Susan Gentry

[MUSIC]

Hello and welcome to another edition of Undercooled, a materials education podcast. Today, I'm here with a special guest, Professor Susan Gentry from UC Davis. And I've known Susan for a long time, cuz she was a graduate student here at University of Michigan. And so I've known her and I've also gotten reacquainted with her many, many times at education conferences, most recently at the NAMES conference last summer in San Luis Obispo.

And of course, our little plug, we're gonna have the NAMES conference this year in Ann Arbor, Michigan, and I hope all of you can make it. So Susan, why don't you tell us a little bit about yourself and how you got into teaching and what your position is at Davis. Cuz it's a little unusual, but it's really important, I think. So go ahead. Yeah, so I'll start from the beginning. So when I was looking for doing a PhD and started my PhD, I was interested in going into industry.

And so one of the things that interested me about Michigan, going to grad school at Michigan and my advisor was strong industry connections. Sort of through the process, I finished my PhD. I wasn't 100% sold on industry cuz my 3D printing machine kept breaking on me. And so I was trying to look for different options. So I ended up sticking around for another three years in Ann Arbor. I did a postdoc in phase field modeling with Katzio Thornton.

And so sort of during that, during my PhD, I enjoyed my teaching experience. And so I was interested in those types of positions. I was able to get a partial teaching appointment my last year at Michigan. Katzio helped me teaching the undergraduate lab class at Michigan, while I was also doing the postdoc. But sort of through all of this, I was still like, do I go into industry? Do I try an academic job?

And ultimately where I ended was I felt like the teaching job was this thing that I would always feel like, what if? So I started applying for the teaching positions, figuring I'd try them. And if not, my fallback could always be to go back into industry. Well, not go back, but to go back to my original plan of going into industry. And so I was looking at lecture positions and things like that. I wasn't interested.

I don't have a big enough drive of this is my research problem I want to solve to be a research professor and to have to run a research group and write grants. So that was not interesting to me. And so when this position came up at Davis, I was really interested. So it's morphed titles a little bit. The title series though that I'm in is a professor of teaching series. So it's equivalent to tenure track here at Davis. So I'm an associate professor of teaching. I have the equivalent of tenure.

And so I'm treated like an equal. I'm a member of the Academic Senate, treated like an equal in my department. I say that this comes with the honor of serving on three hour qualifying exams. [LAUGH] But I get to sort of meet with students. My teaching load is one and a half times that of the other faculty. So I do teach more, but then I'm evaluated more on my teaching innovations, my teaching excellence with a smaller aspect of my scholarship.

And for my scholarship, then I've been able to go to these education conferences, present on work that I'm doing. I've worked on computational modules for the curriculum, looking at student learning. Now that I've gotten tenure, I sort of tried integrating more active learning and different types of strategies into my graduate classes. And so, yeah, it's just been a really fun time to meet with students and also to think about what is good teaching and how do we help all of our students?

And how do we help them learn and grow? Fantastic. Yeah, I remember your first few talks at NAMES back when you were just starting were all on different computational models. And then of course I noticed in your publication history, you started drifting towards during the COVID days of how to do distance learning. And now you're back doing equitable teaching and active learning in all your classes. So I hope we get to talk about all of that.

So do you want to lead off Tim and ask her a question about one of her topics? Sure. I have a two parter here because the first thing I want to know is digging further back into the past, what caught you into material science and engineering in the first point? What's the origin story here? Oh, the origin is I'm the rare bird who came into material science as who had material science as their degree listed when they started college. I liked chemistry. I liked physics.

In high school, my sister had suggested to me she had gone to Carnegie Mellon. And so she had suggested, oh, you might be interested in this material science thing. I was also at one point interested in chemical engineering till I visited a different university and realized that it was like big pipes and they were like showing off their like two storey like facility. And I was like, I mean, I liked chemistry and physics, but this is that's not what I thought chemical engineering was.

And so just a little bit of trying different things. But I just like the integration of the chemistry and physics. And I took an intro to materials class my freshman year and liked it and I've stuck with it ever since. Yeah, I feel like every time we ask someone a question on the show, the answer is I thought I wanted to be a chemical engineer, but then I really didn't. It's just funny how much that keeps coming up.

But back to graduate education, since that was something you were talking about in graduate courses, especially there's so much rigorous quantitative content, so much of the sort of analytical problem solving that has to be taught at a really deep level in graduate courses, especially compared to say the intro materials class. What are things that you're doing with quantitative problem solving in an active learning sort of way in these graduate courses? What does that look like for you?

