News

Innovative Approaches to Revamping an Online Course: Personalization, Engagement, and Anti-Cheating Methods

Oct 26, 2023, 08:46 AM

Craig Lemoine

In the age of GenAI, how can we keep students engaged, challenged, and help them learn essential knowledge and skills with long-term benefits? How can we effectively counter AI-assisted cheating in online education? Professor Craig Lemoine gave considerable thought to these questions. In our recent interview with him, he shared innovative approaches for revamping his online personal finance course at the University of Illinois. His strategies include personalized assignments that encourage students to reflect on their own experiences, an interesting 'quiz ball' concept designed to deter cheating while making learning fun, and a collaborative grading approach to efficiently manage individual assignments in a large class.

Join us to explore Craig’s methods further in our recent interview with him, and discover how you can implement similar strategies to engage your students while discouraging AI-assisted cheating.

Q: How did you start your journey of revamping your course?

A: Sure. I would love to share a little bit about how we've adapted a very large class to sort of work around AI considerations. I teach an online personal finance class for our campus. It has all sorts of majors from all over the university. And while it's not a Gen Ed, it mirrors and mimics a lot as far as size goes. It's about money, and everyone wants to learn about managing their money. So it's a popular class.

The first thing about our class which is pretty unique is that we spent a whole bunch of time developing an eText to go with it. This eText has a lot of technical content about the class (faculty/staff can get free access to books). Topics like financial planning, how to make a budget, time value, taxes, economic concepts, banking, credit cards, risk management, unemployment, retirement, investing, building wealth, estate planning and then financial aid, and so on. It tends to cover the areas that students want to learn about. And so we've got this great content. This is very technical content. It all sort of has a uniform. So we use this eText tool from a very technical standpoint. There's really no assessment in the eText tool. It serves as a textbook. We keep everything edited. You can experience this on your phone just as much as on a tablet or on a computer itself. 

During COVID, this online class exploded. And so this class went from around 200 students a semester to 700, 800, 900 students. I think one semester, we had 1,100. It's a pretty good size. And I believe this is a great way to deliver content to a student. We kind of have this system in place where we can deliver education at scale. I think we've done a pretty good job segmenting it and chunking it, breaking it into pieces. And then you've got these little knowledge checks at the end in the eText. They're not graded, but it kind of helps the student learn a little better about what the content covered. Okay, so eText is what we used to deliver learning content. The assessment portion is taking place over in Canvas and this is where things get kind of fun.

We have two issues here. First, how do we assess the content that someone's learning in the class? If we use traditional quizzes asking students questions about the technical content of the course, we have to either utilize some quiz proctor tool that students have to buy or locks down their computer. And there are some issues with that. I think that we've run into with kind of false positives. The other fundamental problem is that a student is going to use Quizlet or Course Hero to copy and paste to answer the multiple choice question, especially in an online class. So we have to think outside the box a little bit. 

Q. How did you revise your course to avoid generative AI?

A: We made two major revisions in this course to avoid GenAI. First, we make the major assignments personal. We make the individual projects more personalized by asking students to reflect on their own life experiences with the course topic. The revised assignments become more personal and meaningful. For example, there's a goal-setting project called 10% happier precrastination or procrastination. It is based on the goal-setting content. They learn the content in class plus listen to a podcast. Then they need to talk about it. Do they tend towards precrastination or procrastination? What strategies discussed in class can help you improve your academic success? 

Another example assignment is about their relationship with credit and money. We ask students to pull their credit report and analyze it. Did you have difficulty accessing what credit Bureau provided it? Was there anything on it? But I ask them not to actually attach it. I don't want all their information, but I just want them to tell me about them themselves. An alternative is that international students may not have a credit report. And so just tell me about your money. What is your first memory of money? Okay. My first memory of money is being at a pool in Austin, Texas, with a quarter buying a Dr. Pepper, and how cool it was. I had a quarter, and I got a Dr. Pepper from the machine, and it was my money, and it was my first or second grade and it’s my real memory of money. But this prompts some of the most incredible conversations with students when they talk about their memory with money. Were you exposed to conversations about money as a child? Do you keep a budget? How do you spend your money? Okay, so either way, credit report or not. You can't use ChatGPT here. Because it can't pull your credit report. It can't learn about your journey with money. It can't tell you your first memory of money. So what we try to do is again make these projects about the student and not how much they memorize the technical content.

