Confronting Higher Ed's Grade Economy: A Call to Action on AI

Welcome to another episode of AI Goes to College, the podcast that helps you figure out what in the world is going on with Generative AI, the topic that just never stops giving. I am joined, as always, by my friend, colleague and co host, Dr. Robert E. Crossler of Washington State University.Rob, you have a little professional update for us, don't you?

Yeah. It's the new year, and I am now not only department chair for the rest of the semester, but senior Associate Dean for Academic Affairs.

I'm trying to keep from laughing, knowingly knowing what Rob's life is getting ready to be like.

Getting ready to be like. You assume that yesterday wasn't crazy.

Well, that's true. That's true. I don't know. We need leaders like you, so I'm glad you're stepping into this role.All right, before we really get going, I also want to encourage people to listen until the end, because I'm gonna give you a little tip that could make a big difference in how you work with Generative AI.Alright, but our first big topic is based on an email that you sent to me, Rob, and correct me if I don't get this quite right, but you're saying it's time to start doing something. We spend a lot of time doing nothing or next to nothing and being who we are.We've had a lot of conversations trying to get things perfectly figured out before we start. And if I understood your message correctly, we got to stop with that. It's time to actually start doing something. Do I have that right?

100%.And what inspired this thinking is, because I've been thinking about this going into this semester is we've talked a lot about what are we doing with AI and trying to plan for it and to do things. And some people are, good job if you are, keep doing, keep pushing. But it's the first part of January, we get to enjoy college football playoffs.And if you reflect back, Craig, a college football playoffs. We didn't have one, actually.They declared a national champion based on how ballots were done, computer rankings, various different ways of doing things, and they got to a point where they realized they needed college football playoffs. It started with two teams, four teams, eight teams, 12 teams. It's grown, it's changed.My guess is after this College Football playoff season, they're going to tweak it again. Why? Because they decided that the way they were declaring a national champion wasn't perfect, wasn't working. A playoff was better.They implemented it. They didn't quite get it right. And they continued to make improvements. And I think we need to do the same thing with how we're approaching AI.Yes, we know we need to do something, and no, we don't know what that perfectly looks like.So following that lead of the college football playoffs and their iterative approach to trying to make it better and better and better, we need to do something similar, which is, if it's in my classroom, start doing something. It may not be what your class is five years from now, but it will be better than not doing anything and waiting until you have it all figured out.If you're at a departmental level where you can influence conversations at the departmental level, what's your department doing to be purposeful about embracing and adopting AI throughout the curriculum there?And then as you move up that leadership ladder, deans, presidents, whoever might be listening to this podcast, what can we be doing to making strides in those directions with an idea that as we learn how well what we're doing is working, we can make changes. And the other thing that's going to happen is the world of AI is going to change on us as we advance in this as well.So even if today we absolutely had the right approach, tomorrow it might need to tweak a little bit as well as we adapt to what the world is doing.

Absolutely. I'm glad you brought that point up, because if you hadn't, I was going to.These tools are constantly evolving, Their capabilities are getting better, there are ups and downs, but the trend is decidedly up. This is just going to be the way it is for the foreseeable future. I don't care.You can spend all the time you want, and even if you come up with something that's perfect today, it's not going to be perfect tomorrow. And we pretend that we always have everything perfectly figured out before we do anything, and that's never the case.

Yeah. And I would argue that there are classes that knowledge of certain topical areas is still as crucially as important as it ever was.And as you look at AI and what you're doing today, it's how do you ensure that knowledge is being learned in a way consistent with the fact that that AI is at everyone's fingertips, from a Google search where it gives you the AI summary, to the writing tips that you get from Grammarly to just the temptation of, well, what does Grok have to say about this? It's there and it's important that we learn how to critically think, it's important that we gain knowledge, and so on, so forth.So it's going to be A moving target at some level and pushing into it and doing something is going to help us to at least get. Get closer to that moving target. As we press into this, do we.

Really want to know what Grok has to say?

I hear from my students that's their go to AI tool for asking questions. So we like it or not. I think it's in the world of what we have to be aware of.