You know, it's it's an emphasis, but I think sometimes we get too caught up on thinking that our students actually are getting the quantitative problem solving skills that we think they are. I like to sometimes reckon. So I'm teaching a graduate thermodynamics class is what I do this in. And I've seen what I call the hope and pray approach to science to solve problem solving, which is like they just rearrange their equations and they get an answer.

And if they get the answer, like they're very relieved. As opposed to like being able to explain their answer and, you know, can they reproduce this on an exam, you know, or can they do this on their own? Do they know why their answer is right or wrong?

You know, those are the skills that I just feel like easily are still getting bypassed, you know, and if I have an exam and sure, it's a really hard exam and then the score is a 60 and, you know, half the students didn't get one of the questions like, what is that telling me about their their learning? And so it's just these questions of, yes, we expect it to go deeper.

But I also want to make sure that my students like have these fundamental this fundamental knowledge, have these fundamental skills, because, you know, they're sort of rearranging equations, you know, just trying to figure out the right answer, hoping that it works out like that's that those skills aren't going to serve them well as they move forward in their graduate career. So I noticed that in your one of your talks, you put up very prominently a picture of grading for equity, the book.

And it seems like you followed a lot of the examples in that book. And also the specifications grading book by Nilson, I think her name is. And I've been reading similar things. I just finished reading a book called Grading for Growth. I don't know if you've heard of that, the Robert Talbert Road. And there are very similar kinds of things. But obviously, from looking at your work, it seems you believe that the grading strategy strategy is intrinsically involved in the learning of your students.

So can you talk a little bit about that? Yeah, and I you know, it's wanting their their grades to represent like their their knowledge and their mastery of the course content as well.

And not to be like the thing that struck me about there's this example in grading for equity where like depending on how we are waiting different categories of like homework versus participation and this and that and their exam scores like how much am I sitting there fiddling with this to like when you sort of can look at your students to be like they get it and they don't.

But how can I sort of make my grading scheme represent more realistically like what do they know and and like what is their mastery? And how can we also then get it so that we don't have to like nitpick about like this was an 82 or an 84 on the homework or this presentation like you know they're going to my students are going to go on you know if they're in the PhD program at the UC Davis they take their an oral preliminary exam at the end of their first year.

And so like that's on a pass retake fail basis. And so like let's stop getting into like these little nitpicky arguments of like you know 82s and 84s and like let's focus on like you know are you demonstrating these skills are you not. And it's hard to get there with the grading team.

I'll be the first to admit I'm not there yet but it's just been really getting me to think about like how do we get students to like recognize that they're you know they and they can't just wait for the you know depend on the class to do poorly. But like I don't care if the class everyone in the class did poorly on something like it's they're responsible for you know mastering these topics. And how can we sort of then link that to the letter grades that we we have to get.

So all that sounds great. But of course the devil's in the details. So how do you actually assess whether they've learned it or not. I noticed you even went to oral exams or you're thinking about going to oral exams. That's very time intensive. So what's going on with all that. How are you going to figure out what they've actually they've mastered a topic. The reality is that some of this mastery is still for me just having to be is being done like on an exam setting.

But then also just trying to give them opportunities to like practice mastery as well. And so that they're they're gaining those mastery skills so that the exams aren't this like big thing. But that exams are just a way for me to like to check. We're not there yet. But I really tried to give them opportunities to give get lots of practice for those. Yeah I really this is the thing I really struggle with is how do you I like the oral exams. We could talk about that for for a while.

But how do I do this and make it time timely for me. Also I have some remote students. And so how do I do this in a way that works for them as well. So they're working full time and are watching lectures on their own. And so they can't necessarily participate all the time in class every day. And so how do I create this this class environment for everyone. Yeah. And so there's. Oh that's a really interesting idea. If I can ask a little more about that.

So it if you have in person students who you're doing these sort of you know doing different activities in class with your in person students but then you also have remote students. What are some strategies that you're trying to give you know to give some sort of comparable valuable experience to the remote students.

And I ask because you know in 2020 2021 when I had these hybrid courses as well I was really struggling to actively engage my remote students and never quite found out what worked for me in that. So how are you doing that. Yeah. So one of the strategies that has worked I can you know they do work a lot of them work in the same place. Most of them are through a national lab program. And so you can still give them like group work to do or group homework assignments.

But then also some of this problem solving. I'll make one of the groups. It's my classes video recorded and a video recording classroom and so I'll make them either go up to the board with the microphone or go use our document camera. But so that way they're getting to see sort of other students working working through these problems if I'm doing problem solving and I can set it so that the microphone isn't projected through the classroom.

