Why I use Multiple Grading Schemes

It's about equity

I have been using Alternative Grading approaches in my community college classroom for over 5 years. I started altering my grading approach after reading the book Grading for Equity by Joe Feldman. In response to some of the ideas in his book, I started using Multiple Grading Schemes. In this post, I explain what Multiple Grading Schemes is and why I think it’s more equitable than traditional grading.

Which is more equitable: Including homework in the final course grade, or not?

In Grading for Equity, Feldman makes the argument that a student’s final course grade should only represent a student’s demonstrated learning. The final course grade should not, he argues, include grades from activities that promote learning, like homework, practice quizzes, or draft submissions. (For the remainder of this post I’ll use the term “homework” to mean any type of graded activity whose primary purpose is to help students practice new skills or reinforce concepts, but is not intended to determine students’ mastery of the content.) Since learning is a mistake-making process, students should be able to make mistakes on homework without being penalized for it in their final course grade. His reasoning is this: Imagine a student, Charlie, who makes many mistakes on his homework but another student, Brandon, makes almost no mistakes at all. However, Charlie learns from his mistakes. On the exam, both Charlie and Brandon perform well. Using traditional grading, Charlie’s lower homework score will be averaged with his exam score, resulting in a lower final course grade than Brandon, even though both Brandon and Charlie exhibited the same level of content mastery on the exam. This is not equitable, because the same demonstration of learning earns different grades. Thus, Feldman argues, an equitable grading system is one where the final course grade includes only “summative” assessment scores like those from tests, exams, and other final projects.

The problem with Feldman’s argument is that if you remove grades for homework, many students won’t do that work. That hurts those students, because they don’t benefit from doing the learning-promoting activities that the homework is designed to guide students through. For this reason, in their book Inclusive Teaching: Strategies for Promoting Equity in the College Classroom, Kelly Hogan and Viji Sathy argue that homework should be included in the final course grade. They reference research (Freeman et al. 2011, Eddy & Hogan 2014) showing that high-structure courses, which implement weekly graded homework assignments and regular low-stakes testing, increase scores for all students, but especially benefit traditionally-disadvantaged groups like racial/ethnic minorities and first-generation college students. Thus, in the name of equity, Hogan & Sathy make the opposite argument from Feldman: they say that homework should be included in the final course grade.

As instructors, what should we do? Include homework in the final course grade, or not?

I split the difference and do both. Here’s how.

What is Multiple Grading Schemes?

The premise of Multiple Grading Schemes is simple: rather than having one grading scheme that weights different components of the course differently, use multiple grading schemes with different weights for the different components of the course. Here are a couple examples.

The math class is taught by someone else at my institution. The biology class is mine.

In the Math class shown above, scheme 1 is “test heavy,” with 70% of the final grade based on tests and the final exam, whereas scheme 3 is “test light” with only 35% of the final grade based on tests and the final exam.

In my Biology class, scheme 1 includes all the components of the course, including homework and online quizzes, whereas schemes 2 and 3 don’t include those components at all. Additionally, schemes 2 and 3 flip the relative weight of the two major types of assessments: the final exam and essay exams, which probe students’ mastery of the material in very different ways. Students may excel at showing mastery using one format more than the other, so I offer two grading schemes that heavily weight one or the other type of assessment.

At the end of the semester each student’s grade is calculated using each grading scheme, and they get the best grade. This is important: students do not choose which grading scheme applies to them, because that could unfairly privilege students who are better at predicting how their semester will go if they choose at the start of the semester, or are more math-savvy if they choose at the end.

By using Multiple Grading Schemes, I split the difference between including homework and not including it because one grading scheme includes homework and two don’t. In the math class, while homework is included in every scheme, homework (and labs, which are more like homework than assessments) are weighted more heavily in some schemes and less heavily in others. This way, students can still be rewarded for doing the homework, so the motivation to do it is still there, but students will not be penalized in the grade if they struggle to consistently complete the homework.

