There’s a strange combination of facts in teaching:
Fact #1: You have virtually infinite flexibility in writing an exam. You can make it as hard as diamond or as soft as chalk. You can offer copious partial credit or none at all. You can ask questions identical to the homework, or questions of baffling novelty. You can give the exam in-class, as a take-home, open-book, closed-book, group-style, or as an Instagram post on which students comment their answers and your score is the number of likes you accumulate.
Fact #2: Whatever the resulting exam, 90% means A, 80% means B, and 70% means C.
These two facts are the parents of an ancient and venerable custom in education: the curve. If the exam scores are not resulting in the grades you want, then what do you do? Turn them into new scores!
In that spirit, I give you…
(To answer a common question from Instagram: this only makes sense if you read the grades as percentages. Example: 60 is actually 60% = 0.6, which gets curved to 0.6^0.7, or roughly 70%.)
(Again, we’re reading these grades as percentages. And the sin(x) function takes them as radians, not degrees.)
Always enlightening and entertaining. I only wish there were a Spanish version. There are potentially 500 million new subscribers. My wife is a Math teacher in Panama and she would love this, if only she spoke better English.
Or, if you are a professor at a reasonable university (a shrinking number of places), you can set your thresholds to whatever makes sense for the exam. I routinely designed my exams so that 50% was the target average grade and 15% the target standard deviation, so that I could get information about the majority of the class, rather than spending most of the exam points on distinguishing the failures from the abject failures. Although this confused a few students (who thought that whatever grading scale they had experienced in high school was a universal law of nature), they adjusted fairly quickly.
Other exams, like the Putnam math prize exam, are designed to rank people at the top end and may have a median score of 0—the goal there is to have an extreme-value distribution with no one (or only one person) getting a perfect score.
1. Other than Mastodon, I ignore corp-controled social media. Sorry. (Yes, I am a Canadian.)
2. Marks, I tell my classes early on, are not the goal. They will not be remembered by the vast majority in ten years. It is (wistfully) hoped that some of the subject matter may be retained for the long term.
3. Marks serve a few purposes:
a) As a game, to focus attention
b) For administrative purposes (I am writing a short paper on bureaucratic bloat in which I draw attention to the fact that academic institutions, prior to the 1980s, funded “learning” through 70 to 85% of the total budget, whereas the current situation has flipped, so the administration receives 85 to 95%)
c) As markers to indicate learning progress
d) As a way of very minor boosting of certain student’s low scores to offset learning difficulties (sorry – I am genetically Canadian).
This is such a perfect take — the wild flexibility of test design paired with the rigid expectation of grade bands is almost comical when you step back and look at it. Curving always feels like this quiet acknowledgment that the “percentage” doesn’t actually mean what we pretend it means. Loved this both funny and uncomfortably true for anyone who’s written exams!
This post rocks for site fans! Hypackel
The Eggy Car’s straightforward design is among its most intriguing aspects. The game Eggy Car appears upbeat and welcoming due to its vibrant and colorful graphics. Players typically only need to press two buttons to go forward or backward, making the controls quite simple.
The mismatch between endlessly adaptable test questions and the fixed illusion of precise grading is hilarious in a slightly painful way, and your take captures that blend of truth and irony perfectly.
slope 2