What are we actually being graded on?

A massive hint tucked away in an instructor’s reply to a follow-up on a Piazza post. Eight hours wasted on a problem set because you couldn’t make it to office hours where they told attendees a big clue. Review sessions where the teaching assistant walks through a problem virtually identical to one on the midterm, but you were at sports practice or a job interview or sick at McCosh Health Center. Coding assignments where having beefed-up hardware would have provided a significant advantage. Even in supposedly “objective” STEM fields, grading can be a befuddling, frustrating, and often inequitable process. Add to this the common complaints of humanities students about the subjectivity of grading in their fields — how one preceptor grades more harshly than another, or how some professors use an indecipherably cryptic rubric to assess papers, or even how writing style can shroud a strong argument — and we see that the difference between supposedly good and bad grades at Princeton is often arbitrary.

I’ve had the experience of being given a small set of times to pick up a final exam and potentially contest scores. By sheer luck, I noticed an addition error on my exam’s score sheet, and successfully got my score raised from a 62 percent to a 72 percent — a significant difference in that class. If I hadn’t been free at that time or hadn’t discovered the error until after I left the professor’s office, tough luck — I would have been stuck with the grade I originally received.

Sometimes it seems like the only way to get a good grade in a class is to enable push notifications from Piazza and read them religiously, to build your entire schedule around course office hours (and not just one instructor’s office hours, but all of them, lest one be more tight-lipped than another), and to scrutinize every sub-part of every graded exam problem for scoring errors. In many cases, it is not nearly enough to simply attend all lectures, do all problem sets, and know all the concepts covered in class. In order to differentiate the dozens of bright and hardworking students in a class, the grading system doesn’t measure how much a student learned but how much he or she is willing to sacrifice for a class. For some students, this sacrifice will be sleep. For others, it will be extracurriculars. For yet others, mental health. Rather than a straightforward assessment of work done, some classes operate by attrition, punishing students who are not willing to make that class their number one priority so that the dedicated few can get their reward.

This sort of maneuvering to increase one’s grades doesn’t actually impart more knowledge. In the crapshoot of preparing for exams, having an upper hand on doing a certain problem doesn’t mean you actually know the material better, and you almost certainly won’t remember the test problems a year from now. The 12 hours of debugging that it took to get a program from a state of not even compiling to finally running cleanly may raise your grade from a D to an A, but does it really increase your knowledge by a corresponding amount? Yet I sense that this perceived need to excel not only in mental growth but by the metric of grades contributes much to the stress of the typical student.

The arbitrariness of grading strikes me as a problem that is not going to go away anytime soon. The Princeton academic machine is composed of human beings, and it’s an intractable problem to ensure that every course instructor is applying the exact same standards. Let’s face it — grading is not fun, and it’s not what most instructors became professors or graduate students for. This doesn’t exonerate them from the duty to try to be fair and reasonable in grading practices, but it should encourage students to simply put less emphasis on grades. After all, for many career trajectories, GPAs matter far less than research quality, personal projects, and strong letters of recommendation. In light of this, I hope course staff eventually spend less time devising assignments to spread students out on the grading curve, and more time engaging students’ academic passions. And since it is ultimately necessary to give grades, courses could reduce much unfairness by following three broad heuristics: Create assignments that build incrementally on previous assignments, don’t privilege any groups or individuals in any way during the assignment process (from the classroom to office hours to grading), and attempt to make grades a roughly linear function of time and effort put in by the student.

Students, on their part, can respond by fostering a less grade-conscious culture and promoting a greater appreciation of learning experiences for the sake of learning. If you go to class and learn something new or work on a project and gain valuable practical experience, this is in many ways its own reward. If an interesting topic comes up in lecture that leads you to search for new types of jobs and internships, I’d say you have gotten something out of the course. If a course leads to a productive and meaningful collaboration with a professor or graduate student, it has been time well spent. The grade you get is often only loosely correlated, if at all, with how much you benefited from the class. This is how I have justified no longer caring much about what grades I get. My mental health and sense of purpose is better served by doing my best in my classes but not caring how I’m graded and devoting more time to other pursuits that matter to me. If enslaving myself to a class is the only way to get an A, I’d rather just not get an A.

Thomas Clark is a junior studying computer science from Herndon, Va. He can be reached at thclark@princeton.edu.