The other major revision is what I'm affectionately calling quiz ball. At the end of the student's learning journey, we have thousands of questions about financial planning and the student can opt in to quiz ball. It is intentionally an impossible task. Basically, I'm giving you literally 700 or 800 questions an hour that you cannot get through it. But if you're those that read and studied and did all the assignments in the course and in the eText you are going to be able to go faster than the students that didn't. It is all extra credit. You don't have to do it. You can get an A in the class just by the projects. But if you want to do quiz ball, you get 10 tries. We have such a big exam bank like 3000 in the bank. You're gonna see different questions almost every time. After each attempt, I don't let you know what was right or wrong. When you go back you don't get to see it either. It's quiz ball. 

So we kind of made the best of the resources we had created. We'd spent so much time on these quizzes, and then it just got to the point with using ChatGPT students can copy and paste any quiz question and get an answer. You know what? Forget about it. Let's teach you content. Let's assess your learning journey. And then all those thousands of questions? Let's make the most of them. Let's have some fun with it. We'll have students do that. And the more competitive students will try to win too.  So we incentivize students to go ahead and knock stuff out on their own pace as they go quicker. Yeah.

Q: How do you handle grading when this course focuses more on individual assignments?

A: With this class, I have a unique grading model. I have a faculty extender. That faculty extender is a graduate professional, a non-student, and can work up to 20 hours per week. They have some experience in financial planning. So they have financial planning as a part time job for them.

I also have up to 8 upper-level undergraduate graders. That works in the class as well. So between the 8 graders and the faculty extender and myself and the rubrics we've designed, we have capacity to crank through these fairly quickly. The last week of the semester is rough for me, but that is because we do have deadlines and timelines. It's not that bad. There are 10 of us at any time tackling a project. So 10 people, grade 10 people, great, so forth and so on. So you know, that keeps things moving pretty well. There's another one or two people I can ask to if I really need it, but with 10 people it goes pretty fast.

We also incentivize students to go ahead and knock stuff out on their own pace as they go quicker. They can always start early, too, which also helps.

Q: What are the students’ responses?

A: This is the first semester we tried this approach. I do wonder what happens with quiz ball at the end, but we'll find out. I think that's a neat way to do it. I think it's a better class and it gets around a lot of ChatGPT, Course Hero, and Quizlet issues because you sort of eliminate that as a crutch. 

We do have different ways to collect students’ feedback. The final assignment has built-in feedback like questions about what you enjoy, what would you do differently, and what you would recommend. I am also bold enough to post it on Reddit, which people tend to be more honest. Last semester, students thought there was too much work. There was a post about the class on Reddit. And I jumped in. It was a really productive conversation. And also, of course, we look at ICES feedback.

Q: What are some benefits that you can see?

A: I am very happy about the personalized projects. Through this project, students start on their personal investment journey. For example, we have this long-term goals project where they have to talk about their long-term goals in terms of building wealth and investing. So I ask them to read the content on investing and wealth, stocks and bonds, and go on to explore Robo-Advisors. A Robo-Advisor is a kind of younger-person investing tool where you interface with an AI platform. The AI platform makes stock picks and you invest in that. What it lets you do is invest small amounts of money with very low costs into a platform. So you could put in $5 a week. There's no person, no investment advisor who's gonna invest your $5 a week. But any of these Robos would happily take your $5 a week and then diversify that $5 across dozens of investments. It's really cool. It's meant for people who are starting out. I love them. And it leverages technology today. Okay, it's all done on the phone. It's fantastic. So I make them look at two and talk to me about strengths and weaknesses of the platforms and pick a winning platform. In the end, a lot of students do start investing like $5 or $10 a month or $20 a month. And all of a sudden, now, the students are building wealth on top of everything else. So that's like my happy place. If I can get a student to invest even a tiny bit, my heart is happy because they're starting a journey. If that same student sets their goals, checks their credit report, knows how their insurance works, and starts saving money for themselves, tackle their goal setting. All of a sudden, students are building financial planning and management skills that are much needed for their future. All of this cannot be achieved through traditional quizzes or exams. 

Another benefit is that we discourage cheating in this class and make learning more fun through the quiz ball. We have created thousands of exam questions that are technical. They are multiple-choice questions. It is not fair to penalize a kid that doesn’t cheat. A student who doesn't use Quizlet or Course Hero or ChatGPT tries on their own on a quiz and gets a 70. Whereas another student cheats, takes the quiz, and gets 100. That’s not fair. That’s like fundamentally unfair and it bothers me to my core. That’s why we created the quiz ball from thousands of exam questions. Students can opt in to take this impossible challenge to earn some extra credit. So we make this assessment fun. Those who studied and are familiar with the content will be able to answer the questions faster. Of course, students can still use GenAI tools but they cannot be fast enough to answer all the questions. So we avoid potential cheating through this extra credit project and make it fun and less stressful.