Saturday, March 29, 2014

Same test, different versions

So yesterday, I was talking to a guy who used to work at my university.  He was trash-talking the students saying they were all cheaters, etc, etc.   And yes, students in Korean universities will cheat on tests simply because there isn't the social expectation that they won't.  Nor is there any real punishment if they get caught.  Not really their fault, it's just the system that they're in.  Anyway, he asked me if my students cheated.  I said, "No, never."  His response was that I just simply didn't notice.  Yes, perhaps one or two have slipped by me, but I really don't think so because I do the following things:

1. For conversation/speaking classes, I will only do 1-1 type interview tests.  It's just me and the student.  I will give them a list of 8-10 possible topics or questions and then randomly choose three or four.  How you could cheat when you are having a 1-1 conversation with the teacher, including follow-up questions that are "off-list" is beyond me. 

2. For written tests, I will ALWAYS make 3-4 versions.  It takes like 10 minutes to do in Google Docs.  Just cut and paste and then change the numbers.  For a few questions, I'll change some of the vocab too. For example: "Write 2 sentences comparing a desktop computer (living at home) (Seoul) and a smartphone (living in the dormitory) (Busan)."  I very clearly write "TEST VERSION A(B) (C)," at the top.

Then, I make a big show about handing out the papers from various folders depending on seating arrangements and tell the students that it truly is futile to even try to cheat. 

3. I NEVER do multiple choice questions.  It's simply too easy to cheat (by sending answers to friends via text) or by looking at someone else's paper.  It's almost impossible to catch cheaters this way as well.  But, by expecting students to actually write out a sentence, it's much harder for them to copy it and it's much easier to catch them (ie: the same 10 sentences, complete with the same mistakes on 2 student's papers means that there certainly is a problem of some kind). 

4. For my advanced writing class, 50% of their final grade is the 2 tests that they physically do in class.  10% is attendance and then the other 40%: I guess they could cheat, but the majority of their grade is cheat-proof so the curve should reflect who can actually write and who can't.  I plan on giving the students a list of 30 or so possible essay topics.  To start the test, I will hand out of a piece of paper with their essay topic on it, so each student will be writing something different from their classmate.  How could you possibly cheat on this?

No comments: