The case for changing how we approach grading
I listened to a podcast from The Daily last week (June 28, 2023) about how students were actually using AI and talking to one professor about his response. And I was again struck by how heavily we focus on grades and GPA, as if those were holy predictors of learning and success. Then I read Ethan Mollick’s piece (2023) on what the fall might look like and changes that instructors should make, and my concerns about grades and GPA were reinforced.
For the past decade or so, I have tended to focus as much on process as product. I have flipped my classes where it made sense to do so, moved to more group projects, had students repeat certain assignments until they grasped the concept, and asked students to demonstrate knowledge through individual, pair, square, share discussions. And through all these changes, I have let students know that they have to put in a moderate amount of effort to fail my courses. They won’t get an A without work, but they won’t fail without work either. In other words, I have tried to move away from the grade being the end-all thing that shows knowledge gained.
I haven’t gotten onto the ungrading train. That’s not to say that I haven’t tried, but I haven’t succeeded yet. I had a brief flirtation with considering contract grading, but it also eluded me. That’s not to say that these aren’t great for other classes and instructors, I just haven’t found comfortable ways to make them work for me.
In the US (where I am located and where I can speak to), I have watched high school students stress over “only” having a 4.5 weighted GPA, “only” having so many AP classes, “only” being in the top 1/10th of the graduating class. And for what? I mean that both kindly and realistically. What does all that stress amount to for these students in the long run?
Brief personal story: I wasn’t a great college student. I never had to study in high school and then college hit me like a brick wall. Studying techniques that seemed to work so well for folks around me didn’t help me. I had one professor who looked at the class on Day 1 and said “I’m not here to help you. If you can’t figure this out on your own, you’re going to fail.” And guess what? I didn’t do well in that course. That was a weed-out course in a hard program. So college for me was one partial failure after another after another. Guess who felt dumb and worthless? Sad thing was, I was really interested in my science courses, but the tests were just overwhelming to me. If you judge me only by my GPA, I look pretty miserable on paper as an undergrad.
So coming back to GPA and how that impacts the likelihood of cheating with AI. (You knew AI would show up somewhere, didn’t you?) Think about it this way. If a student feels pressure to get all As, then AI becomes more of a tool for their version of success. Regardless of what guidelines the instructor has set, a student who sees an A as the only acceptable grade will do whatever needs to be done to achieve that. And AI just made “whatever it takes” a lot easier to access. Now, envision a student who is aware that failure can be part of the learning process and who understands that repetition and iteration may be required before knowledge is gained and learning is achieved. This student may be less likely to use AI to cheat because they understand that the struggle is a necessary part of the process; that for knowledge to stay with us past the test, we have to take time with a concept.
As we move into the fall and beyond, we need to think about our courses. Which students do we want to have? And how can we convince those students who have been trained that grades are all that matter to take a learning journey with us that might involve some stumbles?
- Low stakes assignments along the way. Consider small problem sets or writing assignments to help students see where they are in the learning process.
- More small quizzes and fewer large tests. This allows students to again get a feel for what they know and where they still need help. Where possible, consider allowing students to redo their work for partial credit, or write a response to show how they got to the wrong answer and how that compares to the right answer. Have them work through their thought process to see where they stepped off the path.
- Break up writing projects into small pieces. For example, I have students submit a topic so that they are on the right track before they even start. In upper division courses with longer papers, they then submit a scaffold draft that is like an outline but less formal. Students come up with headings and quotations that might support those headings. Then they discuss how those quotations might fit into their larger ideas. This forces students to start researching early and gives time for the ideas to roll around in their heads for a bit. Next is the rough draft with more detail and complete sentences. But I still allow for questions, comments, and “I might talk about X here” statements. This draft can be a conversation between student and instructor. If you choose, peer review can happen at this stage as well. Finally comes the final draft and the annotated PDF copies of their research documents. Breaking up the process helps keep students from putting off the entire project until the last minute, and asking them to turn in the PDF copies of all their sources forces students to get back to some basic research skills. I also have clear AI usage guidelines for each step in the process, to help students see where the AI might be a useful tool versus where it might hinder their work or understanding of a concept.
- Small and large group discussions. For courses where the layout and attendance allow for this strategy, having students write about ideas before they come to class, then share those ideas in small groups is a good way to deepen understanding. After small group time, the class can come back together to discuss what the small groups talked about. In this way, even quiet students can benefit from hearing the main ideas from groups and from having spoken to a few peers. There is much to be said about learning through listening, which can also be helpful for neurodiverse students who struggle in large groups.
While the changes may not be easy and we may stumble through assignments that really didn’t go the way we hoped they would, putting the focus back on the process of gaining knowledge and moving away from the “3-test or paper and that’s it” model can help students focus more deeply on the course content. It can decrease student stress and can increase their willingness to take risks because they know that they won’t fail the course if something doesn’t go to plan.
I’m never quite sure what lessons my students will take from my writing courses. Obviously, I want them to get more comfortable with their writing and thinking skills, but making these sorts of changes to my courses has led to students sharing all sorts of extra skills and knowledge that they gleaned from my courses. And many students have shared that they feel less anxious about writing and class discussion after a semester with me. If our goal is to teach our subjects and help our students progress through their education journeys with a backpack full of knowledge and skills, perhaps AI and its challenges are just the kick we needed to figure out what tools really belong in the backpack and to set all the other ones down because they no longer serve us.
References
Mollick, E. (2023, July 1). The homework apocalypse. Substack. https://www.oneusefulthing.org/p/the-homework-apocalypse?utm_source=post-email-title&publication_id=1180644&post_id=132054399&isFreemail=true&utm_medium=email
Tan, S. (Audio Producer). (2023, June 28). Suspicion, cheating, and bans: A.I. hits America’s schools [Audio podcast episode]. In The Daily. New York Times. https://www.nytimes.com/2023/06/28/podcasts/the-daily/ai-chat-gpt-schools.html