We pretend that we have things figured out and we never have to change classes. And that's just not the case.I mean, just in our time in higher ed, at least in my time in higher ed, I've seen it go from I didn't have a computer when I started my doctoral program to not having to carry change around to make copies of articles, actually being able to use your ID card, which was a massive breakthrough, to not having to go to the library to do your searches, to not having to go to the library, period. And then we have all the Internet stuff that came along, and it's just been one thing after another. I think generative AI, is that on steroids?But we can take the same sort of approach. Do something. Some of it will work, some of it won't work. Fix what doesn't work, keep doing what did work, get started, and fix it in the mix.But there are a couple of caveats there. One is what do faculty do to the potential hit to their teaching evaluations and potentially to student learning?

I think you're transparent about what you're doing and why, and talking with your leaders about the approach that you're taking and why you're doing it, I would hope gives you some cover to try and fail. It's the learning in the process and being transparent about it that I think every academic leader should be looking for in their faculty right now.So that's what I would encourage, is don't just go out on a whim and say, this is what I'm doing, and I'm not having that conversation with those in my sphere who will influence those evaluations and those different things.And if you're paying attention to learning outcomes and what students should be learning, if you have a semester that's off a little bit, well, a, how do you make lemonade out of lemons? How do you adapt a little bit to ensure that they did learn? And then how do you do it different next semester to fix what didn't work this semester?

Yeah, it's a tough problem, and I agree with what you're saying. I worry that the reality on the Ground is very different for a lot of people. You're an administrator who understands all of this.I was an administrator who was willing to take on smart risks and accept that risks sometimes come to be. But that's not always the case. And I'm not even sure that's the norm. I think it sounds like it is at Washington State.It certainly is the norm at Louisiana Tech, but it hasn't been other places I've been. And so I would encourage faculty to not let that make them the deer in the headlights and just freeze. But to acknowledge that that risk exists.Keeping your department chair in particular informed is a good start because often department chairs are experienced academics and they may help you avoid some potential problems. It's like, you might not want to do that.Maybe you can scale it back a little bit, or maybe you can figure out a way where it's not going to affect the student's grades if it doesn't work. Or we've all done these kinds of things that haven't worked and have figured out ways to mitigate the problems.It also does give you a little bit of COVID if the evaluations are not so great. I would also encourage faculty to be a little bit prudent about what they share with students.So you and I are in fairly privileged positions where we've been around a while.We're male, I'm certainly older, but I've heard from a number of faculty who are in other circumstances that if they admit that they don't know something or they're going to try something that might fail, it absolutely comes back to bite them where you or I can get away with and say, hey, look, you try stuff, sometimes it works, sometimes it doesn't. And they accept that from us. They may not accept that from others. It's not fair, but I think that's a reality of the world.Have you experienced anything similar or different?

No. I actually have seen a lot of students that like the transparency and to know that the faculty are figuring this out with them.So at some level, I think, being honest, that, yes, you're figuring this out. We're figuring this out too.It is the paradigm changing sort of thing where, no, we don't have all the answers, but we have seen new technologies come and go. Here's what we've learned from the past, here's how we've learned it.I think including students in those conversations is preparing them for exactly what they're going to do when they enter the workplace, which is figure out new technologies, whatever those might be.

Yeah, that's a good point. I think what you can do if you feel like maybe students are going to be less accepting of ambiguity from you is you can have bounded ambiguity.Here's what I know, here's what I don't know, and here's what we're gonna do, and I want your feedback on it. The other thing you can do that absolutely will help is make sure that if an experiment fails, it does not negatively affect student grades.Hopefully it won't affect learning. But let's be real. What a horrible saying.

Keep it real, Craig.

Keep it real. My favorite Chappelle sketches when keeping it real goes wrong. But the, the thing that students really care about are their grades.I've always had a policy that if I tried something new and it didn't work, I might give everybody all the points for that assignment. And students seem to be a lot more accepting if they know that they have that sort of COVID So that's something to think about.Also, I don't know that I would massively change how I'm doing my tests right now until I figure this out.I mean, I might change the modality doing face to face or something like that, but I don't know that I'd go so far to try to completely redo how I think about learning assessment for big stakes assignments. That's the kind of experiment that could go, if it goes wrong, could go horribly wrong. You're absolutely right.We've got to get started or we're going to be having these same conversations a year from now, two years from now, three years from now. Now's the time to learn. But I do think you can be prudent about how you do these experiments. Feel free to email us.It's Rob Crossleri goes to college.com or craigiigostocollege.com and we're happy to bounce ideas back and forth with you. To quote Animal House, that don't cost nothing.

