This blog was kindly contributed by Dennis Sherwood who has for a long time been covering all things A-levels, Ofqual and exams for HEPI. You can follow him on Twitter here: @noookophile.
The A-level and AS results will be announced next week on Tuesday 10 August and the GCSE results on Thursday 12 August. Judging from this past week’s newspapers, we know what to expect. The Guardian’s front page had the headline ‘Grade inflation forces ministers to pay for extra medical school places’, followed by this first paragraph:
The government has been forced to fund hundreds of extra places on medical and dental courses at universities in England, as rampant grade inflation unleashed by cancelling exams has meant unexpectedly high numbers of students qualified.
Given that the results will be made public on Tuesday, it is, I think, a safe assumption that the relevant Ministers know what the results will be a few days in advance. So ‘rampant grade inflation’ is unlikely to be mere speculation, but rather yet another example of the Government trailing a future truth. Just how ‘rampant’ grade inflation will be will soon be evident: my hunch is that for A-levels, grades as awarded will be very close to UCAS predictions; for GCSE, there is no UCAS-like comparator, but since I think that few teachers would wish to ‘award’ grade 3 when a grade 4 is plausible – especially for those critical subjects of Maths and English – that suggests (many) fewer grade 3s and more grade 4s, with knock-on effects for higher grades.
‘No grade inflation’ has, of course, been the key policy since around 2010, so if this summer’s results have even more higher grades than last year – which itself was exceptional – then it might appear that there will be a considerable amount of egg on the face of the exam regulator Ofqual.
Ofqual, however, alongside the Department for Education, have remarkably egg-repelling faces and the entire emphasis of this year’s process had been to deflect those eggs very much towards teachers. This year’s three letter acronym TAG – Teacher Assessed Grade – immediately identifies the bestower of grades as the teacher, and this year’s appeals process has the teacher as the first port of call.
My conclusion is that this year’s ‘rampant grade inflation’ will be blamed squarely on those teachers. This is a grave injustice, for teachers, for the most part, have worked very hard and have acted with integrity. Their task, however, was impossible and for the second successive year Ofqual have designed a singularly flawed process. Oh, what a lost opportunity! If only some grain of common sense, let alone wisdom, had prevailed last year (see, for example, ‘Alternative 2, Trust the teachers’ here); if only that had been a platform for something even better this year … if only …
I fear, however, that by attributing the blame for this year’s grade inflation on teachers, Ofqual and the Department will use this as a pretext to return not just to the status quo before the virus, but to something harsher. ‘We gave teachers the chance to show that they could be trusted, to deliver good grades. And for two years in a row, they failed. In the interests of the country, we have to take back control…’. Oh dear.
The return to exams-as-usual has already been declared, and – recognising the need to prevent comparison of 2020 and 2021 results with those in the future – some ideas for a ‘reset’ have already been trailed, so getting that oh-so-important grade inflation under control. One idea is to introduce numeric A-level grades to replace the current lettering (A*, A, B…) and having a greater number of numeric grades will make valid historic comparisons more difficult; there is also talk of introducing a new top grade, 10, at GCSE.
One consequence of having a greater number of grades is to make the average grade width narrower. Which brings to mind these words, to be found on page 21 of Ofqual’s landmark report Marking Consistency Metrics, published in November 2016:
Thus, the wider the grade boundary locations, the greater the probability of candidates receiving the definitive grade. This is a very important point: the design of an assessment might be as important as marking consistency in securing the ‘true’ grade for candidates.
This is of more relevance the other way around: the narrower the grade widths, the lower the probability of candidates receiving the ‘definitive’ grade – where the ‘definitive’ grade is the grade that the script would receive if marked by a senior examiner (or, as page 20 of that report says, the ‘true’ grade).
Ofqual’s former Chief Regulator, Dame Glenys Stacey, has already acknowledged that grades are ‘reliable to one grade either way’, so by introducing more, necessarily narrower, grades, this will make grades that are already reliable only to one grade either way even less reliable. What use is that?
Messing around with grade names and grade widths might enable Ofqual to reset the benchmark for grade inflation, but surely the fundamental unreliability of grades is far, far more important. Yet Ofqual show no sign of wishing to do anything about it – it does not feature in their corporate plan, and when the new Chief Regulator, Dr Jo Saxton, who takes up her appointment on 17 September, was quizzed about this by the Select Committee on 6 July, her awareness of this key issue appeared to be slight at best.
Grades that are reliable to one grade either way are worse than useless, for the fact that a certificate shows, say, AAB implies that those grades are reliable, and that they really are AAB. Anyone that doesn’t happen to know that AAB really means ‘any set of grades from A*A*A to BBC, but no one knows which’ is likely to be misled; even worse, someone who does know, but fails to take this unreliability into consideration, could be inflicting a grave injustice on the certificate holder.
Which leads me back to Friday’s newspapers, and the headline on the front page of the Telegraph: ‘Universities set entrance exams amid A-level chaos’, with this first paragraph:
Universities are turning to entrance exams because they can no longer rely on A-levels to find the brightest students amid a “tsunami” of top grades.
That’s all about a university’s confidence in relying on school exam grades. And any university that lacks this confidence has to find other ways to select their intake. Yes, contextual admissions are important, but does that imply that exam grades are totally ignored? Another possibility, as this article suggests, is that more universities introduce their own entrance exams. Which is one approach, over which that university has full control – but at a cost of time, effort and money – time, effort and money that would not be required if school exams could be relied on.
Other than the individual students themselves, higher education collectively is by far the largest and most important ‘user’ of A-level grades – and so by far the largest and most important voice that Ofqual should listen to. If the individual institutions within higher education speak with different voices, Ofqual can choose which particular voice to hear. But if higher education can speak with one voice, and speak loudly too, then Ofqual is obliged to listen.
This year, there are no alternatives to the TAGs, and so there will be no repeat of last year’s fiasco, in which the results of the algorithm were (after public pressure) abandoned to be replaced by the Centre Assessed Grades (CAGs). But much of the muddle of these last two years has been underpinned by the fatal flaw that has been present with Ofqual’s grades for the last 10 years – the fact that even the grades resulting from the so-called ‘gold standard’ exams are reliable to one grade either way.
When exams are reinstated, this will still be the case. Even if some form of Post-Qualification Admissions (PQA) is introduced, if the actual grades on which that process relies are still reliable to one grade either way, then the only result will be to substitute the unreliability of UCAS predicted grades by the 1-grade-in-every-4-is-wrong unreliability of the ‘awarded’ grades. What use is that?
School exam grades that are reliable and trustworthy are a ‘public good’, benefiting higher education, employers, schools, students and most importantly, the students. And they are easy to deliver. Ofqual, however, has refused to change their policies to enable that to happen.
The mess of the last two years has thrown a spotlight into the hitherto murky corner of the fact that exam grades are reliable to one grade either way. And the mess of the last two years has highlighted, across higher education, the importance of grades that are indeed reliable and trustworthy.
So, this is an opportunity for the higher education sector to ride to the rescue. This is your opportunity to speak with one, loud, voice, and say:
Ofqual, you have a statutory duty to a “secure a reliable indication of knowledge, skills and understanding” as represented by exam grades. Being reliable to one grade either way is not reliable enough. Exam grades must be reliable, full stop. Do it.