By Gordon Rugg
Every year, huge numbers of new students start university, and are surprised to discover that their timetable is very much a work in progress (and sometimes, a work of fiction). Every year, understandably, huge numbers of new students react to this discovery by wondering why universities crammed with alleged geniuses can’t sort out something as simple as a timetable. It’s not an encouraging start. This article is about the reasons for this state of affairs.
The main reason is that timetabling actually isn’t simple. In reality, it’s hideously complex. The timetable for a single university has to handle thousands of students, hundreds of modules, hundreds of academic staff, and hundreds of rooms. Very few of those students want lectures first thing in the morning or last thing in the afternoon, or on a Monday or Friday, so some slots are much more in demand than others.
Reconciling all of these issues is a huge, messy problem, but it could in principle be resolved by using smart software; some universities already use cutting-edge software that can perform impressively well, if other things are equal.
Unfortunately, the big spanner in the works is that other things usually aren’t equal. Here’s a classic example of why timetables are often fluid until well after the first week.
Suppose that you’re in a practical computing class of twenty students, and that you’re timetabled in a room that has twenty computers. That’s nice and straightforward.
Now suppose that one more student decides to sign up for the module. This means that you can no longer fit everyone into the room; even apart from the issue of not having enough machines, there’s the more significant issue that Health and Safety regulations take a hard line on room capacity, and quite right too.
So what happens? What happens is that either the session needs to be moved to a bigger room, or if that isn’t feasible, we now need to schedule two separate sessions in the same room. Either way, we’re looking at a timetable change. That change will probably have all sorts of ripple effects, and it’s a change that nobody could have predicted in advance.
If it was just a case of one or two students changing their modules, we could probably work round it by always booking rooms with some spare capacity. In practice, the numbers on some modules can be wildly unpredictable, for all sorts of reasons.
So, in conclusion, timetables are usually works in progress, and usually the reasons are because of the way the world is put together, rather than a reflection on the competence of the university.
That’s maybe not the most encouraging message ever, so here’s a nice picture of some contented kittens as a soothing final note. The next article in this series will be about something positive and encouraging.
There’s more about university life in my books with Marian Petre:
The Unwritten Rules of PhD Research:
Rugg & Petre, A Gentle Guide to Research Methods:
There are a lot of other articles on this site about academic life and education, including topics that students often have trouble with, such as the differences between academic writing and other types of writing. They’re tagged under “education” and/or “craft skills”. We hope you’ll find them useful; if you’d like to see more on some particular topic, let us know via the comments section, and we’ll see what we can do.
Yoiu have written clearly and concisely about some of the issues. I believe these issues are easily resolvable.
For me the issue is one of minmising costs with attempts to maximise efficiency in the use of resources while at the same time trying to give the impression of a customer focused, flexible and accomodating system.
It’s about trying to have ones cake and eat it.
I think of it as the Tesco’s issue.
Interesting post, especially for timetablers and accountants.
There’s been some very interesting work on using Genetic Algorithms to handle the scheduling and resource allocation issues; some universities use this approach to find the closest feasible match between what everyone wants and what is possible. I get the impression, though, that this approach isn’t widely known outside Artificial Intelligence research.
The main problem I encounter with timetabling is not knowing how many students there will be in a given module. Without that uncertainty, timetabling would be a lot more tractable.
Do you think a solution to that would be to put hard limits on the amount of students in a module? It seems to me to be an optimisation problem and so the genetic algorithms approach seems like a great idea but would only be applicable if you had some sort of hard limit to how many possible students in a class.
Pingback: Life at Uni: After uni | hyde and rugg
Pingback: Life at uni, revisited | hyde and rugg