But that way the microphone is getting picked up the audio is getting picked up for the recording and so that they can sort of be there. One thing I did two years ago I skipped it this year and but I think I'll go back to some of it is to also have the students do videos where they have to explain this the solutions that they have. And so the remote students if they weren't in a group they had they could have easier problems if the end of all students could do these.

If you were working as an individual there was slightly easier problems group problems had slightly harder problems but that way you sort of are getting examples and videos of how to like you know do some of this problem solving. And see other students solving the problems like we skipped so much of this in graduate education we just pretend that they're supposed to go off and magically do this and so you know sort of saving time for some of that in the class.

And so that they can sort of see some of that and then they're still having to explain it to each other you're using a lot of. Lots of feedback right during class where you're able to see them actually do the work and then give them feedback on what they did and not count that as great right just use that to teach them is that what i'm hearing. letting them teach each other.

Well yeah so just I mean if part of it is is working through how do you start a problem you know you sort of start some of these problems. A phase diagram you give them common you give them the free energy curves and they have to do the common tangent construction to generate a phase diagram and you know students don't even know where to start or how to explain that and so. getting them to practice some of that starting and explaining but then also trying to make sure.

When we would do these problem sessions that then you know things were getting recorded for the distance learning students, so I had seen one scheme to make oral exams more scalable. And the idea was that. If students didn't get a good grade the first time or a passing mark or a mastery mark whatever you want to call it. They would have an opportunity to do it again and so they then to do it again the professor would tell the student I will let you.

take this again or take parts of this again, but you have to come to my office hours and do it in my office hours and let me talk to you about it. That way you only have to deal with the oral exams for the students who actually need the help. You don't have to waste your time on all the students who actually learned it so I thought that was a really clever trick that I read about in a book grading growth. And it's a neat idea.

Yeah, so I did so I would I sent this students a problem 30 minutes advance they had some time to work on their problem so they sort of got that problem solving time without me having to be there, and so then they just had to sort of. explain their problem, it was just I need to figure out better the logistics of how to how to do some of this and make sure I. don't give them too many time slots during the days so that they have to like they have to stack themselves up.

How many students are in your class typically. I had about 20 students so you can probably do it for 20 I mean I just taught a class with 140 students in it so that's kind of hard yet you know oral discussion just communicating talking to somebody you get a real sense right away if they understand it. And that's what my my last course was a total disaster. Everyone got days because they all did what I told them to do but.

But when I walked around the room and I talked to all the teams it only took me about 15 minutes and I realized not a single person knew what they were talking about and it made me really sad so I'm now having to read this. I wish there was I mean. yeah the gold standard for measuring learning isn't oral exam. Too bad it's not scalable that's the biggest problem. I know that there's a group that had that I think it's a UC riverside it's definitely one of these season in mechanical engineering.

And they've been looking at oral exams they started them during during COVID and so you know coming up with rubrics and things so that they could do this with. You know, in some of the you know with 100 students and things like that sort of they wouldn't do all their exams this way, but at least to do one or two check ins and then also wage to you know frame it for the students. As like learning opportunities rather than this like panic you know really scary thing.

You know so there's people who are interested in doing that and then it's also just trying to make you know keep in mind that like time goes into grading and so you know opportunities to use these sort of strategically I think are really really interesting. yeah two ideas in there that really stood out to me one was giving the oral exam almost as the retake opportunity it sounded like to give that extra time and attention to the students who apparently need it most so that's really cool.

But then, as you were Susan as you were describing the implementation that you're giving the students the exam problems just a short time ahead of the interview I'll call it I thought that was pretty interesting because I do a similar thing with my lab classes where the students have to give impromptu presentations and I give them.

just 15 minutes to prepare okay here's your prompt here's what you're going to present on you have 10 15 minutes get your thoughts together and then give like a you know five minute whiteboard talk because one of the professional skills that I feel like we value so much is the really thinking on your feet the extemporaneous oh someone just asked me a

question and I need to be able to explain it even though maybe I haven't thought about this topic for a couple years and so scaffolding is the way to do it. I think it's also important for them to work through their nerves I had students there and they sort of knew what was coming, but you could you can see them visibly you know their hands shaking but like you know as a professor i'm used to that.

you know I'm I'm not put off when students do that I know that these are really nervous like high pressure situations it's just that you know by ignoring them weren't they're not going 15 minutes to prepare for something like that but you know the more you do it the more you get used to it and like you find if nothing else ways to manage some of those some of those feelings and I don't judge them like I said you know I know I can see them shaking and it's like I'm not going to get used to it.

you know we just keep moving on. So I saw in some of your other talks that you use learning assistance a lot do you use those in the graduate courses as well. I won't say that I've used them a lot I've used them for a project based class. I've used them for a project based class. I've used them for a project based class. But I've started exploring how to use them more.

