, hmm, that’s interesting. There’s something here. Anything that would generate that kind of reaction, whether it’s positive or negative, there’s something there. So, it made me think, this fear. And I remember talking to the C AC, the Centre for Academic Communication about this, and their concerns were that if students start in year one of their program using generative AI tools, then how will they develop the skills to write and express themselves? Because they’re really just asking the chat bot to do it for them, and then there’s no critical thinking part. Unless you take that to the extreme and you become a chat bot operator, in which case your critical thinking is one step removed from the actual output. So, that puzzled me. And then I realized, well, but these are fourth year students; this is a capstone course. So, by now, I mean, they’ve had four years of us asking them to do written output, and there’s a lot of that in this program, why not use that and combine the two things? But, then the fear part came up, and I thought to myself, wow, you know what, some of the students are really uncomfortable doing this. What are the ethics of asking them to do something like this? So, I puzzled it around and I thought, well, what if we create a safe space and say, in this course, you are going to use a generative AI tool to do stuff and then do a critical appraisal of it. What would that look like? So, I realized at some point that I need help. So, that’s when I came to visit LTSI.

JIEXING

So, I wonder, were there any obstacles or barriers to getting to the outcomes you hoped for? And maybe you could tell us a bit of the steps you had to take to prepare for this. So, you mentioned that you reached out to LTSI, and maybe you could elaborate a little bit on that, and what are other steps that you’ve taken?

SIMON

Thank you. Yes, I will. It wasn’t my first step. My first step actually was to go and see the director. Because the nature of this course is that I am teaching course in which the curriculum kind of changes every year and curriculum level changes really require faculty buy in. So, my first step was the faculty meeting, and I spoke to the director and the rest of our team. Well, here’s what I think we should do. And the discussion came out. We talked broadly about generative AI, not just in this context but also in the context of what the students are going to experience at the beginning of their program and should we or should we not include this kind of training right at the beginning. So, I won’t go into detail there. When it came to the curriculum change for this course, should we do this kind of exploration work in here? It was actually quite well received; it was quite positive. And the fundamental reason for doing it is because we are a professional program, so our students at the undergrad level will become professionals in health care. And so, they are here to learn their technical skills, and I think we’d be doing them a disservice if we didn’t train them how to use this emerging trend of AI tools especially for writing because we’ve all heard stories of documentation burden in health care and how can AI help solve that. Well, if our informatics professionals don’t know how to use it, then who’s going to help people in the industry deal with it? Chances are they’ll have to learn it on the job, but why not start here? So, there was that. The other piece that bugged me is that, well, we are in an academic environment and this notion of plagiarism and the unauthorized use of an editor and the overuse of certain technologies is a major barrier, and this is what causes the anxiety in our students. So, the decision was taken to… I decided rather to go and speak to LTSI and see what we could do about this. So, I’m glad I did.

THIAGO

So, you talked a lot about the student experience and the concerns you had, you know, how students are worried navigating academic integrity and things like that, especially now regarding generative AI, so we were hoping you could introduce us or let us know how you had that conversation with students and letting them know that in this course they were going to get to use generative AI. Was that in the course outline? When students registered for this course, did they know it? Did they learn it like first time when you had the first lesson with them and how did that go?

SIMON

It was in the outline; it was specified there. It wasn’t mandatory. But I mentioned in the outline if they do use the chat bot, they may have to sign up if they want to get the value-added tier, where you pay a bit of money, you get a better product. So, the premium version of the software. And so, they were aware of that. But before I started even drafting that part of the outline, I did knock on the door and talked to LTSI, and particularly Hayley Hewson was the one who answered my call and was quite interested in what we were doing. And I expressed my concerns that, well, you know, some of the students are not going to like this and yet we have to do this somehow because it’s part of their training. How do we get around it? And so, we discussed a few ways. One way was to have an opt out. So, you can use gen AI to do this part of the work, and, if you really don’t want to use it, that’s OK too. You can craft a capstone project around describing what these things are and the technology, the theories and methods behind them without actually having to use one yourselves. So, I thought, that’s great. That wasn’t in the outline. Just the fact that the tool was available and could be used and might capture their data was in the outline. But the notion that they had an opt out, that I had to introduce in class. So, on the first day of class I asked them, well, who hasn’t used one of these AI tools? And, you know, there was 25 people in the class, not one hadn’t shot up, right? So, fearful or not, everybody’s used it. So, and it could be just as cursory as getting it to answer a question in the same way you would ask Google a question. Instead, you might ask the chat bot a question and play with it and see what happened. That notion of playing with it was really what I was going for here. This is what I wanted to do in a more structured way.

