## Monday, April 27, 2015

### on the toxicity of grading curves

At my university, a campus-wide "curving" policy exists. The grading curve varies slightly, depending on the course being taught, but the basic idea is that the grades that the students in a given class receive must conform, more or less, to a rigid bell-curve distribution. Because I'm a big believer in meritocracy, I find the grading-curve policy to be, as Tom Cruise said in "A Few Good Men," galactically stupid. To my mind, if all my students perform at "A" level, they should all get "A"s; if they all perform at the "F" level, they should all get "F"s. Merit, pure and simple. (That said, the subjectivity that comes with assessing/evaluating language output is a topic for another post.)

In the classes I'm currently teaching, there's only one grading curve, and it goes like this: up to 30% of the class may receive "A"s; up to 70% of the class may receive a combination of "A"s and "B"s; the remaining 30% is open to "C"s, "D"s, and "F"s. I have four classes; my first class has only fifteen people in it; the other three all have nineteen. If we do the math, then:

Class 1: 15 people * 0.3 = 4.5 people. You can't have half of a person, so drop the 0.5, and you see that, in a class of fifteen, only four people may receive an "A" according to this curve. Up to ten people may receive "A"s and "B"s. Five people must receive the lower-tier grades.

Classes 2-4: 19 people * 0.3 = 5.7 people, which rounds down to five "A"s maximum, and thirteen "A+B"s. Interesting to see that the number of possible "A"s differs by only one person between Class 1 and Classes 2-4.

Right now, Class 1 is my best-performing class. On the midterm, there were two "A"s, ten "B"s, and three "C"s. The overall semester grades for all fifteen students, however, is "A." Why? Because up to now, the only grades I've given have been "chump" grades in categories where it's easy to earn a 100%. Attendance is weighted at 10% of the final grade, and all my Class 1 students have a 10 out of 10. Participation is 10%; all students have 10/10. Homework? This is 15% of the final grade, but the students have all kept up with homework, so again, everyone's got a 15/15. It's only now—with real evaluations happening in the form of big tests and projects—that we start to separate the men from the boys.

But that's precisely the toxic mentality I'd prefer to avoid. Unfortunately, the curve has me thinking in such Machiavellian terms: because my students have to fit the curve, they obviously can't all get "A"s: I have to design tests that are inordinately difficult in order to create a bell-shaped spread. Only a select few can earn "A"s; everyone else must be consigned to the no-man's land of "B"s, "C"s, "D"s, and "F"s.

I don't want to design hard tests: I want to design fair tests—tests that evaluate the students as objectively as possible based on the material they've covered. Instead, I find myself thinking about how to make the final exam as hellish as it can be. The risk, of course, is that I might end up going overboard: I could, in theory, design a test that's so hard that no one passes it, and as a result, no one gets an "A" in the end. That would create hard feelings analogous to what happened last semester (even though I'd say I graded quite fairly last semester). Not that I'm primarily worried for myself; it's more that I'm worried about whether inappropriate test design is going to shortchange these students, depriving them of a decent education.

The students themselves have had to live with something like this curve for all of their academic lives. Competition is fierce and Darwinian in Korea; in the end, it's all about your school's prestige and how you rank among your peers. The toxicity isn't limited just to thoughts about the curve: many students, as I discovered last semester, develop an overweening, greedy sense of entitlement that articulates itself as a kind of misplaced self-esteem: "I deserve an 'A'! Why on earth did I get a 'B'?" And that's when I get the angry, desperate emails at the end of the semester—the ones in which students beg for, or outright demand, a better grade. The transformation is an ugly one: students who seemed polite and reasonable in person for 99% of the semester suddenly morph into grade-grubbing ogres in those final days.

Another problem with the curve is that it promotes an atmosphere of secrecy. In an ideal world, I would have no trouble uploading my students' grades to a campus database that they can consult to see how they're doing. (In fact, we have such a database; its use is optional.) The problem is that, in order for all the students to fit the curve, I'll probably have to make last-minute adjustments to their grades. If there are too many "B"s, for example, then some of those "B"s will have to be reduced to "C"s. If a "B" student has been following his or her progress via an open database, then a sudden "C" will come as a shock. It's in my interest, therefore, to maintain a shroud of secrecy about how I calculate final grades, and to be vague when students start asking prying questions near the end of the semester.

I've tried to be as up-front with my kids as possible, however. I've told them, from the beginning, that some student might unjustly be pushed down from an "A" to a "B," or from a "B" to a "C" just to fit the curve. But even though the students are aware this might happen, experience has shown me that, when it does actually happen, they still get upset.

Teaching is often called a thankless job for a reason. The institutions in which teachers teach often do little to make life easier. The grading curve is one example of the many burdens that pedagogues bear. It's said that curving counteracts the problems associated with grade inflation. I understand that grade inflation is a potential danger when teachers are allowed free rein (getting a "B" in American grad school is a seppuku-worthy dishonor, for example; grad-school grades are often way inflated), but in my opinion, the curve creates far more problems than it solves, and it does nothing to improve teacher-student relations.

_