Friday, 9 May 2008

Class Sizes in UK universities

What has happened to class sizes in Russell Group universities? The need for national data.

Oxford Magazine - 2007

Bruce Charlton

In a decade my final-year class size has gone from around 16-24 to 84 and 123. First and second year classes are around 200 students. In other words, aside from a handful of tutorials and supervisions of dissertations or projects, it seems as if students now go through the whole degree in very large classes.

What I would like to know is whether this massive decline in teaching quality is typical of the top 50 (ie. roughly pre-1992) UK universities in general, and of the large Russell Group research universities in particular.

Anecdotally, the answer would seem to be yes, such increases in class size are typical. In the past, introductory lectures were big, but as students progressed groups became smaller. But the remarkable fact is that no one really knows what's going on, because information on university class sizes is not collected nationally – or if it is, it is not publicized or distributed.

In particular, although the national university "teaching inspectorate", the Quality Assurance Agency (QAA), examined a great deal of paperwork (a whole roomful of box files in the case of my department), and indirectly generated vastly more, it failed to measure class size.

Just think about that for a moment. The QAA neglected to measure the single most important, best understood, most widely-discussed measure of teaching quality: class size.

It is no mystery why class sizes have expanded. Over 25 years, funding per student has declined by more than half, and the average number of students per member of staff increased from about 8 to 18. In the face of long-term cuts, a decline in teaching quality was inevitable. Indeed, it was anticipated: the QAA was created in order to monitor and control this decline.

But is class size important? Of course it is! In the first place, the public regard class size as the single most significant measure of teaching quality. Every parent with a child at school knows their class size. Parents who pay for their children to attend private schools are often explicitly paying for smaller classes.

It is not just in schools that size matters. US universities publish class size statistics that are closely scrutinised by applicants. For instance US News gives data on percentage of classes with fewer than 20 and percentage of classes with 50 or more. Around 70 per cent of classes at top research universities such as Princeton currently have groups of fewer than 20, and less than 15 percent of classes with more than 50 students. The expensive and prestigious undergraduate liberal arts colleges offer not only about this proportion of small classes and an even smaller proportion of large classes, but guarantee that classes that are always taught by professors (rather than teaching assistants).

A way of measuring the importance of class size is to see what people are prepared to pay. In a small comparison of public and private universities in America, Peter Andras and I found that students at the private institutions paid on average 80 per cent more in tuition fees, for which they got 80 per cent more time in small classes.

Given the usefulness of a valid and objective measure of university teaching quality, and the overwhelming evidence of public demand for small classes, the case for publishing national data on university class sizes seems unanswerable. I would guess that class size data is already available within the central administration of many UK universities, because they record the number of students registered for each course for their own internal administrative purposes. It is just a matter of collecting and summarizing the information, and publishing it nationally.

However, I doubt that universities will publish class size data unless they are made to do so. University bosses probably feel too embarrassed to admit the real situation: nobody wants to be first above the parapet with shocking statistics. Alternatively, if or when the three thousand pound cap is taken off university fees, some universities with small classes may start to publish this data in order to justify charging higher fees, and eventually all universities may be forced to follow suit.

But why wasn’t the QAA interested in collecting and publishing data on UK university class sizes? I can not think of any good educational reason. It managed to spend well over £53m in data collection and auditing since being set-up, plus many times that amount in opportunity costs incurred by UK universities, but amazingly failed to provide a valid measure of teaching quality.

Incompetence and inefficiency on this scale would beggar belief if the QAA really was concerned with teaching quality - but of course it never has been.