One of the things I have learned this semester (which is now officially over, having turned in my last batch of grades this morning) is the following lesson which I am convinced I must implement immediately: Group work has been playing far too great of a role in my student’s grades. From this point forward, assignments which could conceivably be done in groups — not just those that are designated for group work — will count for no more than 10-15% of the grade in my courses.
I like collaborative learning. I think, in fact, that working with other people on math can be not only a highly effective way of doing so but also carries with it a powerful pro-math socialization effect. The best personal friendships that I had during my college + grad school years were those that I formed with my classmates in my various math classes, as we struggled through material that, to us at the time, was really hard. Not only did those friends help me learn, I also associated good times and shared victories over math problems with learning math.
But here’s the deal: At the end of the day, the grade that an individual earns in a class, mine or anybody else’s, has to be an accurate reflection of that individual’s mastery of the material and that individual‘s ability to solve problems and think effectively. If were reasonably confident that group effort on problems was translating into individual mastery, I’d be perfectly willing to admit as much group work as students want. But the fact is that this has not been the case.
Case in point: In a recent course, I gave out some pretty difficult advanced problems and instructed students on the usual academic honesty procedures, which boil down to “collaborate if you want but not to the point where you’re no longer doing your own work”. I got back solutions which were eerily similar and all basically correct, and in many cases way out of character for the students handing them in. It was enough to make me suspect a breach of my academic honesty policy, but not enough to make a case. So I simply reproduced the exact same problem on a timed test. And guess what? Whereas before, nearly everybody had a really nice solution — the same really nice solution — this time only one or two people had an idea where to start or even how to correctly parse out the terminology in the problem.
And this has been happening all over, not just in that class — a sort of soft academic dishonesty that nominally stays within bounds. Students work together and hand in work that earns points but does not reflect their understanding of the material. I understand earning good grades is important, but equally important is my ability to identify problem areas and help students grow through them.
So I know what all the digital nativists say about how in the modern workplace, people work collaboratively and it’s a 19th century anachronism to give out timed tests and all that. But you know what? You can’t contribute to a group if you yourself have used the group to feign your own competence. So from here on out, the majority — if not all — of my assessments of students will be done in a timed setting, under conditions that I can set and monitor. For example, in calculus next semester, I’ll assign homework problems and let students work on it all they want in any size group they want. But the grade is going to come from timed quizzes, tests, a midterm, and a final. Some variation on that will also be in place for my two sophomore level courses as well. If you do group work properly, contributing where you can and really working to understand things where you can’t, then it will be no problem to do well on a quiz or test. If not, then the quiz or test will show that up as well.
If that makes me an anachronism, or unhip, or whatnot, then so be it. I’m tired of students not learning the material because they have easy workarounds for doing their own work, and one way or another they will get a good grade in the course if and only if they can show me that they know what they are doing.