Our Center for Excellent Teaching without Money sends out a weekly reading with a commentary by one of the supposed experts at the center (people who have never actually taught college courses, of course). This weeks was on the superiority of multiple guess exams, and how it's possible to write really good "distractors" and how you only need three elements in your multiple guess exams, not four or five.
Okay, so I know in big courses instructors feel the need to use multiple guess exams. And, having done quite poorly in my intro zoology course in college on the multiple guess exams, I know they can be really, really hard.
But "distractors." A quick look at a the OED will show that "distract" comes from the Latin past participle "distrahĕre." And a quick look at an on-line Latin dictionary tells me that "trahĕre" means "to pull." So, to distract is to pull apart.
More commonly, though, one distracts in order to move a subject's attention away from where it was, especially if where it was bothers the distractor. So, you distract a toddler's attention from the breakable object they want to play with by waving a toy in front of them. That's benign. Or you bump someone so a third person can pick their pocket. That's less benign.
The thing is, if we're asking students to trust us, then the idea of trying to mislead them, to in a way, trick them, is not what we should be doing. But it's inherent in the way that researchers who study multiple guess type testing think about what they're doing.
I'd never heard the term "distractor" for the wrong answers on a multiple guess exam before, and so I'd never really thought about the term. And as someone who's never written or given a multiple guess exam, I've never thought much about them in that way, either. But now I'm thinking about them, and I'm resenting the zoology class's tests. (Most of my exams weren't multiple guess in college, though I had lectures of 400 people at times. Mostly, I think, some poor grad student or students did the grading.)
This is why I stopped giving the multiple choice exams. In an essay exam, the student and I feel more like we're on the same side -- looking for things they got right.
ReplyDeleteI'm in history.
ReplyDeleteI can't do anything but multiple guess exams in Humanities. I'm against these tests, but when I have no grading help for 200 tests, and I'm supposed to get them graded in a single weekend, I don't feel like I've got much choice. (I don't test students in my other literature classes. They write a lot instead, including in-class writing "pop quizzes.") Anyway, my Hum colleagues think my tests are too easy, but I don't think it's fair to use tons of distractors. The test averages are usually about 79-81%. I can live with that.
ReplyDelete200 tests in a weekend is insane.
ReplyDeleteI like recognize-the-right-answer quizzes for some things (especially if the students are at a stage, as mine often are, where they're more likely to be able to recognize the right answer than to spontaneously produce it. A well-written one can serve as a quick check-in--yes, I noticed that; no, I missed that. And they can also serve as prompts to thought--All of the following could be aspects of X *except*...As with any well-written assessment, you can learn something from them. But teaching something definitely needs to be the goal. I never try to trick them, and I'm always receptive to feedback about the assessments.
ReplyDelete--meansomething (I JUST commented successfully using my Wordpress ID on another blogspot blog, but no luck here today!)
The multiple choice test I hated the most was Constitutional Law Final in Law School. Each of the 160 questions had choices A through H- and answers like A and C and E or A, B, G, and H etc. None were ever just "A" or "B" and the small differences made it difficult beyond all justification. Essay tests aren't perfect, because of the subjectivity and inherent bias within any grader, but it least you feel you have the opportunity to express and develop what you know.
ReplyDeleteI've heard the term before, when doing freelance test-writing for Big Test Companies of Which You Have Heard but Which I'm Not Allowed to Name. I think it's fairly standard.
ReplyDeleteBut it does raise some ethical issues, all the more so because there's some research suggesting that saying "x if false" may serve to cement x as a fact even more firmly in people's brains.
Also -- why, oh, why does anyone hire people who have not taught, or have taught very little (and usually mostly/only in some sort of grad-level education program) to tell people how to teach? I'm lucky that my institution mostly does not do this, but, since we now have graduate degrees in "higher education," I fear we may be heading in that direction.