THIAGO

So, there was an option in place for students who didn’t want to use it. But then you are telling us that all of them were ok with using?

SIMON

In the end, yeah. And also, as we were going along developing this further, there was a second way to opt out. If the concern was that you didn’t want your typewritten, anything that you input on your computer going to this third party, there was another way of downloading a model. It’s fairly large, larger than your average Netflix show if it was resident on your computer and running the model locally on your machine. So, you could interact with the chat bot. It’s a bit slower when you do it that way, but no data goes up to the to the cloud that way. So, that was the second opt out. One group actually did want to do that, just because they wanted the experience. And that was OK by me. Nobody opted out; nobody chose the ‘let’s write it all by ourselves’ option.

THIAGO

And that option you mentioned like running, you know, the server, if I may say on a local device, that means that information that you enter into the chatbot is not absorbed or kept; it’s like saved on your local computer?

SIMON

It would be saved on your local computer, yes.

THIAGO

So, you’re not, let’s say, feeding the beast with your intellectual property, which is a concern, right?

SIMON

Exactly, yeah. Because who knows what you could type in. You could type in your own personal symptoms or whatever. And, if it knows who you are and knows you have symptoms, well, suddenly your insurance gets more expensive. So, I mean, that’s a valid concern and, yeah, they had more than one way to opt out. So…

JIEXING

Ok. So, my next question is, how did it play out with your students?

SIMON

Well, at first they were a little hesitant; they weren’t quite sure if I was actually telling that… they looked like they won the lottery, right? That’s the kind of impression I got. They thought, wow, really? I said, well, no, I’m not going to grade your work. That got them. I always play around in this course. Because it’s such a different format, they’re always surprised to find out there’s no exam, there’s no tests, but then it starts to dawn on them that really, I’m not going to mark what the chatbot generates. I was very clear right at the beginning. So, you’re going to get this thing to write a bunch of stuff, including a small thing and then a big thing, sort of a one-pager and then a five-pager, but I’m not very interested in those things. I want to know a lot more about your experience using it. How did you discover it? What settings did you use? Did you pre-program it? What prompts did you do? Did you only do one prompt? Did you prompt, and reprompt, and reprompt? What was that experience like? And I also got them to make sure they picked from the concept map a topic that they were already familiar with, particularly a topic they had already written on or that they did for their co-op, or that they generally have a keen interest in because that means they’re leveraging their existing knowledge, measuring what the chat bot does. And you’ll be able to tell whether the chat bot is giving a good enough account or not. I do try to adopt a playful tone in this because this isn’t, if you like, mission critical work. I want them to be relaxed while they’re doing this and not stressed out about being correct. So, I give them the option to imagine that what they’re doing is they’re in a small informatics team working for a hospital. And imagine that the entire upper management suite of the hospital is being taken over by a major bank. Ok? And all of the banking managers are now hospital managers. What do hospital managers, these hypothetical hospital managers, know about running a hospital? Nothing! Ok? And that’s, and that’s fun; that sets up a playful mode for them to think about, right? So, they might ask the silliest questions, how come the cups are made of plastic and not paper? Why are there lines painted on the floor in the emergency room of the hospital? What does that door mean? All this stuff. So, to warm up as an exercise, and this was actually LTSI’s suggestion, Haley suggested this, that we have a smaller deliverable in amongst all the other deliverables. And so, this was a memo explaining some aspect of hospital life to people who are managing the hospital but haven’t a clue about and pick something again that you know you like. So, I don’t think anyone actually did lines; somebody… one group did why the ICU, the intensive care unit, in the hospital where the most severely affected patients end up, on life support typically, why does it have its own pharmacy? Now, if you think like an industrial engineer, it’s you’ve got a pharmacy there, but you also have another pharmacy in the hospital that does most of the pharmaceutical work. Why do you have a separate one in the ICU? Why don’t you just merge the two and get the thing out of there and all that sort of thing? And so there’s very compelling reasons to that; You need to have medications quickly, you need specialist things in there and so on. So, they took that on and kind of rolled with it as it were.

