In one of the courses I taught last semester, the exam has asked pretty much the same questions every year for the last five years. This doesn’t exactly reflect well upon those doing the teaching, and this year (my second of covering these lectures) I made a deliberate decision to do something a bit different with the questions from my part of the course. The exam is divided into two sections: a ‘numerical’ part focussed on data manipulation and calculating things, and an ‘essay’ part. For the former, instead of asking the usual question about triple junction stability and migration, I gave them a question where they had to assess the quality of some paleomagnetic data and use them to calculate when a terrane (a highly allochthonous one, no less) collided with a larger continent.
I realised at the time that this alteration could potentially act as a useful survey of the study practices of our undergraduates. The question I set was actually pretty easy – if you’d bothered to revise the part of the course about paleomagnetic field tests and calculation of paleolatitudes (I’d even, after some debate, given them the relevant equation, and therefore wondered if it wasn’t a little too easy). However, students in thrall to the art of ‘question spotting’, who only revised the small bit of the course about triple junctions, might be in a little bit of trouble.
Today I received the scripts from the exam; no-one had answered my numerical question. Not one student.
Although that means I have less marking to do, part of me can’t help but find that a little bit depressing. The weird thing is, plenty of people answered one or both of the essay questions I set – both of which asked about parts of the course which previous years' questions hadn't, and at least one of which requires some knowledge of paleomagnetic techniques to answer properly. Maybe then, this just indicates a degree of mental inflexibility (‘It’s not a triple junction question! Panic!’) rather than damagingly limited revision practices. I suppose I’ll find out when I read the essays…
Update: OK, after a bit of investigation, it seems there are a number of factors at play here:
- Of the other two questions in the relevant section, one, on seismic moment and earthquake hazards, has been mixing and matching the same three subsections for at least the last four or five years; the other, on elastic bending of a plate in response to seamount loading, must have been seriously highlighted by the lecturer because I was asked to help with exactly the same question a week before the exams.
- I have some testimony to the effect that the students pretty much assumed that ‘my’ question would be of the same type as it had been in previous years (despite the fact that someone else had taught the course for most of those years), and revised (rather selectively) with that assumption in mind.
- If I’m scrupulously honest with myself, the way the question was worded might have made it seem that it was a lot more work than it actually was: not every step was explicitly spelt out, and the style (“This is the scenario. This is your data. Calculate stuff” style, rather than “Calculate x. Calculate y. Calcuate z.”) was perhaps a little intimidating. I’m open to the possibility I was also asking to much for a 45 minute question, but sadly I have no data to test that.
So, faced with two other questions they had prepared for, and one which they hadn’t and might have looked like a lot of work, the students opted to shun me. I’m not sure anyone comes out of this particularly well, to be honest. In my own defence, I have a fair number of essay questions to mark, so obviously I wasn’t being a fully bastard examiner from hell.