So, How Bond is That?
Another outworld-ish experience I had at Bond University was when attending examiner meetings.
I never really figured out what these meetings at the end of each teaching trimester were supposed to achieve. But neither did any of my colleagues. The meetings were held after all (centrally scheduled) exams and everyone happily showed up. So I did too.
What we did there - under the competent supervision of the dean of the day - is to look at every students' mark in every single course.
What we looked for were anomalies, such as a student failing every single course except one. And when we found such an instance, then someone in the room had to state that "this is really a weak student" and we all had to nod in agreement. And so it was decided that this student was supposed to have failed in that single course as well. Basta.
And if a student had HD (high distinctions) in all courses except in one with a D (distinction), then we all agreed that "this is a really good student" and it was decreed to upgrade him (or her) to HD. Basta too.
Now imagine 15+ academics, all with a Ph.D., manually scouring over the printed lists trying to find more anomalies.
No, I do not want to hear that a Perl program could do this more effectively. Not now.
It appeared like an alien game. Maybe not overly attractive, but it had its strong moments. And you learned things about students you never expected to learn: "Since he has this new girl friend, he never shows up in class" or "Her mother's sister on the grandmother's side has a wart and because of that the student is so stressed". Important things only life itself could tell.
The other thing we looked for were students who got 84 points out of 100. Because with these marks you would get only a distinction (D) and with 85 a high distinction (HD).
Students might be keen to go for that missing point, and to discourage them the rule of thumb at Bond was to either upgrade the student to 85 preemptively, or to downgrade him by 2-3 points. Basta.
Now imagine 15+ academics, all with a positive IQ, scouring through the printed lists trying to find these corner cases.
No, I still do not want to hear about Perl.
After several hours even the most eager of us became tired of that game. So time for another one.
That one involved walking through every individual course to look for more anomalies: too many HDs, too few HDs, too many Ds, too few Ds, too many fails, etc. Isolated, or in comparison with other courses running in parallel or with the same having been run in the past.
With student numbers around 100 per course this would even make sense. But what is the significance of a distribution with 20 students? Or with 10? Or with 5? Or with 3? Or with 1? Hmmm.
So imagine 15+ academics, all with an IT degree, taking part in a concerted Rorschach inkplot test assessing Excel-produced barcharts with average sample sizes around 10.
And when finally an anomaly was perceived and the immensely competent dean asked for the possible cause, then you are supposed to say "Yeah, that was a particularily (bad or good) class". Basta.
It is not really that difficult to teach at Bond.
And now you can tell me about Perl. And teaching ethics.