And what I'll add to what you just said, Craig, I think that's all great is, I think this is why tenure is important. And I know every institution is a little bit different with tenure, and there's some changes happening at different states at that level.But if you have tenure, this is your chance to lead. And why you're given the protections of tenure is to be the thought leader and to have that academic freedom to do that.If you're an assistant professor at risk of your tenure and those sorts of things, I think the conversation might be a little bit different because those Risks are very real. But I honestly believe why tenure exists is so we can take some risks and lead through changes in the world and in society.

Absolutely. 100% spot on. This is a time for senior faculty to step up.Because the reality of it is if we get bad evaluations, as long as it's not a consistent thing, and it's not so bad that it triggers post tenure review, it's okay if you're a third or fourth year assistant professor who's getting ready to do the final push to tenure. You don't want to have a semester or two with awful evaluations. And so you got to be smart. Got to be smart 100%.So I'm a little worried I've gone politically non correct by citing Chappelle and Animal House. But, you know, I don't know.

That's what people listen for. Craig.

Well, I've got tenure, so get started. But we really are genuine in our offer to help you figure some of these things out. We enjoy having these conversations. We have them all the time.But feel free to email us.

Yeah.So, Craig, I found it interesting what you said about not punishing students with grades, based on the newsletter that went out about a week ago about the great economy. And it seems like that suggestion pushes into the great economy. But yet you called for an end to the great economy.Can you tell us what you mean by the great economy and how all of this plays together?

I can in fact tell you what I mean by that. So an economy I'm going to way oversimplify. So those of you who are economists, please don't be too hard on me.But one of the things that occurs in an economy is transactions. You have something I want, I've got something you want. We exchange that often. It's money for goods. We do that every day.We do that constantly, exchange one thing for another thing. And we've created an environment where students trade effort for grades. So effort is the currency. The grade is the good is the product.They feel like I put in this much effort, I should get this kind of a good for it. I should get this kind of a grade.And if I operate under that sort of a logic, then the smart thing for me to do is to minimize the amount of currency I give for the product. I just bought my wife, Tracy, a new vehicle. It's really nice, by the way, but the local dealer here had a really high price.So I looked around and I ended up driving almost three hours to buy it for 10 or 12% less. Why would I do that because I'm trading currency for a product.And so if we look at what AI can do from that perspective, then it makes sense for students to use AI in ways that minimize the amount of effort they put into getting a certain grade. And I know I've ranted about this before, but have not gone into it in much depth.So students are being perfectly rational under that sort of a logic. And we've created these conditions. We've done it. And not just us.I'm actually doing some research in the history of grades and it goes back into at least the 1800s. But we've made it worse during my time as an academic. And so students are going to minimize input to get the desired output.So what we have to do is we have to shift that thinking and we've got to do it in small ways individually. But it's going to take a systemic change in order to really solve this problem. We pretend that AI created cheating. AI did not create cheating.Cheating has been around a very, very, very, very long time. AI has just accelerated what was already happening. You're probably familiar with chegg.You know, any assignment out there, you can go pay a little bit of money, you can get it done through chegg, submit that and you're good to go.

Well, and what I would say about AI and I struggle with it being considered cheating is with chegg, it's finding the questions with the answers right there. You claim them as your own.With AI, it is, I've got these problems, I go interact with the AI and it helps me to come up with a written solution to those problems, which may be brand new problems that I'm not copy and pasting.And if I were to do that in the workplace, to help me come up with a solution, assuming I put the work into it to find good solutions, that's not cheating. That's called being efficient.And so I do think the nature of what is cheating is changing because so much of human interaction with these technologies is that. And I struggle with even calling that cheating.