JIEXING

What do you think was the biggest impact in your students and their learning?

SIMON

Well, I think the biggest impact on the students learning was gaining the knowledge that the chat tools produce copy that is really unusable. Out of the box, what the thing produces, it looks like words on a sheet, but when you read it and go through it, it’s a bit like the simile there, the lights are on but nobody’s home, right? There’s no real intelligence behind it. And then this is my personal opinion, and actually it’s what many of the students came up with as well, and they realize that it simply wasn’t enough. And so doing the memo was useful because most groups chose to ask one question, receive one answer, and plop that into a piece of paper or, sorry, into a Word document, and that was it. And so, I had to train them how to be critical and how to ask more of the AI to clarify, to omit, and to expand on certain parts, and actually reengage with it several times before you get a reasonable result. And I told them, I says, look, I’m not going to mark your stuff that the AI produces, so don’t go and copyedit it; just try to get it as well as you can with the prompts, and then keep track of what you asked it and what came out and what your impressions were.

THIAGO

So, you told us a bit about the student experience and how it laid out with them, but how about you and your experience with it? Maybe if you can summarize it in, you know, what was the most important lesson you learned with this experience?

SIMON

Well, I guess I learned that it’s counterintuitive. If I can train the students not to trust something, then I think I’ve done a good job in this particular case. One of my colleagues in the School of Health Information Science has based her research work on the notion that new technology breeds new kinds of problems and new kinds of errors. And, remember, the context here, despite the silliness of the set up for this particular problem, but, in health care, when you make mistakes, patients are adversely affected and they can be killed. And this is bad, and you certainly wouldn’t want it to be you. So, it’s important that we get that this knowledge has to be embedded like quick. There are some things that the AI is good at and there are some things it is not good at. And, as long as right now, we have a human in the mix, we’re the better for it.

THIAGO

Is there something you didn’t expect?

SIMON

Yes. One of the things I didn’t expect was the group’s problem-solving ability kicked in. Basically, the subtext of all of this is I’m asking them to evaluate this new tool. And, again, it’s a capstone course and I was hoping they would come to that realization as well that what they are actually doing is an evaluation. How do you evaluate this kind of tool? I suggested a couple of things like examine it for a reading level. Is it a grade six level? Is it a grade eight level? Whatever that means; I’m not an expert in this area. What they ended up doing though is they did a critical analysis, and several groups actually developed their own, effectively, their own research tools. They were simple, but they developed their own set of ways to measure the output. These are multi-access tools they came up with. So, one of the deliverables of the course is to produce a poster in order to reengage in the conversation with others. And so, the tools they came up with ended up being visualized as well. Also, by using the chat bot to help make the visuals. So, I didn’t expect that kind of full circle. They found it generally not as useful as it should be for writing, copy, but it was quite useful to use it full circle to evaluate and then to produce results on that. So, I couldn’t have been happier with the result; it was great.

THIAGO

Nice. That’s great.

SIMON

Yeah, a good learning outcome for them. And yeah.

JIEXING

Sounds very cool. So, you mentioned that the topic changes every year in this course. So, next time when you teach it, it might not be focused on generative AI, but say, if you’d like to do it again, would you adopt your approach in the future?

