The Fallout from A Levels, What Next?
Published on: Monday, August 17, 2020
As the turmoil surrounding last week’s publication of A level results shows no sign of dying down, I thought it might be useful for me to put my own thoughts on the matter down in writing. What follows should be seen against a backdrop of deep disappointment and puzzlement here at King's: far too many of our own students have been harshly treated by the exam boards.
Let’s start at the beginning. This cohort of students was not able to sit A level examinations in the summer. Quite rightly, the government was keen that they should nonetheless receive grades for their subjects in order for them to be able to move on. The difficult part was deciding how best to generate those grades. Clearly there had to be significant school input. Clearly no system was ever going to be perfect (every year we are surprised by some of our students’ results – no system can predict how a candidate will perform with 100% accuracy). And, to my mind, clearly there had to be some sort of external moderation.
This last point is important. If schools were simply asked to come up with the final grades themselves there would have been massive grade inflation. That is inevitable: we all want the best for our students and given the chance we would always err on the optimistic side. So some system of making sure that grades remained roughly in line with previous years was necessary. Otherwise this year’s A level results would have been devalued and would always be treated with suspicion by future employers.
The idea, then, was to produce an algorithm which would tweak each school’s results so that they more or less fell in line with the typical pattern of grades for each department in each school. For very small departments, where there was not enough historical data to be statistically secure, the grades would not be changed. Much has been made of this in the press – but in fact it is, by definition, not statistically all that relevant. The great majority of grades in independent schools were still subject to the algorithmic moderation process.
At King’s we took the process of generating grades and rank orders for our pupils very seriously indeed. At the same time as our teachers were working their socks off to sustain an excellent remote learning programme, they were asked to spend a great deal of time and thought sifting through all the data we held on each GCSE and A level candidate. Reams of data were sent to each department: the results of baseline testing, previous public exams, previous internal exams, significant pieces of work produced etc etc. They were also asked to use their judgement about the character and habits of each individual pupil – were they the sort of person who pulled it out of the bag at the last minute? The idea was to predict how well each candidate would have done had they sat the examinations. Those predicted grades and those rankings were then sent through to a small group of senior King’s staff for internal moderation. That group compared the pattern of the predicted grades in each subject with the historical patterns of the past three years. In several cases departments were asked to adjust their predictions to bring them more into line with past averages. We arrived at what we were confident was a sensible, honest, measured set of grades for our candidates, which properly reflected the abilities and hard work of this particular cohort, and which we then submitted formally to the boards.
Last year our students set a new record for A level results. This, we believe, is part of a generally upward trend for the school – we have made good progress in improving the effectiveness of our teaching and learning, and we continue to refine and develop our methodology in preparing candidates for exams. It was also due to the make-up of that year group. We are not a large school, and inevitably some year groups are academically stronger than others. A measure of the integrity of our centre assessed grades: Last year we achieved 77% A* to B grades. The grades we submitted this year, after that long and thoughtful process, would have given us 71% A* to B. We did not take the mickey, we did not abuse the system.
But, as I say, some form of moderation was inevitable. We trusted Ofqual to get that right.
But clearly they have not. When our final results came out last week there were some, many, jaw-droppingly ridiculous down-grades. One girl had been predicted a C grade in chemistry and ended up with a U. So the board was essentially saying that, in an examination she was not able to take, she failed. There is no chance on earth that she would have failed that exam. She was a solid C candidate. On a very bad day she might have dropped to a D. On very good day she might have raised her game to a B or even better. She could not have “failed”. And there are ridiculous downgrades of that sort wherever you look. 45% of the carefully-considered grades we submitted were reduced by the algorithm. The final pattern for the school, including the now 61% A* to B rate, is wholly out of kilter with what our pupils have achieved before.
So what has happened? I think it is simple: the algorithm stinks. They got it wrong. The idea might have been sound, but the execution was a disaster. An algorithm is a recipe made up to try to do a particular task. They just got the recipe wrong. And it wouldn’t have taken much checking of the results for individual schools to see that it was wrong, and to make adjustments to the algorithm. Ofqual have had five months to get this right. Particularly galling is the fact that most exam boards have reduced their usual exam fees by very little, if at all. We are having to pass onto parents a significant cost to pay for the privilege of being royally stitched up.
What happens next? The problem here is time. There are candidates who need their grades to be sensibly raised in order to secure their places at university in the next few days. At the same time we are about to face a, presumably, similar storm when GCSE results come out. And at the same time were are working flat out to prepare the school for opening in September under tight and complex Covid rules. Bandwidth is an issue, and would be an issue even if we knew what the appeals process looked like. But as of now we still don’t know. Neither the government nor Ofqual are saying anything. It’s a mess.
Frankly, it does seem now that the only way forward is for Centre Assessed Grades to be accepted. At this stage the serious and very real drawback of grade inflation is, I believe, outweighed by the anger and mistrust of students and schools, and there really is no time left for a meaningful appeals process. Which school has the capacity to manage a complex appeals operation while at the same time coping with the (inevitable) fallout from the GCSE results and preparing for a post-Covid reopening in a few days’ time? If my students were simply awarded the grades we had given them I would sleep easily, knowing that we, at King’s, had done as fair a job as possible under the circumstances and that the grades were a much better reflection of their abilities than those they have been awarded by the boards. I can’t answer for other centres, and it may be that some have pushed the boundaries unrealistically, but for my own students I believe that would be a sensible outcome.
The original plan made sense and most independent schools supported it. The philosophy was sound, but the implementation was bungled, to the detriment of hundreds of thousands of angry and disappointed young men and women. I don’t think, with the national mood as it is now, that any attempt to sort the mess out while maintaining some sort of downward moderation of inflated grades will either be accepted or be completed in time. That leaves the grades predicted by the schools as the only practical way forward.
Update
I posted this blog online only 45 minutes before the announcement that Ofqual were, after all, going to accept the CAGs! We are pleased with this outcome, though it is not ideal. The effect of grade inflation will have an impact on this cohort, whose results will, I suspect, be looked at with suspicion in future. It will be interesting to see how universities manage what is now a very tricky situation: on the basis of last week’s grades places have been secured or denied and most courses are now full. Universities are in an impossible position: having turned down a student on the basis of last week’s grades, what do they do when that same student now finds she has the offer grades after all?
For students taking a gap year and hoping to reapply to university, things also look difficult. Firstly, their grades will be treated with some suspicion and, secondly, universities will have over-recruited this year and possibly have less space next year. But, as I have said to my son, bright, hardworking, interested students will always be in demand. Let’s hope that’s true!