I don't disagree with any of that.But I would also point out that a lot of what students are doing is pasting a question into ChatGPT and then copying and pasting the answer, which is cheating. But my point is we've created the conditions that have incentivized that and that's what we've got to change. We can chase the cheating all we want.We've never solved cheating 100%. We're never going to, especially not in online classes. And Less structured environments. And the problem is that this great economy is unnatural.When we're kids, we learn because we want to do something. When I was trying to play basketball, I worked really hard.I got most improved player one year because I was pretty clumsy and crappy when I started out. And I worked really hard on all the drills and ended up being a lot better player by the end of the season. Not because I was getting a grade.I didn't get a grade for that. It's because I wanted to be better. You've got a child who's a fabulous musician.Maybe they need prodding to get started, but once they get into it, they're internally motivated and they do it. They're not going to want to say, hey, AI, play this guitar for me. Oh, look, now I can play guitar. No.And so we've got to do something that feeds into that natural desire to learn.

And Craig, I think those students still exist. I think many students have a natural desire to learn.But what they see are those that find the shortcuts and see how their finding a lot more free time in their day or various different things because they press into faster, easier ways to do things.And even those who are motivated to want to learn eventually are drawn towards doing things in a way that maybe they wouldn't in a world where there wasn't those comparisons.

Well, especially if they don't see the point of what they're doing.I mean, I think that's a big part of it is if they don't, if it's just busy work, then why should I put a lot of effort into it, even if I am motivated to learn? I saw this in my doctoral students versus undergrads.So, and I'm sure I've talked about this before, but my doctoral seminars, I encourage them to use AI. I show them how to use AI. They use it a lot. They tell me how they're using it. But I also kind of don't give grades.I mean, I assign a grade at the end of the term, but when they hand in papers, I don't put a numeric grade on the paper. I don't do it. I said, look, here's where you were really strong, here's where you were weak. Here's how you can make these arguments stronger.I give them feedback and they're internally motivated because they want to learn how to be good scholars. The undergrad class with 80 or 90, 100 students, that kind of thing goes out the window a little bit.And so I'm a lot more Concerned about students using AI in ways that actually hurts their learning and prevents skill development. And so we've got some big heavy work to do here, we being higher ed in general. But I think to your earlier point, it's time to get started on it.

And so Craig, what can we do?I mean, because I think about at wsu it's expected that to earn an A in a class that you're spending two to three hours outside of class time per credit hour doing work. So a three credit hour class is almost this expectation that you're spending another six to nine hours outside of class working for that class.

You know that has never been true.

I know it's never been true, but it is what is defined. And there's actually calculators at various different institutions.We've looked at them at WSU to calculate is there enough outside work to justify the workload for this being a 3 credit hour class? AI definitely makes things more efficient depending on how you adopt it.What does this look like as we push towards not having the great economy when there are institutional rules around what is a three credit hour class mean for work outside of the classroom and contact hours and all those sorts of things.

So I want to separate work outside the classroom and contact hours because I think they're two different things. I'm pausing here purposely because I'm trying to gather myself and not give the really hot take.But this X amount of outside work for a three credit hour class is complete and utter hogwash because people are different. One person, it might take 15 hours. Another person might be really talented in this area and it takes two hours. So wait.The 15 hour person who doesn't know it as well should get more of a grade than the one hour person that knows it really well. Differences in talent exist. They do. I think we just need to. Again, this is a role of senior faculty.We need to be vocal and pushing back against that kind of administrative nonsense and pushing back against accreditors. I mean, accreditation has its place. I used to be an AACSB peer, whatever they call it, Peer review team reviewer. That seems redundant.So it has its place and it can help schools be better. But it goes too far a lot of times. And so I think we need to push back. I've always pushed back against targeted GPAs for a class.So wait, Rob's a great teacher, teach an Intro to Information Systems and I'm a crappy teacher.And Rob's such a good teacher that all of his students get A's and half of my students get Ds and Rob gets in trouble if somebody's giving away grades, solve that problem.

Right. So should we change the focus then to talking about in an individual class level, what do we hope the students know at the end?And how do we go from A where they started, which everybody's A is going to be different to B? What is it that we wanted them to know at the end?And if you could demonstrate B, then you've learned what it is you should have learned in this class.And then we stack that to be able to say this degree should be able to have this variety of things that students have learned and that demonstrates they have earned the degree associated with the major.

