Grading as an issue of justice in this time of transition

by Brian Katz and Kate Owens

BK: There are multiple, foundational justice and ethics questions involved in HigherEd responses to COVID-19. Personally, I’m pleased how central these issues have been in the discussions I’ve witnessed recently, including questions at faculty fora and conversations in the Twitterverse. For example, I see faculty acknowledging their institutional power and stability and hence that they have responsibility to advocate for staff members, for whom institutional changes represent much larger dangers or who might be required to put their health at risk by being present on campus.

One theme looms particularly large in these discussions: grading. In part to advocate and in part to share ideas as many faculty are in the throes of designing new remote-learning courses that without the time or training to do that well, I wanted to try to summarize the themes of these discussions for a wide audience.

First, some information about my perspective as it filters this summary. (i) Grades are a fairly recent invention, and they have never been objective or precise measures of students nor could they ever be. (ii) There’s lots of evidence that grades are not very useful for decision-making, and there is lots of evidence that grades undermine learning. (iii) To the extent that grades are “useful”, they appear to measure privilege and other contextual factors more than skill or learning. (iv) Grading is dehumanizing, for both students and educators.

Here is a position piece penned by Dr. Kate Owens, Associate Chair of the Department of Mathematics at the College of Charleston, that lays out a strong case.

—–

Hello Colleagues,

A growing list of institutions making the P/NP leap: https://oudigitools.blogspot.com/2020/03/feedback-alternate-grading-in-crisis.html

I support the switch to P/NP by choice — either that of the instructor for a specific course, or of the students within courses. Here’s why:

I am concerned about our ability to deliver high quality online instruction in the given circumstances. In my courses, I strive to have course grades accurately reflect the demonstrated content knowledge of each student. Right now I’m teaching calculus so I will refer to that course. When assigning a grade of, say, 83%, what I’m really saying is that “Given the data at my disposal, I have a reasonable belief this student’s knowledge of calculus is somewhere between 81% and 85%; I assign a grade of 83%, and give a B-.”

This approach for grading does not work well in our current scenario. If you imagine calculus as a hurdling event at a track meet, the course has a predefined track length, a prescribed number of hurdles, and each hurdle has a preset height. Given the reality of the situation we are facing, all three of these things will need to be accommodated in some form (distance, number, and height). I cannot expect my students to do as well at mastering the content of calculus as they would have done before. Yes, I can modify how I assign grades; but to change the hurdles mid-race is going to be difficult and stressful on instructors and students alike.

Outside of my class, burdens will not be shared equally among my students. I worry about their stress of keeping or maintaining their grade at midterm. I worry that not all of my students have access to what they need to be successful in the course (high speed Wi-Fi? A quiet place to study? Time away from care-taking duties?). I worry that some of my “regulars” to my office hours will be less successful now that I am not as available as I was before. I worry about equity, both in access to learning and knowledge, and also in ability to master the associated content.

Additionally, I have some concerns about the type of measurement letter grades (B, B+, A-, etc.) impose. My uncertainty bars around any learning data will be larger than they were before. I am no longer confident that a student with an 83% average really knows between 81% and 85% of what I’m measuring. Given the new delivery system, the new assessment system, the lack of in person proctoring of examinations, and so forth, it’s probably more fair to say that an 83% really means I have reasonable belief that a student’s knowledge is between 78% and 88%. Reasonable grades to assign in this range would be C+, B-, B, or B+. It is not clear to me how I am to distinguish between them.

Lastly, I have seen the argument that if we make an institutional switch to P/NP that some students might do the least amount possible to secure a “P”. I think this is a feature and not a bug of the system. Our students are juggling a lot right now. They are rational actors. If we, as instructors, set a minimum bar for passing, and some students meet that bar, then they have done what we asked. I trust my students to make decisions about their priorities. On the other hand, I know several of my students will aim much higher, maybe because of personality, or curiosity, or because they are hoping for a glowing Letter of Recommendation at some point, or because they really love calculus (who doesn’t?!?). This is great, too. We need to trust our students to use their judgement about how best to allocate their time and energy.

—–

BK: Kate highlights many of the key themes I have seen in these discussions:

• Concerns about students’ access to computers, internet, and quiet time/space and equity of this access.
• The need to adjust goals. My current Provost encouraged faculty to identify the critical heart of our objectives and let go of the rest in support of achieving this primary goal.
• Changes in our ability to support students, which was always part of the course into which students enrolled.
• Allowing students to make the hard choices based on their contexts and priorities.

