Grades. Hot topic right now. And honestly, this is my least favorite time of year. I have to bite my tongue almost daily.
“What’s my grade in the class?”
Response I can’t give: What’s preventing you from figuring this out yourself? Maybe something that indicates your grade shouldn’t be too high?
“Is there going to be any extra credit?”
Response I can’t give: Sure. Here’s some extra credit: calculate your current grade in the class.
“Is there going to be a curve?”
Response I can’t give: you don’t actually know what a curve means.
Currently the “curve” thing is what irks me the most. The word makes me cringe.
Let me explain why.
First here’s what I believe a (graded) assessment should accomplish:
(1) determine the mastery and retention of specific material covered in class by an individual in the class
(2) determine the ability of an individual student to apply class material to new (to them) situations.
Depending upon the students, the material, and the institution, these two aspects may be weighed differently. For instance, at Carnegie Mellon when teaching a 400-level number theory course, I leaned more towards (2); next semester most students were proceeding to grad school. The final was predominantly filling in the details of proofs and examples “left to the reader” in published papers—an activity graduate students often need to perform. In a 100-level business calculus class at UGA, the emphasis was more (1). It was a service course to a non-science department, and algebraic proficiency and “basic” applications were what was deemed by people far above my pay grade to be most relevant. We were basically glorified tutors.
While when you first start teaching this balance can be difficult to manage, it shouldn’t take too long to figure it out. Yet when a faculty member says they “must” curve an exam it’s often because they’ve failed at balance, probably because they didn’t think about it much or at all when writing the exam. If faculty regularly give exams without any qualms about later “having to curve”—and we all can name plenty of people in this category—then they’re doing students a disservice. If the exam is comprised of predominantly “cool” problems that may not actually apply to the course at hand, the professor is not thinking about the students or how best to prepare them, or the college and its expectations. And think of the psychological and pedagogical message it sends to students (especially in lower-level courses) that they should “expect” a raw score of (say) less than 50%. That that’s actually somehow “passing.” Or that they should “expect” a curve, regardless of initial outcome.
On that…since curves often change grades, let me say what I believe is the purpose of a grade.
A grade is meant to provide student, current instructor/supervisor, and future instructors/supervisors with a measure of the student’s individual proficiency and mastery of the course material. This course material should be readily accessible via online course catalogs so one could, say, start to distinguish between a “C” and an “A” in two similarly-named courses at two different colleges.
But herein lies the first major issue. Is the proficiency and mastery absolute, or relative? That is, does an “A” indicate mastery of some list of topics, or does an “A” indicate mastery of some list of topics relative to everyone else in the class/school, or even relative to the individual student’s starting point?
This question is ESPECIALLY relevant now. Many during COVID have thought students need to “catch a break” when it comes to grades. More holistic (i.e., relative) grading systems have been suggested.
I think it’s dangerous to make grades relative, even during a pandemic. Why? With respect to the pandemic, people are going to forget soon enough. From a promotion and tenure standpoint, many schools are already forgetting what impact this has all had on course evaluations and research, or they think the impact is at most one-semester’s worth.
But suppose it’s “the before times” or “the after times.” Statistically it is entirely possible that a particular class is “exceptional.” It happens, though rarely, that a class honestly has 30% functioning at an “A” level. Similarly, there could be classes where 0 students perform at an “A” level. But if grades are relative, those two classes would have eerily similar GPAs. Certainly both classes would feature at least one “A”.
Another reason relative grades are dangerous is psychology. If students know that it’s their ranking relative to their peers that determines the grade, competition will skyrocket. Some level of competition may be healthy, but it can quickly spiral into a cut-throat, toxic, grade-obsessed environment. Think of the classic scenarios where people are constantly ranked: med school, law school, and the military. Those are high-stress professions filled with people who mentally and/or physically already perform well above the “average.” Probably shouldn’t treat any and all (under)grad classes like they’re full of officers, lawyers, and surgeons.
There’s one other big reason I’m so gung-ho on not grading relatively: the future instructors/supervisors. They weren’t there; they won’t know what the “average” was. They don’t know what you’re grading relative to. Regardless of COVID, the person teaching Calc II at my institution won’t know what my Calc I class was like. And if I give a Calc I student an “A-” because of their relative class performance, but in terms of overall mastery of concepts on the department syllabus they function at a “C+”, how is that going to end well for anyone?
I’ve worked in enough departments to see enough fights with those teaching the calc sequence or those teaching whatever immediately follows intro to proof on this exact issue. The claim is that students are being passed who “shouldn’t” be passed. Students are getting high grades when they are ill-prepared for the next course. The term “grade inflation” is tiptoed around, yet that’s basically the accusation.
But hold up…grade inflation is different from curving, right? Yes in theory, no in practice. Traditionally, grading on a curve meant normalizing the grades literally to fit the normal Bell curve. This is a highly relative grading scale where a “C” is an average and there are equal numbers of As and Fs (try telling students that’s what curving is!). With this “fit to a Bell curve” mentality, technically speaking the drastically not-normal situation of a 60% average and no grade above an 80% would be just as problematic as the other extreme of non-normality of an 85% average and no grade below a 70%.
Of course, this is not how curves are used in practice. In practice, it’s only the first instance of 60% average that’s deemed eligible for a curve. There is no curving “down.” In practice there’s a mixing of the absolute and the relative scales to the benefit of the students’ GPAs (and also possibly the faculty course evaluations). And because there’s never a curve down, students begin to expect more and more the beneficial-to-them situations where averages are “B”s.
When students ask if there will be a curve, they are not asking you to risk making the class more competitive by enforcing a system where only the top x% receive an A. They are not asking for the same number of As as Fs. They are asking (slash demanding depending upon their aggression) for their grade to be relabeled with a higher letter.
Perhaps one of these days I’ll understand when a student asks for a curve versus extra credit. Both are basically attempts to raise grades. It’s interesting when the student specifically is willing to put in work for it.
Still, students and faculty alike will tell you that a curve will and indeed SHOULD raise the grades of everyone in question [definitely implying a one-way normalization.]. And while mathematically it’s usually true curves raise all grades they don’t do so uniformly. Different curving “systems” benefit different students in different ways. When an instructor curves, do they curve in the way that benefits most the students who asked most loudly for the curve? Do they curve in a way that benefits the most students period? In a way that benefits the students with the worst numerical grade? And no matter what system is used, there’s always a notion of a curve “breaker.”
Again, no one would think to curve a situation where 60% of students earned over an 85% on the exam; indeed fewer and fewer see a problem with that scenario in the first place, writing it off as “the students are exceptional” (again, statistically rare), and never considering other options like “the instructor needs to rethink their exam writing” or “maybe more topics can be covered.”
And that reality is what makes curving in practice akin to grade inflation. That reality makes me understand more the faculty who have no qualms about curving exams on a regular basis. If the grades are going to be effectively “made up” anyway, why not write a “fun” exam with a 40% average? Who cares—you’ll fix the numbers later just like an accountant at Arthur Andersen. Better than that: you’ll be an “unsung” hero to all those students giving you anonymous evaluative comments later on. Write an exam with a 40% average to make yourself feel good and intellectually superior, tell students that anyone who scored at least a 40% earns at least a C, and you’ll be a God to the class for turning their “F” into a “B.” Everybody wins.
Except there’s never a situation where everybody wins. The faculty member who goes rogue and decides not to curve when the rest of the department does loses. The faculty member teaching a follow up course to a bunch of 40% = C students loses. The students when they finally reach a time in their lives when there is no curve lose.
And that is why I cringe every time I hear “curve.”
Happy grading, everyone.