I’ve always been uncomfortable with the grading part of being a professor. As a student, I loved learning, particularly writing, but dreaded the judgment of an instructor and the harsh disapproval of their red pen. When I started teaching, grading was my least favorite part of the job. I’d obsess over minor discrepancies and second-guess myself constantly. Even when I felt I was completely fair in handing out a less-than-stellar grade, I worried about whether the student truly deserved a permanent mark on their record for something minor, such as forgetting to include a cover page. But I plugged along as a reluctant grader anyhow.
Then came the age of AI.
I am constantly seeing AI-generated papers, AI-generated emails, and even other AI forms of written communication from students. This has made grading, particularly grading written work, seem more unfair.
Article continues below
In other words, generative AI has poured gasoline and lit a match on top of an already flawed grading system.
While I understand that grades are still necessary and grade my classes in accordance with the institutional policies of the colleges I teach at, increasingly, I’m convinced that the widespread nature of AI provides more of an argument against grades in favor of some other model, such as standards-based grading or mastery learning. Here’s why.
1. AI Adds To the Randomness of Paper Grading
Most educators, myself included, believe we grade papers fairly. Unfortunately, data does not really back this. In fact, much research that goes back to the early 1900s has found that when educators grade the same paper, there is huge variation in the scores given; for instance, one 2011 study asked 90 different high school teachers in the same school district to grade the same paper according to the same rubric. Scores this identical paper received ranged between 50 and 96.
I’ve noticed similar variations myself when I’ve observed the teaching habits of other educators I’ve supervised. What earns a B in my class might earn an A or a C in a colleague’s, or worse, a D.
AI further unbalances this already unfair system. It’s clear that when we assign take-home papers these days, we’re creating a moral conundrum. Some students will use AI and receive good grades, while others won’t and might receive a poor grade as a result. This is not what any educator I know wants, but it is the default system that has emerged for handling AI.
2. AI Undercuts Motivation Argument for Grades
One of the main arguments for grades is that without any, students won’t do any work. Again, studies don’t really confirm this. According to one review comparing grading to feedback without grades, students who received feedback without grades achieved more and were more motivated to learn.
That said, I’m willing to grant that fear of getting a bad grade does motivate some students. Even I admit that I would worry about getting quality work from my students without some aspect of the fear-based motivation that grades provide.
The problem is that AI has already undercut this argument. In many instances, students no longer fear grades. Instead of grades motivating students to write, grades are motivating students to use AI to do their work for them.
3. AI Makes Grading Useless As A Ranking System
Another rationale for grading is that it helps sort students into various abilities, both for employment and for the next level of education. I never loved this as motivation for grades. Helping companies decide who to hire may have its uses, but it’s not likely to show up as an answer if you ask educators why they got into teaching.
Regardless, given how widespread AI cheating is these days, I wouldn’t place as much weight on grades as I once did. What I’m seeing in the world of writing education is that more and more, grades are merely a measure of how effectively a student is using AI without getting caught.
4. AI Makes It Clear Teaching To the Grade Isn’t Best
Almost all teachers I know rail against teaching to the test, but how different is that really from teaching to the grade? In fact, I believe that part of the reason students are turning to AI so frequently is that they’ve been taught implicitly that the grade, not their learning, is what matters. Therefore, when a tool comes along that can get that grade without work, there’s no reason not to use it. That is at least the way many of my undergraduate students seem to feel.
On the other hand, with my graduate students, AI use is nonexistent. That’s because they are in graduate school to learn to write, not just earn a grade. Of course, duplicating graduate student motivation for undergrads and K12 students isn’t easy.
I also know that eliminating grades overnight and switching to a mastery-learning method isn’t plausible. That doesn’t mean we shouldn’t strive toward creating a new motivation for students; we may not succeed in this process 100%, but we can do much more as a profession to move classroom motivation away from grade-based fear-mongering. If we don’t, I think our students will keep turning to AI, no matter how hard we try to catch it, and take-home writing assignments will become a sad remnant of the past.