As I’ve mentioned before, this year I’ve served as a IA Examiner. As part of my duties, I showed up at the Guildhall last Friday to examine Paper 3, gown and all. The role of Examiners in a physical examination is to answer questions and issue clarifications concerning examination content as needed. Well, that is the idea at least...
You see, in subjects like English or Physics, it is a reasonable expectation that any member of the faculty would be knowledgeable about all the topics represented in a given paper (and, in fact, it’s common for each paper to be set by a single person). Computer Science is a little different, in that probably no single member of the faculty could get a passing mark on any one of our papers, much less set the entire paper. I don’t know a thing about machine learning, and likewise our excellent faculty in machine learning would not be to muddle their way through even the easiest supervision sheet in Semantics. I wish that weren’t the case, for all our sakes, but that is how Computer Science works. We are not a single field; we are the refugees of a dozen other fields, thrown together and then bound by a shared ethos whose roots lie deep in the history of the Computer Laboratory.
Long story short, there is no hope of any Computer Science Examiner being able to answer almost any question posed by students during the exam—except in the vanishingly remote chance that it happens to be concerning a question they have themselves set. So our policy is instead that Setters shall sit by their phones during the pertinent examination and await a phone call from the Examiner. That would work very well, if Setters would actually wait by their phones rather than setting their phones to send all calls direct to voicemail, taking meetings during the examination period, etc. As a responsible Setter, I actually had my wife call my phone to make sure that it would ring, and then stayed by it for the duration of the Paper 2 examination on which I had set two questions. This was not difficult, but it took some forethought and a sense of responsibility.
Even more bafflingly, Examiners are not required to sit in the examination room for more than 20 minutes, which means that the invigilators must phone the Examiner when questions arise, creating a long-tailed game of Telephone in which important queries go unanswered for the longest time even in the best case. I decided to buck this bizarrely maladaptive practice, and just stayed throughout the entire exam to ensure that emerging matters could be dealt with swiftly (I actually got some writing done!). Indeed, if I were to take my duties as Examiner remotely seriously, then no time would have been saved by me leaving the examination room anyway (since I must stay by my phone the entire time and certainly could not take meetings or get deep into other work). I would like to see other Examiners take up this practice.
Examinations happen only once per year, and they may seem of little consequence in the scheme of things to someone who sees all aspects of the life of the University as “necessary evils” ancillary to Research. But the careful conduct of examinations is immensely consequential to students, and we should therefore take our role in the process with commensurate seriousness.