In yesterday’s Underground Back Channel post, I wrote:

Because students are engaging in forbidden activity these conversations are underground. Well, the smart ones. Some are having these conversations on Twitter where one party of the conversation is not private and anyone (like a nosy DBA like myself) can see it. If they are used to quasi-cheating, does real cheating become easier? That might explain much of what I see.

I decided to write more about this.

Back in July, I ran across posts on Twitter where a student claimed another named (Twitter account) classmate provided the answers to a test. The moral was: “Students who cheat together will not repeat together.” A few days later the student bragged this activity happened the previous term as well. Then the student asked the classmate whether the next test was taken, presumably so they could cheat again. We handed the information over to the institution. I have not heard whether disciplinary action was taken against the students.

At the time, I attributed the students writing about the cheating similar to hubris in the Congressman Weiner scandal. It may also point to a lack of understanding about privacy online. Hard to know what was running through their heads at the time other than what they are willing to say or write which will be biased.

The NYU Prof Vows Never to Probe Cheating Again story hit the week after the initial detection. Pursuing the cheaters hurt his evaluations. The advice of plagiarism detection tools is to use them for teaching students what is correct behavior not as punitive evidence.

But then I ran across Classroom Ethics 101. The student experiment points to students being very willing to take an early look at test questions and answers (69%) even when given a notice it might be cheating (41%). Ariely does not think the actual cheating was very much as on the final exam few made 90% scores. The experiment is similar to the UCF cheating scandal where 1/3 of the 600 students were thought to have cheated.

Occasionally we, those running the LMS, get a request to look at the Mail tool messages or Chat conversations between students in a class to see who are trying to share answers. So far none have. The professors fear the students are using private communication tools to cheat. Maybe they are smart enough to use Facebook where we do not have the data.

Cheating is happening. The activity is underground except where students make a mistake. Professors have plenty of tools to help them detect it. Maybe as analytics become more widespread, they will be used to identify more cheating, though perhaps it should come from deans or academic affairs or student judiciary not the professor.