The other theme, which is hinted at in Kate’s letter, has been discussions of cheating. As long as we assess skills that can be demonstrated by putting forward the knowledge of others, there is and will be some cheating, including in whatever was planned for courses before COVID-19 forced us to revise our plans. Moreover, as long as we connect these kinds of assessments to high-valued outcomes like grades, we are setting up a situation in which it could be rational to try to cheat. I feel as betrayed by academic dishonesty as anyone when it happens, but it is a systemic problem. We could talk at length about redesigning assessments to address these issues, but I commend the efforts I’m seeing to envision either “open world” assessments that assume students will have access to many resources (scaffolded so that this doesn’t replicate privilege) or synthesis assessments for which the demonstration is the meaning-making rather than results and hence is largely independent of resource use. More broadly, I support the efforts to re-envision assessments as opportunities for students to show their skills rather than comprehensive measurements.

My personal stance is that any system that requires students to engage in our courses during this crisis is holding the first half of this semester hostage and potentially forces them to choose ongoing coursework over their own health and the needs of the communities in which they live. There will be loss of “productivity” in a global pandemic, and whatever meaning you give to the problematic term “rigor” is clearly met by students’ learning under duress. For me, the question is who has to make the hardest choices about which losses to accept, and do they have enough power in their systems to make that choice freely. Faculty and some institutions have suggested extending the withdrawal deadlines even past when grades are submitted, which mitigates the consequences of some choices in ways I like, but still makes credit for the first half of the term contingent upon continued work.

My preference would be that we allow students to accept credit for smaller courses for the first half of this semester (eg converting 4-credit courses to 2-credit courses, although assessments in these last two weeks have certainly been inequitable), allowing them to walk away and choose to focus on their other needs, and then offer supplemental, non-credit opportunities either for those who want to do some learning in this time [though I am concerned about this replicating privilege]. I also think it would be good to declare the semester over, giving students credit if they were in good standing and offering very generous incomplete or withdrawal options for those who need a little more time to get there, largely as Berea College seems to have done. Some faculty have announced that final grades will be no lower than current midterm grades, which accomplishes much of this same goal.

Barring that, Kate and I support what Smith College is doing: making all course grading Satisfactory/Unsatisfactory. This is much like Kate’s recommendation that students or faculty can elect to do this for any course, but we prefer it because it shifts the challenge to the institution rather than individual faculty and students. The Princetonian recently advocated for something similar. In Smith’s implementation, letter grades will be reported but not included on transcripts, so students who want these letter grades shared (e.g. with graduate schools) could have that done formally from the Registrar. We prefer a formal move of all courses to Satisfactory/Unsatisfactory or Pass/NoCredit because individual faculty and students are not in a position to move between the options at this time. We also think that faculty will design better courses knowing that all students are in a P/NC course, but this option seems to have all of the same choices for students and might allow faculty to put their limited work time into other critical course issues.

A few people have responded to me or Kate or others advocating these ideas with concerns that this move will hurt students who need a certain GPA, who can’t have P/NC courses count toward their major, and who might lose NCAA eligibility. These are important issues to consider as we make decisions, but all of these examples are issues within our power to change. As Kate put it:

“If we can’t do something in the best interest of the mental and emotional health of thousands of people because of (obscure regulation), then I maintain the regulation should be expected to adapt, not the people. We are all being forced to adapt to stressors and situations none of us ever imagined a month ago. We need to lighten the burden felt by all of us. Cut the red tape — trust me, it’ll be easier to repair that than emotional baggage when things go back to normal (which I hope is soon).”

For example, at Smith a D is coded as Unsatisfactory; at the roll-out of these changes, the Provost was explicit that students for whom course credit from a D would have been sufficient will have that manually adjusted despite what the transcript reads. And perhaps we simply treat Pass this semester as 4.0 in terms of any GPA computations we are forced to make.

One concluding thought that is helping us: future courses will have to adapt, no matter what. There is no version of our spring courses that will allow the fall to be “normal”, and students will have different experiences from each other. Moreover, we always need to meet our students where they are. When building courses, we need to keep the prior diverse experiences of our students in mind, and build the courses in such a way as to invite them in. So we should make choices now that treat students and faculty with respect and humility and that allow them to take care of themselves. And perhaps this is a moment to acknowledge that we should always do this, and that grades are (always) interfering with our ability to do that.

[Editorial: I have chosen not to tag individual faculty in the comments above because of the potential to increase their workload today.]