I don't use them in a I don't use them in the grad class and I don't know how I could use them in the grad class in you know similar to Michigan we have to be careful with rules regarding especially now that all of our our teaching assistants and research assistants are unionized And so it's really unclear even in other situations of opportunities for graduate students.

I mean they could sign up for a one-unit class and do this but it why would they and so it works a lot better with the undergrads and so I've had this junior level project based class where I had senior you know I get senior students who then sign up to do it and then they get they've been getting course credit and they get to do it.

They get they've been getting course credit and then I'm very careful to delineate their responsibilities versus the responsibilities of the T.A. or the greater but then you know these are senior students who need leadership experience you know or want you know to try you know to get more involved in the department so I'm sort of I'm thinking of using some of them next quarter and teaching intro to material science class and and so I'm probably gonna get like two or three

to help me out with some of the lab opportunities or some of the office hour times. - Can you talk a little bit about the, you know, you were talking about how you were starting to use videos that graduate students would make to teach a muddiest point or something like that as part of the instruction. Now that you've done this for a while, do you see impact of that? Is that something you're gonna continue to do?

- I mean, I just like, you know, it goes back to the learning assistance, it gets back to the graduate education. Having people have to explain how they solve the problem or answer a question is just such a good experience for them. So I've done these videos where they would do like, in my grad class they have to answer like a common undergrad problem of how can entropy increase or decrease, you know, how can be, I calculated the change in entropy and it was negative.

And so just getting practice, having to give words to those answers, and then also just trying to build some of these up as a repository so that they can watch all of these. You know, it's one thing, I've been teaching for a while, but sometimes I struggle with answering questions in different ways.

And so to hear someone else answer it in a different way, come at it from a different angle, you know, I think that's why it's so important in these like peer to peer instruction opportunities for learning, you know, to hear the explanation from someone else or to hear the explanation from someone who's had to master the concept more recently than I have.

You know, I took Intro to Material Science as a freshman in college, back in 2005, you know, and so to have someone who had to struggle with these questions more recently than I have, sometimes they can just answer it a little better. And for some of the things that you're starting to do, you had written that you wanted to do in-class problem solving and student-led instruction of problems. Is that for both graduate and undergraduate?

Yes. I say with the hesitation only because I teach a lot of different classes, and so it's also the reality that, you know, how I have to come up with different activities for different classes. This quarter I'm teaching a lab class, and so it's just, you know, it looks a lot different. And so for me, it's been thinking about my educational, my teaching as slow innovations. And so like how can I start trying things out?

I've been trying out a lot of these things in my grad class because it is small. It's only 20 students and they're a little more forgiving. So that I can see how things work and then, you know, bring more of some of those things that I'm doing, bring them into my bigger undergraduate classes, and just, you know, try out different things. Again, though, you know, 100 students in my intro class, you know, versus 20 students in a grad class, sometimes it's just the mechanics are a little different.

So why don't we take a look at both of those? Let's start with the 100-student undergraduate class. What does your class look like? Do you lecture a lot? Do you break students up into teams? What kinds of activities do you do for a very large class? This is where I wish I did more. So the honest answer is that it's mostly lecture, but then I try and, you know, bring in, you know, once a week. I try and one out. It's three hours of lecture a week, and then the students have a lab section as well.

And so, you know, trying to take advantage of those when they are in class to, like, bring in, you know, I have designed some activities for them to do in class, like that, you know, take maybe half of a class, you know, think pair shares occasionally as well. But trying to sort of use that as additional opportunities for them. And what does the space look like for the 100 students? Is it a traditional sloped lecture hall?

Yeah, you know, that's-- they've been building some new camp-- they recently built a new building that has more of those, you know, long tables with chairs and things. But often when I-- I haven't looked at which classroom I'm given next year, but next quarter. But a lot of times, yeah, it's been those sloped lecture halls with the mini little desks and, you know, these ideas of turning around.

I mean, this is where, like, think pair share or, like, do a computer activity but work with your neighbors, you know, sticking with things like that as opposed to things that involve completely moving around. We have-- like I said, we have some of those spaces, but the challenge can be making sure that we're getting-- we're getting assigned those spaces from the register. So if you got a flat classroom with movable furniture and stuff like that, would you teach the course differently?