SIMON

Yes, I would. I would consider doing something like this again. There is already another topic in the pipeline. I don’t want to let the cat out of the bag at this stage, but I would like to do something like this again. And perhaps we could find another way to incorporate it into our curriculum; in a way I think we have to. So, one of the pieces of feedback I got from the course evaluation surveys recall that the deliverable wasn’t really the five-page paper that they got the AI to write or the one-page memo that they got the AI to write. The deliverable is really their reflection. So, the paper that they had to write about that was quite lengthy, and they had to work quite hard to get it done. However, the course evaluation surveys reported that they found the course workload for this course was much lighter than previous course evaluation surveys indicated. The previous ones were about average, so neither too light nor too heavy. This one was quite distinctly on the light side. So, that wasn’t necessarily unexpected because we’re getting the AI to write the majority of the work. And, if you’ve ever used it, if you’re listening and you’ve ever used it, ask it to make a five-page paper, it’ll figure out the number of words. It just generates five pages in a matter of seconds. It’s pretty effortless. And, if you were to take that and then start copyediting it and changing it, you might end up with a unique paper very quickly. Right? And that’s I think why they felt the workload was lighter. What would I change in the future? I need to may take steps to not to increase the workload, but if the workload was too light, that means I have more room to push the students to get them to learn more and dig deeper. How exactly to do that? I’m not sure. Perhaps my first step will be come back to visit LTSI again and ask for more advice. I expect more critical analysis. I did have a thought about this earlier and that I said to them, as I said to you today, that I wasn’t going to grade the outputs of the AI. Yet, I asked them to go back and prompt, and reprompt, and reprompt until they got something that was reasonable. Well, I’m not measuring that other than to look at their reflection and see what happened. We could come up with a scheme where we actually do look at the output of the AI. Was it sufficiently prompt engineered? So that it’s usable or not. We could look at that, but I don’t like doing that because then I’m giving them grades for something they didn’t write. So, again, back to LTSI for more advice on that, because it is a bit of a sticky area.

THIAGO

It is.

SIMON

It is a bit of a challenging area.

JIEXING

Yeah. So, we know many instructors are also interested in integrating generative AI in their teaching and in their course. So, what suggestions do you have to offer to another instructor who wants to try this out?

SIMON

I would encourage you to do it, first of all. Don’t shy away from it. I would say certainly keep the opt-out, whether it’s a technological opt-out like I discussed earlier, or simply some people may want to draw a line in the sand. OK? And say, well, I can use generative AI for, say, business writing. OK? But I would not use any generative AI, zero, never cross that line, for my own research work. OK? So, some students may still want to opt out of using the tool at all and concentrate on developing their own writing skills as we know that they take a lifetime to develop. So, I would keep that; definitely keep the opt-out. And don’t shy away from making the tasks more challenging because, as my results showed on my survey results, and they weren’t exhaustive, they didn’t reach saturation, I only got a few respondents, but, despite that, the ones that did respond did say the course work was light. Particularly, consider that this is group work. So, we’ve got three people sharing the output of gen AI. It’s, you know, I would make more use of smaller projects instead of one big one as well, and try to spread that learning through; not only spread that learning throughout many projects, but raise the stakes at each one.

JIEXING

And that mock up session that you had your students do probably have also helped to make students feel more familiar and more comfortable with the tools.

SIMON

Yes. Definitely.

JIEXING

Ok. Thanks so much for sharing the story of your course with us. Before we say goodbye. I would like to ask you one final question. What is one question you wish we had asked you and how would you have answered?

SIMON

I would say that what I’m hearing from my colleagues is that we are in need of embracing this technology. It still concerns me that if we do so at the early stages that we are subverting the development of our students as professional writers, or even academic writers for that matter, because they’d be relying on the tool far too early, I think, in their stage of development. It’s not the same for graduate students, but I think for undergrads, we really do need to not clamp down on it but warn them, educate them, tell them what skills they’re learning. We typically don’t mention that writing is one of the skills you’re learning as an undergrad. This just happens by osmosis, right? Well, that may stop. So, I still think there’s a, we have an obligation to choose when to do this sort of thing. And I think the final year of the undergraduate program is probably the earliest that I would recommend.

THIAGO

Nice. That’s great. Thanks again for joining us today, Simon. That was a great conversation, and we wish you all our best.

JIEXING

Thank you, Simon.

SIMON

You’re very welcome. Thank you for having me.

JIEXING

That brings the episode to a close. Make sure to visit our website at uvic.ca/teachanywhere, where you can learn more about teaching and learning at UVic. We will talk to you again soon.

• Hosts: Jiexing Hu & Thiago Hinkel
• Guest: Simon Minshall
• Technical Production: Jiexing Hu & Thiago Hinkel
• Transcript Preparation: Jiexing Hu & Thiago Hinkel 
• Theme music: “freesound1.wav” by freezound5 (⁠⁠https://freesound.org/people/freezound5/sounds/588258/⁠⁠). Available for use under the ⁠⁠CC0 1.0 Public Domain Dedication⁠⁠ at ⁠⁠freesound.org⁠⁠.