With all the questions about students using AI and how we can verify that it's their work, what does it mean for something to be truly their own? Students can use AI, and professors can use methods to detect AI. It's a complex arms race that kind of defeats the whole purpose of education. This got me thinking back to my favorite exam ever.
This was the first time I was teaching. My theory is that exams are not about grading; they are about motivating. I decided to forget about the accuracy of grading. My goal was to get the students to really study well. Here's what I came up with.
In the second class of the semester, I told my 30-some students, Here are 50 essay questions. These are the questions that will be on the exam. If you know the answers to these questions, you know the exam. The questions covered all the material, so if they mastered these, they were set.
But there was a twist. On the day of the exam, I would pick one question at random and read it out loud. The exam would be in person, so everyone would be there. Then, I'd read one student's name, and they would have five to ten minutes to answer that question. Afterward, I'd ask if anyone had anything to add.
This method ensured that students had an incentive to study all the questions thoroughly. They couldn't just study 90% because they might get the one question they didn't prepare for and get zero.
I added another component: subjective evaluation. After a student answered, I'd write down what I thought their grade should be from zero to 100. Then, each student would say out loud what grade they thought they deserved. If their grade was up to nine points above mine, I'd give them their grade. If it was more than ten points above mine, they'd get my grade.
This way, students had an incentive to slightly exaggerate but not too much. It also reduced arguments later because if a student gave themselves 87 when I gave them 80, they couldn't later claim they deserved 100.
We did it this way, and it was amazing. I can't tell you how many times I saw students in groups studying in the library throughout the semester. They didn't wait until the end; they had the material early on and reviewed it weekly.
When exam day came, it was delightful. Students answered accurately and honestly evaluated their grades. It was a stressful semester for them because they studied hard, but it was a good approach.
It was a tremendous experience for both me and the students. Many years later, a student even wrote to me about it. This method could help against any kind of AI by getting students to truly engage with their studies. Maybe we need to go back to this or something like it.
