This is my final post on the topic of exam statistics. Previously, I described my use of the mean, difficulty scores, and point-biserial correlation. This time: the dreaded Q score. (Just to clarify: I'm not describing the other "Q Score", which represents the public's familiarity with--and appeal of--a person, product, company, or television show. That's not dreaded at all.)
The dreaded Q score is not a statistic that I regularly receive with all my other exam stats. I have to put in a special request. It's extra work for the people over at TS&QS, which means there's an additional cost that must be paid by the Department of Psychology. TS&QS has to go into their database of exam results for my class and perform a statistical comparison between two (or more) given exams.
Here's where the dread comes in: Why would I want to statistically compare two (or more) students' exams? If I suspect them of cheating, that's why. Sometimes, the cheating is blatantly obvious. The cameras in the classrooms (you know about those, right?) may clearly show one person peering over at the exam of another. Other times, it's not so obvious. Why is that guy jittering in his seat, looking everywhere except at his exam? Maybe he's nervous, or has exam anxiety. Why is that girl acting squirrelly, shifting her eyes back and forth? Maybe she drank too much coffee, has caffeine overload, and now really, really has to pee. Whatever the case, the exam proctors will not interrupt any student taking the exam. Nope. We'll just let you do what you do. If that's cheating, so be it.
However, at the end of the exam, the answer sheets from any suspicious students are set aside. (Think you can fool us by not leaving at the same time, or handing in your exams to different proctors? Tsk. You don't know how many eyes are watching, do you?) Those answer sheets will be analyzed, and I will get the dreaded Q score. I don't want to say too much about how it works, so suffice it to say that it gives a probability that cheating has occurred, compared to chance. Maybe I'll write more about how I have to deal with cheating in another post, but for now, let's just say it involves a lot of dread.
Although I have caught several cheaters over the years, I'm glad I've never had to deal with anything like what happened in Professor Richard Quinn's class recently. Yeesh.
(Cartoon by Frank Cammuso. It's important to give credit where credit is due. Otherwise, it's like, um...cheating.)
Why aren't you studying?
The Exam Statistics: The Q Score
Subscribe to:
Post Comments (Atom)
Find It
Huh. You certainly gave me a behind-the-scenes look at how cheating is handled.
Whatever happened to the mice in your office?
Now, if you'll excuse me, I need to go find the cameras in the classroom.
Cameras - Do they just look at the video if something happens in the class? Like if there was vandalism or suspicions of cheating or something? I mean, there's no one constantly monitoring it when the class is empty...right?!
Cheating scandal - I read that he just took all the questions from a test bank, in which case it's not the students' fault. How were they to know he'd copy those exact questions? They probably just thought it would be a good study guide (that's what at least one student said anyway - http://www.myfoxorlando.com/dpp/news/orange_news/111110-UCF-cheater-talks). Now I am not sure about whether the students had to lie about who they are to the publisher to get the test bank questions, or if they are freely available to anyone. If it's the former then they are at least partially to blame, for sure. But still - using the test bank to make questions? Seems pretty lazy to me, no? And it was not right to punish the rest of the students who had nothing to do with this. As for threatening them with legal action - really?? I shake my head.
Also, some of the students made this video in response: http://www.youtube.com/watch?v=pJG7aCQtI8E&feature=player_embedded
I think the funniest thing about this is that after 21 years of teaching, he is surprised by the fact that some students cheat. Most people are not moral - welcome to the world, professor Quinn.
@Anastasia: I don't know much about the details of the video monitoring. When the (very expensive) LCD projectors were first installed about 10 years ago, they kept getting stolen, so cameras started appearing in classrooms as a security measure.
Test banks have strict usage requirements--publishers will not release them to anyone who is not an academic staff member. If anyone else gets a hold of them, they may be in violation of those terms, which could lead to a lawsuit. (Although I did find a test bank for one of my textbooks freely available in the library one year.)
I've never made up an exam entirely of provided-testbank questions, but that's my decision. It's a lot more work, but then if I don't ask questions about my lectures, what's the point?
Well, in some regards, I understand Professor Quinn: I assume the best of my students, that they're working hard, they're honest, they're not cheating. In my experience, when someone is cheating, it's not hard to catch them at it. (Of course, that could just mean I never catch the really clever cheaters...)
That's really nice. I usually just assume the worst of people...less disappointments this way.