3 ms·

> No one was pulling all-nighters two weeks into the semester. As a physics student, I can say we definitely pulled 2am-ers 2 weeks into the semester. Our Cond

> No one was pulling all-nighters two weeks into the semester. As a physics student, I can say we definitely pulled 2am-ers 2 weeks into the semester. Our Condensed Matter professor gave us some of the most random questions that no one could anticipate. To be honest, I was one of those "diligent" students who got good grades. I'll tell you right now his quizzes were not hackable. The following held true: |quiz subject matter| >> |lecture matter| He tested on the former, so we had no way of memorizing or hacking the quiz. You had to really understand the concepts to do well on his quizzes. Ditto for my grad Quantum Mechanics class. On the other hand, I've also had professors who give exams pulled from the internet and allow use of the internet while taking those exams. I hacked that immediately during the exam, then almost got in trouble for it. > ...if the professor tells you that there were three underlying causes of the Schism of 13... you'd better know them. A professor like my Condensed Matter Physics professor would go on and on about 3 fundamental causes and then quiz students about the 5 fundamental causes, giving a zero to anyone who couldn't think of all the base 3, then grading the remaining students who thought creatively of 2 additional causes against each other on a bell curve. :)

majos 7y ago
> A professor like my Condensed Matter Physics professor would go on and on about 3 fundamental causes and then quiz students about the 5 fundamental causes, giving a zero to anyone who couldn't think of all the base 3, then grading the remaining students who thought creatively of 2 additional causes against each other on a bell curve. :) To me, this sounds like a plausible argument against such tests. The issue of creatively grading responses to an ill-defined question often pops up in discussion here about interview practices. In those discussions, typically someone will say “I don’t do a generic whiteboard interview. Instead I do [idiosyncratic thing x]. It really gives me amazing insight into the candidate.” Then someone else says “Yeah right, a weird test with unclear metrics just gives you a big empty space to fill in with all your biases and pick someone who answers the way you would”. Of course, both sides are exaggerated here. But it’s not clear to me that “creative” tests are necessarily any better.

majos 7y ago

> A professor like my Condensed Matter Physics professor would go on and on about 3 fundamental causes and then quiz students about the 5 fundamental causes, giving a zero to anyone who couldn't think of all the base 3, then grading the remaining students who thought creatively of 2 additional causes against each other on a bell curve. :) To me, this sounds like a plausible argument against such tests. The issue of creatively grading responses to an ill-defined question often pops up in discussion here about interview practices. In those discussions, typically someone will say “I don’t do a generic whiteboard interview. Instead I do [idiosyncratic thing x]. It really gives me amazing insight into the candidate.” Then someone else says “Yeah right, a weird test with unclear metrics just gives you a big empty space to fill in with all your biases and pick someone who answers the way you would”. Of course, both sides are exaggerated here. But it’s not clear to me that “creative” tests are necessarily any better.