A Cheat-proof Exam

I was reading Matt’s “Cheating and Chinese Students” entry on Metanoiac, and it made me reminisce about how I dealt with cheating students during my tenure at ZUCC in Hangzhou. My own experiences might be helpful to some teachers out there, so I thought I’d share.

Throughout my 7 semesters of teaching spoken English at ZUCC, my class format never stopped evolving. I just want to share some of the major evolutions here and the reasoning behind the changes. At the root of the changes were always two questions: (1) Will this improve the students’ spoken English? and (2) How will I assess the students (i.e. grade them)?

My first semester I relied heavily on vocabulary. My reasoning was that it’s easiest to grade (by quizzes/tests), and if the students can incorporate the new words into their spoken English, they’ll become better speakers. I focused on common, useful idiomatic expressions and slang. Nothing tricky.

I quickly discovered this system was flawed because: (1) the focus was on learning vocabulary, not improving spoken English proficiency, (2) the students weren’t really learning to incorporate the new vocabulary into their spoken English, and (3) it was really hard to stop the rampant cheating on tests and quizzes. Another unwanted by-product of this system was lots of time-consuming paper grading.

I moved on to a skit-based class. Vocabulary was still important, but grading was based on the use of vocabulary and overall spoken English in performed skits instead of on paper. My students felt very awkward about performing at first, but soon got really into it and had a lot of fun.

Later I learned that the skit-based class only works well for certain groups of students. One of my subsequent classes was almost totally devoid of imagination, and their skits would inevitably be translations of some story they heard elsewhere, with the target vocabulary forcibly (and often painfully) inserted into the dialogue. Furthermore, I began to realize that although skits can be a great fun tool, they’re too fake to be the mainstay of a spoken English class.

I came to understand that as boring as they can sometimes be, discussions really should be the meat of a spoken English class because they’re the realest way to practice spoken English. Over several semesters I developed a discussion-centered class model which proved surprisingly successful, and worked even on somewhat large class sizes. [I plan to put this complete model online in the near future.] Vocabulary was reduced to the role of a tool — as it should be — and was provided to help students through the discussion rather than to drive the activity. Vocabulary was kept to a minimum to promote talking. Grading in discussions was based on preparation and participation.

Finally, I return to my point for this entire post. Following my discussion-based model, I think my exam format was pretty much cheat-proof. I decided that in spoken English class, all tests and quizzes should be spoken. The final exam consisted of small-group discussions (2-3 students each) in which the students discuss amongst themselves one of the topics that we had discussed as a class that semester. They knew all the topics; there were no secrets to obtain from other students. The topic for each group was determined by random, and each group got 5 minutes to mentally prepare (no notes allowed, and nowhere to hide them), followed by 5 minutes to talk. All I had to do was listen to the testing group and make notes or guiding questions while keeping an eye on the preparing group in the back of the class. A new group would be moving in every 5 minutes, so a 40 student class could finish in about an hour and ten minutes, in theory. In practice it wouldn’t exceed an hour and thirty minutes.

The advantages of the system are as follows:

• The students have no way to cheat.
• Students who have regularly attended classes all semester are at a definite advantage.
• The material on the exam has all been covered in class before, so it’s unquestionably legit material (as opposed to some teachers’ finals, which may have nothing to do with the semester’s content).
• The teacher can give adequate attention to each student’s spoken English performance.
• The exam can easily be finished in a normal university exam time allotment.
• There are no papers to grade. When a class’s exam is over, the teacher should already have all the exam grades for that class.

There are some issues, though:

• Classes should have less than 50 students.
• A significant portion of the semester must be spent teaching students how to have a discussion (no joke). This exam format works well as a result of an entire semester of the discussion-based class, and is not expected to work as well independently.
• This will not work on students with only low level English (but should be doable for intermediate).
• Some students feel “cheated” by their foreign teacher if a large proportion of the semester is devoted to the students’ discussion and they don’t get to hear the foreign teacher talk as much as they are used to.
• Some students can’t understand that language proficiency is a skill rather than a form of knowledge, and, as such, must be practiced. That this practice may or may not involve much new material is hard for many to accept.

In conclusion, I’d like to say that the system I developed worked well for me in a certain part of China at a certain school for a certain class at a certain point in time. I don’t pretend to have all the answers. I just hope that some of the ideas I present here might help other teachers with their classes.

Related Link: Sinosplice’s Teaching in China: A Guide for the Uninitiated.