From David Maughan Brown in York: No better way?

August 13th

I hope regular readers, if there are such, will bear with me if I ride another hobby-horse off in pretty much the same direction as I did in my last entry, and write about exams again.  My excuse would be that having spent 12 years at high school and university being expected to write them, and my entire 43 year working life at universities either setting and marking them or dealing with the copious fall-out from them, I’ve spent a fair amount of time thinking about them.

So, half-listening to the Today programme yesterday, I sat up and took notice when our callow Secretary of State for Education, Gavin Williamson, who always manages to look like an adolescent rabbit in the headlights when interviewed on TV, articulated his (and by implication the government’s) key principle of educational faith when questioned by Nick Robinson:  “There is no better way of doing assessment than exams.”   In case we hadn’t been listening properly the first time, he reassured us later in the interview, not once but twice in successive sentences, that we had indeed heard him correctly: “No system we put in place is going to be as good as exams.  Every system we put in place is going to be second best to that.”  So lots of equal seconds, then, but no question whatever about what gets the gold medal.

However imperfect exams are, it has to be admitted that they will in many instances be better than assessing students via an algorithm that can somehow manage to increase the proportion of pupils achieving A* and A grades at private schools by twice as much as that at comprehensives.  Just as racist computers don’t programme themselves, so algorithms aren’t self-generating.  Our Prime Minister has declared the system to be ‘robust’, which his minders should know by now would fatally undermine any lingering confidence any half-intelligent observer might have had in it.  But anyone who has ever been involved in education knows that three-hour examinations, which is what Williamson is talking about, are a very much less than perfect way of assessing much beyond a student’s capability at writing three hour exams. And three-hour exams tend not to be one of the frequently encountered hazards of working life.  In an examination in the humanities, for example, if what you are looking for is a student with a disposition not prone to nervousness and the ability to spew large quantities of verbiage, much of it memorised, onto paper in a wholly arbitrary three hours, then examinations are your bag.  Quite what that ability is supposed to be useful for is not entirely clear.

Many parents would be very happy to let Williamson know that some children are very much better suited to the peculiarly artificial exigencies of sit-down examinations than others.   My own siblings are a case in point.   I was relatively good at exams because I enjoyed the challenge, had worked out how to work the system, particularly with regard to what was likely to come up in an exam, and in those days had a half-decent memory.   As an undergraduate I devoted about as much time to honing my bridge skills as to covering the extensive lists of set books (which I didn’t), and good results were in no way an accurate reflection of my knowledge of the curriculum as a whole.  My sister, who is no less intelligent and capable than I am and was vastly more diligent, has a brain that functions in a different way from mine, never came anywhere close to completing any exam paper, and consequently came out with consistently lower results, which didn’t stop her from becoming a successful computer programmer.  One of my brothers is dyslexic, was at school in an era when teachers had no idea how to identify or respond to dyslexia and assumed, wholly incorrectly, that he just wasn’t very bright.  He was petrified by exams and, unsurprisingly, didn’t get good results, which didn’t stop him from having a successful career as a primary school teacher.  And our Secretary of State thinks there is no better way of doing assessment than exams.

On top of their artificiality, the stress and anxiety they occasion, and the questions they raise about what they are supposed to be testing, exams also have unintended and pernicious consequences in encouraging both a narrowing of the curriculum and ‘teaching to the test’ (see nesta link below).  Continuous assessment would be a vastly ‘better way of doing assessment than exams’ if one could be sure precisely who it is one is assessing.  The only thing sit-down exams have going for them, pace their full-throated endorsement by the likes of Williamson, is that with proper security you can, at least, be quite sure who is responsible for the answers.   In that single respect they are the least-worst of all the many alternatives.  But, as I suggested in my last entry, the current A-level shambles is forcing people to confront exam and assessment-related issues in a way they haven’t had to before, and there may be hope for a revisiting of continuous assessment on the horizon.   It has been suggested that developments in Artificial Intelligence may eventually make sit-down exams obsolete (https://www.nesta.org.uk/feature/ten-predictions-2019/beginning-end-exams/). One can only hope so.  In the meantime our Prime Minister and our Secretary of State for Education, among others, might be well advised to start praying that continuous assessment via AI is never applied to their performance.