You know, it just starts to bring in more opportunities to have students work together. I don't-- you know, I like what you're doing in your class, Steve. It's just-- it's hard for me with all my other classes to have to, like, redesign that class to be an entire problem learning. And so if I think of it more, though, like, how can I get from where I am at now? How can I add in, like, one more activity a week? How can I, you know, expand that activity? That's what I see.

Where I see makes the most sense for me.

And honestly, I think makes a lot of sense for instructors is, like, rather than going all in and having to say, "I'm going to completely redesign my class next quarter," in addition to everything else I'm doing, like, if I can add one more activity in or think about how can I get them to do something in groups where they're having to explain things more, you know, get to those higher levels of learning where it's more interactive and problem-based and I can, you know, be bringing things in,

like, that's-- you know, that's my goal for my classes. And what about a graduate class? I think there's really good advice there for any newer instructors out in the audience that you don't have to reinvent the entire course every time you teach the course. And the incremental change is so much more manageable and it does get you really significant improvements if you're thoughtful about where you make those little changes to where they're needed most.

So what does your graduate class look like with just 20 students? Do you teach it differently than the undergrad? Is it lecture or do you do a lot of problem-solving in the classroom? So this is where my ideal schedule is-- so this one has four hours a week of class. And so my ideal schedule that I like to go to is aim for about three hours of content and one hour of problem-solving.

I've tried to, at different times, do my problem-solving only on Fridays because that sometimes has worked better for my distance learning students, sometimes to be able to attend live, if they're working from home or on a 480 schedule. But then-- so that's just been practically-- other times it's sort of then mixing that up. And so I think keeping in-- we have to cover thermodynamics and so there is a fair amount of lecturing in that.

But then are there then opportunities rather than only do Friday problem-solving to spread out the problem-solving on some of the other days so that it's more timely of where we're learning about it or you're getting students to start a problem and think about the hard concepts. And then once they've thought about those hard concepts, letting them go work on the problem overnight and come back so that we can not spend the time actually solving the problem.

It's interesting you have distance learning students in the same class. Do they attend synchronously but just over Zoom or is it asynchronous? So this is a program we've had at UC Davis for about, I don't know, longer than I've been here. I'd say at least 20 years with Lawrence Livermore National Lab where they used to watch class would be video recorded or sent over there.

And so now the program, what it is, is the classes are-- because it's a special agreement, the classes are in a special room that is set up for live-- where they can attend live but they're not required to attend live. You can require them to attend presentations live, things like that. But some colleagues used to make them come to campus for presentations. Now I think pretty much we're OK with virtual presentations but occasional live attendance.

Otherwise they're just expected to sort of stay up on the content on a weekly basis. But you could say break away from lecturing for 15, 20 minutes and have small teams of students work together on Zoom in like a breakout room. And even if they're sitting in a sloped lecture hall, they're still in a more intimate setting and that would be very inclusive with your distance learning because everyone's getting the same experience. And then you can hop from room to room and see how they're doing.

Have you ever tried something like that? I haven't tried the Zoom rooms because that would make the students then have to actually sign into Zoom when they're there as opposed to just being in the same room where I stop by the different groups. But I fiddle with the thing. Do I have them work on homework problems where they present their homework problems to each other on Fridays where they sort of need a set day? Instead of do this Friday thing, do I break it up on different days?

And so we only do like half the class. That's something I'm just actively trying and just trying to figure out what structure works for this class. And again, because this has the additional complication of the distance learning students, what I do in this class isn't necessarily going to be the exact same structure that I do. If I teach an upper division, if I taught our undergraduate thermal class, I wouldn't necessarily keep everything the same way. I see.

So something I've been wondering about with the graduate courses, which to be clear, I don't teach any grad courses, so my opinion here is purely hearsay. But I'm aware, at least at Michigan, that our grad students come from many different undergraduate fields, different engineering majors, different chemistry, physics, math even.

And so there's really not a lot of shared experience or maybe really any formal education at all in MSC that students have in their foundation when they start grad school in material science. So is that similar at Davis? And if so, how do you handle that? How do you handle the fact that you're teaching, like for example, grad thermo and some of these students maybe never even saw a phase diagram before? How does that work?

Yeah, you know, it's definitely a problem here at Davis and I think is a problem at many, you know, many graduate programs in material science sort of across the board from, you know, the top programs on down. We talked at one point about doing like a pre-grad school like boot camp where in one day or in two days, we were going to like tell them everything they were going to need to know. I felt like that wasn't going to work. We replaced it last year with a programming boot camp instead.

