Regular readers know I'm a True Believer in test corrections. What better way to cement a student's understanding than to have him write out an explanation for each problem he personally missed on a test.
I discovered a new difficulty with test corrections in a general-level class today, one that I had never really noticed before. My class spent an inordinate amount of time asking me and each other what was wrong with their original response; as a result, what should have been simple corrections took them for friggin' ever.
Interestingly, I have never had this issue on multiple choice corrections. My students, whatever the level, just justify the answers quickly, without trouble. See, on multiple choice questions, the scantron simply marks the answer right or wrong; there's no ambiguity about what part of the original response isn't right.
I've not had this issue with AP-level students on their free response corrections, either. That's because I've always asked very detailed, targeted additional questions on the corrections. They focus on answering my new question... and that question more often than not leads directly to the right approach.
But today was the first day I have done in-class corrections on a general-level all-open-response test. The problems on this test are too simple to make my AP approach of targeted additional questions to be useful. As they faced an incorrect test and a blank page, I had a parade of students asking "did you mark off for just the wrong units, or was the answer wrong, too?" "Is it my diagram that's incorrect, or the explanation?" "Would this have counted had I said..."
Note that no one was arguing that they should have earned more points, or debating my ability to grade consistently. We're well over that issue. No, these were earnest questions, motivated by the desire to get the correction right, the desire not to make the same mistake twice.
Nevertheless, the interminable focus on "why did I get it wrong" kept too many students from focusing on what I wanted them to -- namely, just doing the problem right. Even as I kept asking students not to try to figure out their mistakes, but to start each question from scratch, they couldn't let go: "But, just tell me real quick so I know, was this part of the answer right or wrong?" Several students spent 20 minutes trying uselessly to find the source of a mistake, but then finished in only 3 minutes when I forcibly removed their original problem from them.
And therein lies the solution. Next time I want students to correct their open-response questions, I'm not returning the original test at first. Instead, they'll get a blank test with the questions they need to redo circled. Once they do their corrections, THEN they can have their original test back. Not only does this new approach solve for the "why did I get it wrong?" issue, but it provides even more incentive to work steadily without distraction. After all, they can't see their test grade until they get the corrections right. :-)
I'll let you know how this goes when I try it next trimester some time.