Learning and Ranking

Should colleges keep the pass-fail grading system that many adopted in spring of 2020?

Rebecca Schuman makes a good case for it. Her piece is well worth the read, but the very short version is that purely extrinsic motivation gets in the way of learning. When students don’t have to focus on the difference between a B-plus and a B, they can take some risks. Getting away from grading generally allows something much closer to a competency-based approach, in which students can try as many times as necessary to master a skill. And although Schuman doesn’t linger on this point, pass-fail also gets away from some of the false precision in distinguishing a B-plus from a B.

She points out, correctly, that the bar exam to practice law is pass-fail. Pilot certifications are pass-fail. (Nobody wants a pilot who got a C …) As The Girl can attest, driver’s license exams are effectively pass-fail. Dissertation defenses are pass-fail. There’s no lack of precedent for pass-fail measures in areas that we, as a society, consider important.

In my own undergrad years, January term courses were pass-fail. I loved them. They allowed students to take some risks. I took a course on gay and lesbian politics in 1988, when even the course title raised eyebrows. I remember much more of it than I do many of the semester courses I took for grades. It was one of the best courses I ever took, even if all I got on my transcript was a pass.

I’ll also admit that grading is, by far, my least favorite aspect of teaching.

In spring of 2020, we went to pass-fail in light of the monumental disruptions caused by the pandemic. The sky did not fall, though I did notice a dip in the pass rates for the NCLEX exam in nursing the following year. How much of that was due to pass-fail and how much was due to students and faculty being overstretched (or even due to random variation) is impossible to say.

The experiment didn’t last long. As Paul LeBlanc notes in his excellent new book Students First, entire systems within higher education are built on the existing infrastructure of credit hours, letter grades and semesters. For example, federal financial aid requires that students maintain “satisfactory academic progress” to be eligible for aid; they also have to complete courses within set amounts of time. In the absence of grades, SAP needs to be determined differently. Many scholarships are based on GPA; take GPA away, and they’d have to be restructured. (To be clear, LeBlanc advocates for creating space for a competency-based system that’s effectively pass-fail; he’s just honest about the obstacles to creating one.)

Neither Schuman nor LeBlanc spends much time on the ranking function of grades. I don’t blame them—it’s not my favorite topic, either—but ranking serves multiple purposes. My own college didn’t introduce letter grades until the 1990s, previously relying on a “mastery learning” system that bears a striking resemblance to what’s now called competency-based education. The primary reason it moved to grades, I’m told, was transfer. GPAs are legible between institutions; from a student perspective, they’re almost a form of currency. A student with a 3.8 GPA and a student with a 2.2 GPA appear the same when reduced to pass-fail, but most of us would probably assume that the 3.8 student is academically stronger. When students tried to transfer without grades, they ran into all manner of roadblocks and often found themselves ineligible for transfer scholarships that were based on GPA.

Eliminating a method of ranking doesn’t eliminate the demand for ranking. It just creates an opening for a new form of ranking. That could be good or bad, depending on the alternative. A graduate program (or law, or medical, or …) that gets, say, 100 applications for 15 seats needs to base its decisions on something. Standardized tests are already considered highly suspect; if you throw out both tests and grades, selection will happen some other way. The recent wave of test-optional admissions at selective colleges doesn’t mean they’ve thrown open their doors; in many cases, their acceptance rates have hit new lows. They’re just using different criteria. The need to narrow down the field remains. Without specifying what the preferred criteria are, it’s hard to say whether the change would be beneficial.

At the level of the individual course, that may be fairly abstract. Within a given class, pass-fail can be a blessing. It’s also more intellectually honest, in many ways, than trying to decide whether a given paper merits a C-plus or a B-minus. When the pandemic hit, I advocated for statewide acceptance of pass-fail grades in transfer for that semester, and I stand by that. In a semester in which most courses changed modality halfway through, the charge of false precision carried new weight. But moving to a permanent pass-fail system at the community college level—where transfer is a core function—could put our students at a real disadvantage. Given the elitism to which academia is prone anyway, it could become that much harder for our best students to get to the next level.

Ranking student performance is distasteful in many ways, but at this point, the external demand for it remains strong. Extrinsic motivations exist, whether we want them to or not. Fixing that would require some much larger changes than any grading (or nongrading) system could deliver. I’m very much in favor of creating spaces in which students can take risks, and in which we don’t have to pretend to a precision we can’t justify. But it has to happen systemically, or we could wind up putting good students in a bad spot.