That one's more broadly applicable. And so what I've had to do is, or what I've chosen to do is I've created, it helped a lot because I got a lot more comfortable with making short videos during the pandemic. But on my Canvas page, I have like review pages of here's the things that I expect my undergrads to know about a phase diagram. This is like intro level.

Here's what a phase diagram is, you know, and so pointing students to those and making it clear in the first week of class, like if you, you know, watch these videos for review. The other thing we see even with material science students is their comfort and knowledge of math is not very good.

And so in thermodynamics, like knowing what the triangle, the triangle is a difference delta versus like, you know, whether you're integrating like versus, you know, the D, you know, the lowercase VD and the delta, like knowing that these, and how do you like, you know, how do you take a partial derivative? Like knowing some of these things are actually like we, they should know it from math, but they sort of glossed over it.

And so just making sure that like there are these resources there that students can go to. And then also I realized this year, I need to find ways to not force, like you can't force the students to go there, but how do you really encourage them? I had some supplemental videos early in the quarter where since I was doing some more problem solving, I was like, watch this video, watch this practice problem outside of class. I don't have to do this practice problem in class. Go watch it.

And like some students would get to the homework and they're like, I don't know how to do it. And you're like, did you watch any of the practice problems that I told you to watch? And so just making sure not only that the students can find those resources, like I've tried to make my, you know, canvas page clearer, but, you know, really trying to tell them like at this point, you know, I'm not going to do this, but there is a video for you. There are some resources here.

There are some practice problems. I've posted my intro to material science, like how to do the lever rule. Go do those practice problems. I have plenty of them. And so getting students to do that and especially trying to do it early in the early in the quarter, just because before things really get busy, you know, they have the first week or two, it's, you know, week and a half. It's slow anyway.

So that's a great opportunity if you haven't learned those things to spend some time and go, you know, catch up on what you do, what you need. And some students do really great with that. And some others just need a little bit more, you know, push and check ins. And so, you know, it's funny you talk about math. Our last episode is all about our problems with our students with math. And Tim has the solution. He's going to teach a new sophomore level math course for our students.

So you can listen to that to hear about that. I have the nucleus of a potential solution. Let's not oversell this too much. But I think it's really good. Yeah, I think in talking to our department chair, Liz Holm, who's been meeting with other chairs of the materials community, you know, at the UMC, she told me that every single program is facing this problem that they don't know whether it's because of COVID and students just didn't learn as much math or if this has just always been a problem.

But it's a real problem across all of our materials programs that our students just aren't at the level of math that we really think they need to be to understand the concepts and thermo and kinetics and things like that where you need partial differential equations. So I feel your pain. My other comment related to that, Steve, is I'll put a shout out, not a shout out, a call out to the community.

I think that someone should create a series of like an online class or something that's intro to material science for like. So Jim Shackelford has 10 things every engineer should know about material science. And I've looked at that, but that's like to applied. We need 10 things every new graduate student in material science should know about material science. Like here's the concepts that you might have missed if you were a chemistry student or mechanical engineering.

We're going to give you some fundamentals so you can get started. That's a great idea. MIT used that idea for math in graduate school. So they have a boot camp for math. And I went and enrolled in it and I looked at it. And in 10 minutes I realized this is terrible because you couldn't read any of the writing. They used some weird script that got pixelated. You couldn't even read it. And then they had these people who were teaching it who just didn't sync with me at all.

And they were writing on the fancy light boards behind the glass and all of that. But their handwriting was terrible and I couldn't understand what they were writing. And I was like, oh, I'm going to write this. I'm going to write this. It was a great idea. But it has to be implemented well to be useful. So I think that's a great idea. You should write a research proposal and start soliciting people to help you out. That would be great.

See if anyone would actually give me money or if it just becomes another one of my projects that I work on. And I'm not even sure if it's a good idea to do that. You could. And it's hard to do. I tried to have an open source textbook. And the logistics. I'm not even sure if textbook is the right thing to do. Maybe a series of videos would be a lot better. I just don't know. Maybe we need to train a personal coaching AI model on fundamentals of material science. And I'm happy to not volunteer.

I'm happy to not volunteer. Oh, you are asking a fun question. I just got back from a workshop, an NSF sponsored workshop on using large language models in chemistry and materials education. I have to admit, it is one of those that seemed like an interesting conference to go to. I didn't have a lot of experience. And then they asked me to give a three minute update. They are asking a bunch of people to give updates on what is going on on campus.

