Reflections on Grading Bias and Alternative Grading Practices

On January 11, 2024, I attended the Major Lecture of the Association québécoise de pédagogie collégiale (AQPC) on ungrading. I found the presentations by Jesse Stommel, Caroline Cormier and Bruno Voisard very interesting. I already had a keen interest in alternative grading practices, but this conference further fuelled my thinking and prompted me to read more on the subject.

Grading is biased, even when it’s based on self-assessment

During the conference, Jesse Stommel mentioned a meta-analysis that found grading is biased when graders hold “irrelevant” information about students (ethnic background, education-related deficiencies, physical unattractiveness, quality of prior performance, gender, etc.). Bias can occur when the grader looks at a student’s written production alone as well as in a collaborative grading process.

Just a few days before the AQPC conference, I had read the testimony of Jayme Dyer, a biology teacher at an American college, explaining why she abandoned collaborative grading. In “Ungrading Has an Equity-Related Achilles Heel. Implicit bias is unavoidable in ungraded courses,” which I read through the Grading for Growth newsletter, Jayme Dyer uses examples from her own practice to demonstrate that students themselves have biases about their own worth (or the worth of their work). It is those with the most self-confidence (and therefore possibly those who have been most advantaged by the system up to this point in their course) who are most likely to argue with their teacher to get a good grade.    

In her text, Jayme Dyer presents her solution for assigning final course grades to her students while resisting traditional grading methods.

I find this multiple grading schemes approach very interesting. I’m curious to know if some people use it in the college network. (Share your experiences in the comments area!) Personally, I don’t think I’ll try it, as I’m looking for a simpler system, but the approach certainly has merit.

 Of course, this approach does not eliminate the problem of bias. To paraphrase Jayme Dyer, the impact of bias is less important when the grade awarded is only part of the final grade, rather than when the final grade is awarded all at once, as is often the case with ungrading.

Is it ungrading if there’s a grade at the end?

As Jayme Dyer writes, with ungrading, there is an important difference between a course for which the teacher is forced to assign a grade to their students and a course for which this is not required. In CEGEPs, teachers are generally required to submit grades to their institution. And that’s the heart of the problem…    

I myself tried ungrading in one of my courses, relying on collaborative grading to assign marks. I really enjoyed that session, because I liked having plenty of opportunities to talk to my students individually (and my students liked it too!). Unfortunately, it didn’t allow me to fully achieve my initial goal of focusing on learning rather than grades… as grades remained a central concept in my discussions with students.

2 versions of ungrading

Yet Jesse Stommel seems quite satisfied with ungrading, even if the school where he works requires grades to be assigned to students. After the conference, I read his book Undoing the Grade: Why We Grade, and How to Stop and realized that his approach had evolved (or varied from course to course, no doubt) since his testimonial that I’d read before my own ungrading experiment.

Focus on self-assessment

In the 1^st testimony I read, Jesse Stommel asked his students to grade themselves at the end of the session. If he felt that someone was undergrading themselves out of modesty, he would raise the grade. But, on the whole, he did very little to moderate students’ grades.

At the end of the term, every institution where I’ve worked has required me to issue a final grade for students. So, I ask the students to grade themselves. I wish I didn’t have to do this. I wish the conversation I had with students could focus purely on authentic assessment, process, and formative feedback.  […]

I’m frequently asked what I do when I disagree with the grade a student gives themselves. I don’t intend my answer to be flip, but I say some version of, “It isn’t really my problem.” If I’m going to give the responsibility of grading over to students, I have to let go of my attachment to the accuracy of that process. Instead, I give feedback, and the need for objectivity or accuracy gives way to a dialogue—one that is necessarily emergent and subjective. I do make clear on the syllabus (and in class) that “I reserve the right to change grades as appropriate.” But I do this only very rarely, and I usually have to raise grades. The most common change I make is from an A- to an A for students who offer no good reason other than modesty for giving themselves the A- grade. (I have observed a distinct gender imbalance in this, with women students much more likely to give themselves an A-.)

– Jesse Stommel, How to Ungrade (2018).  

There’s no doubt that this approach is highly prone to student bias (though Jesse Stommel undoubtedly often manages to correct the situation). I hadn’t implemented it as such in my own practice because the reality is that grades (or the difference between passing and failing a course) have an impact on students’ progress (or even their lives!). Yes, I wanted to avoid disadvantaging the humble or the insecure. But conversely, I also wanted to prevent the more confident (or bolder, less modest) students from being unfairly advantaged.

An A for everyone

In the book Jesse Stommel published online in 2023, this 1^st testimony reappears, but another version is added.

In chapter 9 of his book, Jesse Stommel describes a method as simple as it is bold: “Everyone who participates in our course community and completes their self-reflections will get an ‘A.’“

This approach certainly has the merit of being simple. And it’s significantly more faithful to the essence of ungrading than my own approach.

Trusting students, but also evaluating competency achievement

I admit that giving everyone “an A” is too daring for me (at least, in my context).

Jesse Stommel, both during the conference and in his book, puts a lot of emphasis on the notion of trust. Maybe I don’t trust my students enough. But I do know that, unfortunately, effort does not equal competency. It often happens in my courses that students do all the assignments (formative and summative, even the optional ones) and still fail.

Without wanting to agonize (to borrow Jesse Stommel’s idea) over the difference between a 95% and a 96%, I feel it’s my responsibility to make a critical assessment of each student’s attainment of the passing grade. I’m not ready to take the “everyone gets 92%!” step. But I do recognize the interest and appeal of that approach. It’s undoubtedly the key (or, at least, the simplest idea) to helping students truly focus on learning (the process, etc.) rather than performance (the “mechanical” response to expectations).

Other alternative grading practices

Christian Mercier, a teacher at ITAQ, shared his experience of specifications grading in a very nuanced way. I think there’s a lot of potential here, but also some pitfalls, which Christian describes very well.

After the AQPC conference, I contacted Caroline Cormier about joining the community of practice on alternative grading that she and Bruno Voisard have started. I can’t wait to find out more!

What’s your version of alternative grading? What works for you, in your context?