Results fiasco is symptom of the lack of trust inbuilt in the way England’s schools system has been set up

In slides made available for a presentation by the exams regulator Ofqual last month, one set of statistics stood out.

A page entitled “The distribution of Centre Assessment Grades compared to last year’s grades” saw Ofqual showing what grades teachers had submitted for their students at A-level, AS Level and GCSE, in this unprecedented year when these grades were to given weight, in the absence of students taking exams.

Ofqual’s statistics showed that, if grades were to be awarded based on these figures, national results would have soared, with the proportion of candidates awarded A*s at A-level, for example, nearly doubling from 7.8 to 13.8 per cent; double-digit percentage point increases at grades A to B at A-level and dramatic rises, too at the key GCSE comparison points of grades 7 and 4.

Ofqual, therefore, reacted by using a statistical model to downgrade many results, pulling the overall national statistics more in line with those of previous years, as has been its defining mission under the “comparable outcomes” policy.

This has resulted in many very concerning tales in terms of the impact on individuals, with reports of students downgraded by Ofqual’s algorism to grade Us seemingly particularly staggering. And at first reading there seemed a strong claim that this new system might be institutionally biased in favour of private schools running subjects with small numbers of pupils, where teacher judgments were given more weight.

However, underlying all of this is a sense that this latest shambles is an indirect product of the pressure being placed on England’s education system by trying to use student results for too many purposes.

For Ofqual’s lack of trust in institutionally-submitted exams data, and the statistical modelling behind “comparable outcomes,” came about after years of steadily-increasing grades before this system was introduced.

That rising trend, in turn, was widely seen as a product of the results pressures on both schools, and on a government which had asked the public to judge it on pupil results. For average grades were simply expected to increase every year, making these past statistics a judgment not only on the students taking the exams, but on their teachers, on their schools, and – under New Labour at least – the politicians overseeing the entire system.

Results statistics have arguably been given even more weight at an institutional level since the Conservatives took power, as the success of schools grouped into multi-academy trusts – including high-stakes outcomes including taking over and losing schools – now also hinge on student “outcomes”.

This year, ironically, some of the institutional pressures on schools to raise results are not in place: the government has said league tables will not be produced this year, for example. And Ofsted inspections - which place high weight on pupil grades - are postponed until at least January. The DfE has said, in a comment made following its observation that league tables are suspended this year, that “no school or college will be judged on data based on exams and assessments from 2020,” which if taken at face value would mean Ofsted inspections would have to disregard these data, too.

Yet “comparable outcomes”, which essentially caps grades and ensures there can be no big rise statistically from year to year, has remained in operation*.

To put it another way: implicit in coverage I have read of the results controversy, including the reported reaction by the Labour leader, Keir Starmer, is trust that teachers’ judgments of their pupil grades may be correct. But Ofqual’s algorithm, as dominant at least in larger subjects, implies that the regulator does not trust those judgments.

This suggests, again, that England’s generally low-trust structure for assessing how our education system operates, which endures in Ofqual’s statistical modelling, struggles when confronted with what would have been a high-trust element this year, of allowing teachers and institutions to grade their own pupils.

Ministers, of course, have made all this so much worse by acting in such a chaotic, last-minute manner. But sadly, both chaos and the pressures put on the system by many years of complex, politically-driven and often poorly-thought-through reforms are recurring themes of recent English education policymaking.

*Notwithstanding the effect of mock exams now being a possible factor in the results appeals system, which I wrote about here following reports earlier this week. The overall statistical effect of this move, though, is still being debated.