And I'm like, you know, we are not actually doing that much. But it was actually interesting. There were two camps. There were people doing a lot or had more of a programming background and thinking about fancy tools we could develop. And then there was all of us who were like, we even think about how we should use it more. And so it was really great, though. It was a two-day workshop. And so the second day was just breaking people into groups.

And so the first day we had brainstormed different deliverables or things that we could develop the second day. And then breaking up into groups to develop some of those things. And so to start playing with the tools a little bit more. So some people were proposing how you could use a chat GPT tool to sort of create study things for students or to give automated feedback on their writing.

And so we started looking at the group I was in was looking at how do you use it to help do a better job with figures. And so we specifically then started looking at how you could use it to generate Python code for you to do some of this data analysis. And so just to really think about in this case, I got fiddling with stress strain curves that I do in my intro class. And so just thinking through things like how could we, rather than thinking about how do we use it to like, for good.

And so in my intro class, you know, it's important for them to be able to they collect data, they do tensile tests in lab, but they need to be able to like, you know, determine not only the tensile strength, that one's easy, but they have to figure out how do you select the points for the elastic modulus and fit a line. How do you determine the yield strength and that like the cognitive load of having to try and figure out how to do that when I get some students in this class every year.

Who don't know how to use Excel. And so we're having to like, you know, get them up into the basics of here's how to use Excel. Here's how to like insert equations. Here's how to import data, you know, and then like, that's just a big jump that we're really expecting in this intro level class of having to fit a line to only a subset of points, move that line over by 0.2% and look at where that line intersects. Sorry. In theory, intersects interpolates, you know, more than one point.

In theory, intersects interpolates, you know, discrete data points. It never intersects some nice pretty curve, but these are discrete data points. And like, that's a lot of things that we're having tasks that we're having to get students to do just to analyze this data.

And so, you know, are there things like that that we should maybe consider about lowering the cognitive load so that students can focus on the material science or the things that we want them to learn and less about this like data analysis and maybe if they could generate some Python code for them to do it and they want to do it that way, like maybe there are some opportunities there. As long as you test whether or not it's telling you the truth. Oh, I mean, I could not.

I spent a while trying to get it to give me the yield strength correctly and it gave me a lot of wrong yield strength. And it's very polite when you tell. That one didn't make it into our examples that we were putting in, you know, but it's just realizing, I think, you know, realizing I need to take my head out of the sand. You know, I can't sort of sit here and pretend that it doesn't, you know, chat GPT isn't a thing. I know it's here and it's just going to get better.

And so, like, are there opportunities rather than to see, say, it's just going to be here for bad, but like, are there opportunities for it to like critique figures and, you know, give students some of that feedback to help, like, you know, manage our workload but also give students, you know, just in time, timely feedback. One way. So that was just interesting discussions to have about that. One way to get better quality answers is to only use a curated data set.

So at Michigan, they've got this thing where they can scrape your Canvas website, pick up all your recorded lectures, and that's the only place where the large language model looks for answers. So it's your own material. And so we've just set that up. And, of course, students can ask any question. And then at first, I was very critical of it because it's like, come on, that's not a very active learning activity to just ask a question and get an answer. And they said, no, no, no, no, it's coming.

It's not there yet, but it's coming that we're going to be able to respond with a question to a student's question and give them places to look. Well, that day has already come. They told me this two months ago, and they just set me up for it yesterday. And it is amazing. I ask a question like, what is an edge dislocation? And it comes back and says, ah, dislocations are a very interesting topic. They're used to describe, you know, deformation in materials.

But to answer your question, you should consider the following few things and think about it. And that's a much better answer for a student. It won't give them the answer, but guides them on a path to discover it themselves. So I'm very excited that that's here. So we'll see. But I agree with you. You cannot put your head in the sand. It's here. The box is open. It's out in the wild. We either learn how to use it or it could consume us all. Yep. We didn't throw away calculators.

We didn't throw away computers. We didn't throw away. Well, OK, some people threw away their slide rules. But, you know, it's a tool. We have to figure out how to use it well. Exactly. As students said, let's use it for good. My biggest fear with it is if it's going to end up turning assessments into, you know, because we need to make sure that students are mastering concepts.

My biggest fear is how do we keep it from turning assessments and just being like, well, we'll only do in person, in class exams or, you know, assessments like that. Whereas I used to give a lot of my ability to give assignments, they say read and summarize a paper as a part of their homework decreases. And so how do I, you know, assess some of that as well and get students to do those those learning activities?

