Exams! What are they good for?
Are you expecting the answer “Absolutely nothing”? That’s what a lot of people would say, I know. Since we’re in GCSE and A-level season at present, it’s something on many of our minds at present.
We all know what’s wrong with exams. They’re affected by how you feel on the day and the luck of the draw as regards the question paper. They can be prone to be largely a memory test. Some people get in a terrible state and don’t do themselves justice, whilst others do far better than they deserve because they are naturally good at exam preparation (that’d be me, for one). They certainly encourage teaching to the test, rather than to learn the subject (as evidenced by the number of examinees who remember at most a tiny fraction of what they “learnt” a week after the paper).
But… what are the alternatives?
Coursework? No – who knows who’s actually done the work for it, or how much help was received. And as I’ve said elsewhere, the controlled assessment version of that actually makes it worse, not better, because cheating still goes on, but the exam boards pretend it doesn’t.
Continuous teacher assessment? No. Sure, as a teacher I tend to feel I have a far better idea of the kids’ abilities from teaching them throughout the year than can be demonstrated in an exam. But this is way too subjective – we can all over- or under-estimate kids, based on all sorts of factors from personality clashes to how pleasant or cooperative or articulate or tidy they are, or our perception of how hard they work. And as long as schools are assessed on pupils’ results, there would be a huge pressure on teachers to assess a child who could barely write their name as a budding genius.
So should we just do away with them altogether, and let employers and/or universities use their own entrance tests instead? After all, they’re always saying that GCSEs and A-levels aren’t “fit for purpose”. That does almost sound appealing… but what are the chances of all those individual institutions producing assessments that are valid and reliable, if organisations like our exam boards, with their years of experience, mountains of data and substantial resources, still can’t perfect it? And think of the opportunities for abuse of the system, and the mountains of accusations of “bias” if we went down those lines.
So what about changing them to “Open Book” exams – where you could take in books, resources etc? In many ways I quite like that idea – you wouldn’t be able to do well if you didn’t know your stuff (you can’t teach yourself the subject during the exam, and you haven’t time to look everything up). There are downsides though. Questions that don’t rely at all on memory tend to be per se much tougher since all the assessment is of higher-level skills (I did one open book exam in my Masters’, and I tell you it was evil!). The exams would be much harder to write too – since candidates could take in old questions and solutions with them as part of their “resources”, you couldn’t use old questions or anything similar (and even if you limited the resources that could be taken in to a textbook, authors would soon include old questions in the guise of “worked examples”). Perhaps the most serious problem is that such papers would really favour those who could think on their feet, even more than the current system.
So I think overall I’m sympathetic to the view that exams are the worst way of assessing learning… except for all the others that have been tried!