(Aside: I buy Feldman’s argument that since learning is a mistake-making process, students should be able to make mistakes on homework without being penalized for it in their final course grade. For this reason, I only grade homework on completion [I made a video about it!]. In the math classes at my institution, the homework is graded on correctness, but students have multiple retry opportunities for each assignment.)

Thus, Multiple Grading Schemes is a way of providing multiple paths to success for students. They can be rewarded for consistent, diligent performance on homework and other activities that promote learning, but they don’t have to be penalized for poor performance on homework because other grading schemes weight assessments more heavily. Multiple paths to success is, theoretically, more equitable because it means students with different strengths can have their strengths recognized using different grading schemes.

Is Multiple Grading Schemes more equitable?

Let’s start with what it means to be “more equitable.”

In a 2024 paper, researchers analyzed a large dataset of student grades in university-level Science, Technology, Engineering & Math (STEM) classes. They found that students with more Systemic Advantages had higher grades, on average, than their peers with fewer Systemic Advantages. They define Systemic Advantages as: being white, male, continuing generation and from a higher-income background.

Data from “Systemic advantage has a meaningful relationship with grade outcomes in students’ early STEM courses at six research universities” by Castle et al. (2024)

These data tell us that STEM education at the University level is not equitable. An equitable system would not have these grade disparities based on identity or background.

While the data above are from 6 large research-intensive universities, we see the same trends based on race at my institution (a community college in NC). Here are average course grades from all math courses at my community college, broken down by students’ self-reported race. I’ve highlighting white, Hispanic, and Black demographics because together they represent 85% of our math students¹.

Average course grade for math students at Durham Technical Community College from Fall 2021 through Spring 2024. Data are from NC Community College Data Dashboard.

Based on these data, an equitable policy would be one that helps to level the playing field. It isn’t that Blackness makes students bad at math; rather, there is something about the system that makes it harder for Black students to earn high grades in math classes. An equitable policy is one that changes the system to reduce, or eliminate, that barrier.

Back to Multiple Grading Schemes: is it more equitable than a single grading scheme?

Luckily, there’s data to help answer that question. In the Fall of 2021 all of the math and physics classes at my community college started using Multiple Grading Schemes. As a result, after 3 years of implementation we now have final course grades from over 2500 students whose grade was determined using Multiple Grading Schemes. My colleagues and I are analyzing those data and writing it up for publication, so I won’t go into many details here. But here’s one major punchline:

For all the math and physics classes, grading scheme 2 is the “original” grading scheme. Grading schemes 1 (“test heavy”) and 3 (“test light”) were added as part of implementing Multiple Grading Schemes. When we look to see which students got a higher letter grade in either scheme 1 or 3 (in other words, they would have gotten a lower grade if we had only used the original grading scheme), we see that more Black and Hispanic students get a higher grade because we implemented Multiple Grading Schemes. Importantly, no students saw their grade go down because of Multiple Grading Schemes. Also, since grades are not curved in any of the math and physics classes, the fact that some students’ grades were bumped up had no cost to other students’ grades.

Note that Multiple Grading Schemes is applied equally to all students - there is no differential treatment here based on race - yet the grading system especially benefits students who are systemically disadvantaged in STEM classes. This is exactly what we would expect an equitable policy to look like in our current system.

If you’re curious why we see this result, we are too! Stayed tuned for our paper where we dive into the data to find out what’s going on.²

One last thing: I made a video about Multiple Grading Schemes, which includes some tips for implementation that I don’t mention here. Check it out:

Not everything you read on the internet was written by a human. For full transparency, here is how I used AI to help me write this post:

I drafted the entire post, then I asked Claude to give me feedback on the draft. Specifically, I asked for feedback on ways to reduce redundancy and to identify areas that need more clarity. I made changes based on some of its feedback. All of the text in this post was written entirely by me.

1. In Math classes, 38% are white, 24% are Hispanic, and 23% are Black.

2. When our paper publishes, I’ll post about it here on Substack. If you’re subscribed, you’ll be the first to find out.