While it's fraught with problems, my dream is that we can use generative A.I. to actually do the assessment for our students in terms of, say, an oral exam, like we talked about. Have you played with the OpenAI Apple, you know, the OpenAI app? You push the little headphones and you can talk to it and it talks back to you. It can hear. It can understand what you're saying. And very soon you get into a conversation, you forget that it's a bot. You think it's a human.

In fact, Tim and I were playing with it and we said, what's your name? I was like, no answer. And we said, can we call you Mary? And all of a sudden she says, well, you can call me Mary if you like, but my name is Max. And if it could get to the point where you could in the prompts that set it up, interact with the student to assess learning based on rubrics that you put into it, it might make an oral exam scalable.

Of course, it's going to take a lot of work to make sure that it's telling you the truth and not just hallucinating. In fact, in the thing I just got from the university or Mazey, they call it, there's a little temperature slider. And they said, you can think of it as temperature, but we like to think of it as the hallucination meter. If you put it up to two, it really hallucinates. If you say, what's the color of the sky? It'll say purple if you're on this planet or something.

Yeah. And you can adjust how creative you're going to let it be. And this is happening so fast. And, you know, they're going to learn how to that. They're not quite there for figures and graphs. They're not quite there with Greek and equations, but it's going to happen soon. So who knows? I mean, it can already take your exam. Liz Holm told us she gave us her advanced thermo exam. And this was then this was almost a year ago. Well, yeah, almost a year ago.

And it scored like eighty seven percent on her really hard thermal exam. So who knows? I think it's going to be very interesting. You know, I saw the data of how they're bragging how well it can do on the AP exams and the bar exam and other. On the other hand, a lot of times it gets things dead wrong. Like if you ask it to explain the pedagogical value of midterms midterm exams, it'll just go off and talk about how wonderful they are.

And nothing's better than an exam. And then when you ask it, what about the research that, you know, people forget everything that they just presented two days after they take the exam, it'll go, well, there are a few studies that say that. But it's very funny. People also at the workshop got talking about different, you know, everyone here is a chat G.B.T. But there are other like A.I. tools that can be useful. I don't have them. I don't remember them.

But, you know, in terms of like, oh, you know, chat G.B.T. Very bad at doing a literature review. But there are sort of, you know, we're identifying papers on a topic, but there are other tools that are out there that are more developed for helping you find that. And so, you know, again, opportunities to use some of those tools to identify, you know, key pieces of literature or help us as researchers when we're, you know, we're trying to find out what we're doing.

You know, doing a literature review, you know, so there are sort of, you know, multitude of tools. And so not just necessarily having to think about, oh, it's chat G.B.T. or nothing. And so it's not just, well, chat G.B.T. versus Google Gemini versus, you know, this and that. But other tools that are using some of these these technologies. I think that's a great approach when I'm thinking about what tools do I want to use or teach with my students.

The question is always, do I use this in my own life, like my own professional life? If so, yeah, it's probably worth teaching because there are some tools that have helped me out and saved me, you know, a lot of wasted hours. So may as well start the conversation there. Looking at the clock, though, we've had a great conversation here. I think it's time to wrap up.

Susan, before we call it a day, we always like to offer the guests if there's anything you want to brag about any plugs you want to make just any any info you want to get out there. This is your free advertising platform. So take it away if there's anything you want the audience to know. I don't know that I have anything to advertise. I just encourage you to get involved with the materials community.

You know, there's a variety of conferences and and things I'm involved in the American Society for Engineering Education. We have a materials division. There is the North American Materials Education Symposium NAMES that, you know, Steve was talking about at the beginning.

I'm also on the TMS Education Committee, you know, so just opportunities for those who are looking to to reach out and just to get more connected and learn more about teaching tools and teaching opportunities in material science. Fantastic. So I'm not going to be at the TMS meeting, but I think Tim is. Are you going to be there? No, I'm not going to be there. I'm not presenting and it's the last week of the quarter. Yeah, I'll be there next year when they move the Judson Symposium there.

But at any rate, I hope we're going to see you this summer at NAMES. Are you coming? I am planning an excuse to come out to Ann Arbor. I think in August is free. So yeah, that's great. We're working very hard to keep the registration costs low. It's going to be 250 bucks for the early bird. So that should make it easier for people to get here. Well, great. Well, thank you so much. And thank you to all of those who are listening to this. And so I think with that, we'll say goodbye. So thank you,

Susan. And talk to you later. Yeah, see you next time. Bye.