An experiment on cheating

 My students have good morals — most of the time. When it comes to cheating on quizzes, though, their standards are different from mine.

I give a weekly quiz that includes a little reading comprehension, a few tone words, and a couple literary terms. Although not a huge part of the curriculum, the quizzes are important in making sure that students have breadth of knowledge for the upcoming AP test.

At the beginning of the year, I told my students I trusted them not to cheat on the quizzes. In fact, we went over the answers immediately, and students corrected their own quizzes. This process worked until I noticed that the students’ scores seemed high — perhaps too high.

I am not the type of teacher who distrusts his students. After all, I believe in them. But to test my suspicion, the following week, I had students grade their peers’ quizzes. The first time, grades plummeted, but after that, the scores quickly rose again.

Then there was last week. Two students who sit next to each other had almost identical (wrong) answers. I approached them, and both students denied they’d cheated.

So today, I decided to do an experiment. I made four versions of the same quiz. (Teachers do this all the time. I’ve done it before, too.) In my opinion, it’s a complete waste of a teacher’s time, but in the school game — which includes cheating — it’s sometimes a necessary evil.

(One may make the rebuttal that the quiz itself — my decision to assess traditionally — is setting up the students to cheat. I don’t buy that argument.)

What were the results of my experiment? Well, quiz scores declined drastically, but not as much as I’d feared. My analysis is that many students have likely been cheating on 1-2 questions per week (and that some have been copying the entire quiz). My hunch is that students chose to cheat because it’s easier and takes less time than studying.

Now that my little experiment on cheating is over, the question is what I do next. Do I tell them what happened? Do we talk about cheating? Or do I continue making four versions?

What do you think?