Christian speaks truly. I don’t think there’s a write-up.
I can give a very quick version here. One minor correction to the query though: I didn’t have the students grade each other. I had them grade themselves. Part of the whole point was to tighten the feedback loop the students were getting so that the delay between “I tried this math problem” and “Here’s what you got right, and here’s what you could do to do better going forward” was as short as I could imagine making it given the constraints. I also wanted to give them mental practice correcting their own work.
So with that, here’s the method in outline:
I taught two 90-minute math classes a week, Tuesday & Thursday afternoons, with a group of about 30 students. (This means I could see them all.)
Every Thursday there was a quiz. I didn’t grade homework, but I drew inspiration from the homework to create quiz questions.
I set up the quiz questions on a PowerPoint. Students were to answer one question at a time on their papers (which I’d printed out for them) using a black or blue pen.
After giving them some time with the question, I’d have them put down their black/blue pens and pick up their red pens. I’d then click the PowerPoint forward to reveal the answer. We usually had a brief discussion to answer questions or get clarification about whether a certain thing “counted”.
We’d go back and forth like this until the end of the quiz, at which point I’d have them hand in their papers.
My weekly flight up to Berkeley was on Thursday nights, so I usually reviewed their papers in the airport or on the plane and entered grades as needed. Part of this was to look for obvious cheating, and part was to help me stay familiar with the work and style of each student.
Some bugs I noticed:
Sometimes I’d discover during the quiz that our discussion after answering one question would make the next question downright trivial. It’s long enough ago that I don’t remember clear examples. But there were a few cases where I wondered about just flat-out skipping a question I’d put up there.
Obviously, cheating is a potential issue. I consciously decided not to care. I figured it was rare enough to not worry about, and that if it happened and I missed it then so be it.
It takes a long time during class to do this. A 10- or 15-minute quiz will take well over an hour. I’m not sure this is really a bug though. This meant the students were spending class time practicing the right skills. If the goal of a class were to train the students to actually have the skills, this method seems pretty good and could probably become excellent with iteration. But at the time, it was quite tricky because in practice I had only about half the lecture time of a normal iteration of the course, and I was still expected to “cover” the same amount of material. So the curriculum sometimes felt rushed outside the quizzes, and I felt a kind of pressure to make the quizzes take less time in class somehow. (That said, this is an artifact of classes caring more about talking about ideas than about encouraging students to actually master a level of skill. This conundrum is secretly baked into most undergraduate math courses I’ve seen.)
Students end up practicing a different skill than the one they’ll need for the final exam. Time- and question-management mostly don’t appear in this style of doing quizzes. Most college students have plenty of practice with that anyway, so maybe this isn’t actually a problem. I just flat-out don’t remember whether my students had a problem with the final exam this way. (That would have been May 2012, which was also when CFAR was running its first workshop. I cared a lot more about the workshop than I did about those students’ grades.)
I’m sure there are others. This is off the top of my head, inspired from memory.
Christian speaks truly. I don’t think there’s a write-up.
I can give a very quick version here. One minor correction to the query though: I didn’t have the students grade each other. I had them grade themselves. Part of the whole point was to tighten the feedback loop the students were getting so that the delay between “I tried this math problem” and “Here’s what you got right, and here’s what you could do to do better going forward” was as short as I could imagine making it given the constraints. I also wanted to give them mental practice correcting their own work.
So with that, here’s the method in outline:
I taught two 90-minute math classes a week, Tuesday & Thursday afternoons, with a group of about 30 students. (This means I could see them all.)
Every Thursday there was a quiz. I didn’t grade homework, but I drew inspiration from the homework to create quiz questions.
I set up the quiz questions on a PowerPoint. Students were to answer one question at a time on their papers (which I’d printed out for them) using a black or blue pen.
After giving them some time with the question, I’d have them put down their black/blue pens and pick up their red pens. I’d then click the PowerPoint forward to reveal the answer. We usually had a brief discussion to answer questions or get clarification about whether a certain thing “counted”.
We’d go back and forth like this until the end of the quiz, at which point I’d have them hand in their papers.
My weekly flight up to Berkeley was on Thursday nights, so I usually reviewed their papers in the airport or on the plane and entered grades as needed. Part of this was to look for obvious cheating, and part was to help me stay familiar with the work and style of each student.
Some bugs I noticed:
Sometimes I’d discover during the quiz that our discussion after answering one question would make the next question downright trivial. It’s long enough ago that I don’t remember clear examples. But there were a few cases where I wondered about just flat-out skipping a question I’d put up there.
Obviously, cheating is a potential issue. I consciously decided not to care. I figured it was rare enough to not worry about, and that if it happened and I missed it then so be it.
It takes a long time during class to do this. A 10- or 15-minute quiz will take well over an hour. I’m not sure this is really a bug though. This meant the students were spending class time practicing the right skills. If the goal of a class were to train the students to actually have the skills, this method seems pretty good and could probably become excellent with iteration. But at the time, it was quite tricky because in practice I had only about half the lecture time of a normal iteration of the course, and I was still expected to “cover” the same amount of material. So the curriculum sometimes felt rushed outside the quizzes, and I felt a kind of pressure to make the quizzes take less time in class somehow. (That said, this is an artifact of classes caring more about talking about ideas than about encouraging students to actually master a level of skill. This conundrum is secretly baked into most undergraduate math courses I’ve seen.)
Students end up practicing a different skill than the one they’ll need for the final exam. Time- and question-management mostly don’t appear in this style of doing quizzes. Most college students have plenty of practice with that anyway, so maybe this isn’t actually a problem. I just flat-out don’t remember whether my students had a problem with the final exam this way. (That would have been May 2012, which was also when CFAR was running its first workshop. I cared a lot more about the workshop than I did about those students’ grades.)
I’m sure there are others. This is off the top of my head, inspired from memory.
Does this answer your question?
Thank you very much, it does!
I think you answer is worth to be published as a separate post. It will be relevant for everyone who is teaching.