You've hit on what we ought to be working towards, and that's demonstrations of learning, demonstrations of capabilities of knowledge, whatever label we want to put on it. I don't even want to say learning, because learning is going from point A to point B.And what we really want to do is know how well they got to point B. And you can do that in a lot of ways. You know, portfolios, big exams.I mean, I'm not enough of an education expert to know what the right thing is, but I think the first thing we need to do is have some very frank discussions about why we give grades. Grades can have legitimate purposes. They can credential learning. When the grades are done well, they can provide feedback.They can create some accountability. They can motivate students, although the great economy has set up some perverse incentives around that.The problem is that we've kind of done this Taylorism of education, going back to measuring everything so we can improve it and breaking things down into tiny components. That's production line thinking. And it works well on a production line, doesn't work well in education and never has, really. Never has.And so we get points for participation, attendance, for doing the reading. So learning becomes secondary to the accumulation of points. And that's what we've got to change. I think, about how music works.You learn, you perform a piece. Somebody judges how well you perform that piece. If you're really good, you get the job. If you're not, you don't.I'm oversimplifying, of course, but when I was playing basketball, I got to be a starter because I was a better player than the guy that was a starter before me. That's what we need to be working towards. And get rid of this production line mindset.

Yeah, we've got this mindset that that which gets measured is what gets done. And so if I want you to show up, I'm going to measure showing up. If I want you to participate, I'm going to measure participation.If I want you to read the textbook before you come to class, I'm going to measure reading the textbook before you come to class. Whereas at the end of the day, what we care about is do you know the things that you should know from this class?And using your basketball example, you yourself said you were clumsy and not naturally talented at basketball and you worked really hard to become a starter.I'm willing to bet there was somebody on your team that they were just born very good at basketball and they didn't have to work nearly as hard as you did and they were a starter, right?

Yeah, absolutely.

And we're going to see that in our students as well.There will be some that just naturally get it, whatever that topic is and don't work as hard as the person who says, I really want to know how to work with databases. And so they throw themselves into that. And both got to the same level of demonstrated knowledge, but it took one longer to get there than the other.

Yeah, I mean, the world ain't fair, right? Talent is not evenly distributed, although pretty much everybody's got talent in some areas. But it just may be in a different area.We really need to think hard and have a lot of difficult conversations about what grades are for. One of the ones I'm going to push back on big time is grades for sorting. In the pre call, we were talking about doctoral student admissions.By the way, if anybody's interested in pursuing a PhD in Information Systems, both Washington State University and Louisiana Tech University have find programs. I would encourage you to get in touch. But we use grades to sort. Employers do that. You've only got a 3.48. We've got a 3.5 cutoff.What kind of nonsense is that? Okay, maybe if you've got a 2.2, maybe that says something.And I know as a pragmatic matter we have to have some way to sort, but I don't know, are we letting the measures drive what we're doing instead of trying to figure out how to measure what it is we ought to be doing? That's the fundamental question. Yeah.

And how do we make that?I want to use the word easy, but that may not be the right word because I know from an employer's perspective, when you have 1,000 applicants, how do you discern A from B? And some of these numbers are the easy ways for them to do it.And AI has made it so easy for people to apply for every job and then AI is used to sort for every job. It's creating a whole new set of problems.If you're a human resources student listening to this, I think there's real opportunity to look at how these processes can change and can be different to bring the human back into what we're doing.

Yeah, absolutely.

Because we have definitely made it into a numbers game where it probably shouldn't be.

And it amuses me greatly when employers complain about applicants using AI when they've been using AI for what, at least 20 years for applicant selection. So, yeah, I kind of don't want to hear it. Usually if I point out a problem, I try to have at least a start on a solution for it.I don't know that I have a good start here other than we have got to address this problem.Going back to the first segment, we've got to get started on it because until we solve this underlying problem, I think we're just going to be putting a band aid on a mortal wound. Did that work? I'm not sure that little analogy worked.

But what you're saying though, is we can't do minor tweaks to get to the right solution that we truly have to be thinking, I like to say blow it all up and put it back together again. But we have to be thinking about how is this done differently? And I will tell you, our institutions get in the way of this at some level.You've got the rules that a student starts on one catalog, that's the catalog they finish on.And if you're going to drastically do things different without special permissions, you've got four years of that person who entered by these current set of rules following them, even if what needs to happen needed to happen right now. So there are some real challenges that we need to be having these conversations.I'll put a call out to AIs, the association of Information Systems, and there's probably other associations that should be having these conversations. I would love guidance from our associations on what this looks like. And I don't see a lot of it.And I think that as we have a bunch of little pockets that of places that are trying to figure things out, we're going to end up with a whole bunch of different solutions, which is great natural experimentation. Some things are going to work better than others.But at the end of the day, the power of our associations to help guide institutions to getting this right will be a whole lot better than a bunch of individuals fighting against the status quo. Or the inertia of the way things have been.

