Behind the Curve

A curve is a mystical machine that takes a 60% and ramps it up into an "A." Seemingly, then, a boon to students everywhere.

Yet, grades properly ought to reflect one thing, and one thing only: absolute mastery of material. At the heart of curve-based grading, on the other hand, is the enticing but pernicious premise that grades should instead reflect relative mastery of material; in particular, whether a student is on par with or lagging egregiously behind his peers.

In general, except under extreme applications of the scheme, nobody has reason to notice the problems underlying curve-based grading, and therefore nobody complains. Indeed, I’ve personally observed that many applications of the scheme are so thoughtfully implemented that the problems I’m about to list become practically nonexistent. Still, there’s always a potential for these problems to arise.

On the student-side, problems stem from the bare fact that students like high grades. At this point, a phenomenon of human nature comes into play, and can quickly turn class into a bizarre Keynesian beauty contest. The phenomenon is the same one at work when high schoolers obsess over which SAT date and test center to choose in order to obtain the most generous curve: If you dangle before people a reward for gaming the system, game it they will. The intentions implied behind flippant calls of "Everyone fail the test!" usually find themselves abandoned at home on exam day, but there is truth in the assertion that students think competitively in pursuit of high grades. Curve-based grading retards their education by diverting their effort from learning to finagling their transcripts.

Especially at top-tier institutions, furthermore—where many students are accustomed to receiving high grades—when learning and grades become disparate instead of directly interchangeable metrics, students can be driven into a mercenary mindset and pursue grades exclusively. In the absence of a curve, however, the two are inextricable. You’re forced to get all or none.

On the teacher-side, on the other hand, more severe problems arise from the fact that a curve obfuscates instructional feedback. It can become tough to readily perceive whether your students are learning effectively if you curve a 20% up to a perfect score and assign everyone "A’s." Whereas it’s only sometimes true that this discourages teachers from investing effort into their teaching, it is always true that this reduces pressure on teachers to expect mastery of material from their students.

This reduction of pressure on teachers to expect mastery can even, ironically, elevate pressure on students: Instead of painstakingly curating manageable syllabi or writing reasonable exams, teachers can set overwhelming syllabi or exams and then smooth things over with their pupils by curving everyone up. Students’ transcripts remain flawless at the end of the semester, so they of course don’t complain. But if they haven’t mastered the entire syllabus, what assurance do they have that they haven’t missed something crucial?

Curves, moreover, destroy comparability, as things that look the same end up different, and things that are different end up looking the same.

Specifically, individual students’ experiences in the very same course can end up different, as curves vary from section to section or year to year. This can complicate matters as students shop courses.

Along this same line, but opposite in effect, is the distortion that results in GPA, supposedly a standardized evaluative scale. Two different classes, one full of shiftless slackers and the other full of diligent geniuses, can appear equivalent. And, most concerning of all, a shiftless slacker can be curved up to become, at least GPA-wise, indistinguishable from a diligent genius within the same class. Because there’s no straightforward way to determine precisely which curve a particular student has faced, it’s natural for an observer to ignore the differences entirely.

The effect can be especially insidious when tasks for a class begin to verge upon the impossible—an exam that a teacher declares was never meant to be finished, for instance. Admittedly, the scenario is rare, but it reveals an interesting result of curve-based grading. Consider: In these cases, because the task is so infeasible, talent and effort can become obscured as the difficulty of earning an additional point increases exponentially, and therefore each additional point requires exponentially more talent and effort, while the visible payoff to score continues increasing only linearly. Because scores can then end up separated by a small distance, two students separated by a large degree of competence are likely to become lumped together by the curve.

I must emphasize that the blame for these faults doesn’t correctly belong with students or teachers; rather, the flaws are intrinsic to the curve-based grading scheme itself. Curve-based grading, then, is not the unmitigated blessing it might initially seem. Even if the scheme persists, these issues are worth considering as we use it.

Sparsh Sah '19, a Crimson editorial writer, lives in Stoughton Hall.