Yeah, absolutely. We need collective action of some sort. AI is exposed. The fundamental flaws in the system. I mean, it really has. That's what my view is, at least.The system is not working now and it's going to take big change.I think this could ultimately be a very, very good thing for higher ed, where we end up being much better at what we should be doing than we have been for the last few decades. So. Okay, sorry, I've got myself worked up here.

Breathe, Craig, breathe.

In through the nose, out through the mouth. So, anything else on that? I think I've ranted enough.

No, I think we got a little excited today, Craig.

I did, I did. Got a little pumped up. But seriously, start talking about this. Start with your colleagues and don't just make it a whining session.Start to see the possibilities of what could be. All right, let's go from that big esoteric conversation to something eminently practical. Have you explored Chrome Split View?

Yes. I actually like it a lot.I do Split View on my Windows all the time, but this does it naturally inside of the browser, which makes things a lot more convenient.

So what is Split View, one might ask? Well, it's really cool. It does what it. So we've done this with Windows.This is one of my irritations about the Mac is it does not arrange Windows as well as Windows does. But you have one application open on the left, another application open on the right, or top and bottom, and you work across those applications.Well, Chrome and Edge both have that built in where you can select a tab and then right click on a second tab, go down to. I'm going to have to walk through it. So I get it right.New Split View, it'll bring up a context menu, new Split View with a current tab, and then those two tabs will be side by side.And it doesn't sound like it's that big a deal, but holy moly, if you're working with AI on a document, it is amazing how much friction that gets rid of to just have those two views side by side. So I do this all the time.I'll have Chrome open, I'll have Claude, and I use Lex Page, but you could use Google Docs or whatever you use to work on a document. It's just. It sounds simple. It is amazing how well it works. Like I said, Edge.I can't walk you through how Edge does it, but Edge has the same sort of thing. Yeah.

What I think is really cool about that, Craig is as a Copilot user.Microsoft Word does that for me with Copilot baked into where I have Copilot on the right and my Word document on the left and it's just part of Word. But what this does is it gives you, the user, more granular level control over which AI you want to be using within which text editor.I mean, you could even open up your office 365 Word document with Claude on the right doing the same sorts of things. So you're no longer stuck in a if I want to be Microsoft, I have to be Microsoft.You truly can pick the tool that you want to use, whether it is that text editor or the development tool, if you're writing code or whatnot and put those things side by side, assuming that they can run in Chrome.

Yeah, absolutely. I have a question for you. This is off script. I do the same thing with Gemini and Google Docs that you were talking about with Copilot and Word.It doesn't seem like Gemini is as good in that context. And I don't know whether it's just me or if it's really not a full model that's running inside of the application. Have you noticed any differences?

I find it works really well to use Copilot right in there and then it will rewrite paragraphs for me or it'll take what it's done and paste it into my document and then I can edit on it. It creates some nice iterative processes of being able to write.

All right, good deal.But if you haven't tried this Chrome Split Viewer, Edge, Split Screen or what I think it's called Split Screen and Edge, I really would encourage you to do it. It's a small thing, but it's really helped my workflow. All right, Rob, anything else?

No, Craig, I think we've given a lot of information and hopefully people take action with what we've talked about and Craig mentioned if you want help to go ahead and reach out. The other thing we would love to hear is if you've tried something or planning on trying something, let us know.It would be great to hear how our listeners are taking what we're doing and putting it into action as together we can change the world of higher education one decision at a time.

Right. And you can do that by going to aigostocollege.com and using the contact form that you'll see down.I think it's the bottom right hand corner somewhere on the webpage. Or you can email rob.crossleri goes to college.com or craigi goes to college.com. all right. So nothing else?

I'm good.

Okay. We'll see you next time on AI Goes to College. Thank you.