Archive for August 14, 2020

The Great A-level Scandal

Posted in Education, Politics with tags , , on August 14, 2020 by telescoper

The full scale of the scandal of this year’s A-level results is now becoming clear and it is bad enough to bring down a Government. Unfortunately there are so many scandals surrounding the UK Government (e.g. corrupt procurement deals, collapsing economy, terrible Covid-19 mortality figures, fiddled Covid-19 testing statistics, not to mention Boris Johnson himself) that nobody seems to care about that one more probably won’t make much difference.

Yesterday Qfqual released its report on this year’s A-level results which reveals that in arriving at the final grades, the algorithm deployed was based on past performance of pupils at the candidate’s school. in many cases this has resulted in students being downgraded by several grades in a manner that is both arbitrary and cruel.

Update: I’ve just heard from a physics, in an institute in which I once worked, that a student with original grade A* in Physics A-level has been assigned a final grade E. Unbelievable.

That starting point of the Ofqual approach is indefensible. A student’s examination grade should be determined by the student’s own performance, not by the performance of previous generations of students who happened to go to the same school at some time in the past.

Not surprisingly, the Ofqual approach has benefited students who went to private schools and severely disadvantaged students at less privileged establishments. The rightwing media are justifying this on the grounds that teachers at some state schools have inflated their students’ estimated grades. The attitude is that working class kids can’t possibly deserve an A* so their teachers must have cheated! I can’t believe this bias is unintentional. The Tory message to the less privileged is that they need to know their place. You needn’t ask who is behind this deliberate demographic* profiling. It stinks of the unofficial Prime Minister Dominic Cummings

But even within its own flawed terms the Ofqual algorithm is garbage. Table E8 in the report shows that when applied to last year’s input data (mock exams and centre-based assessments), even in the best case subject (History) the prediction was only accurate for 67% of students; the figure falls to less than 50% for, e.g., Further Mathematics. When the Ofqual panel saw that they should have abandoned their algorithm immediately. The fact that they pursued it knowing how deeply problematic means that they are more interested in serving their political masters than the students whose prospects they have deliberately blighted.

In my view a system should be introduced that gives the student the benefit of the doubt. Grades should be awarded based on what the student has achieved. If that ends up being too generous to a few students then that’s surely better than the opposite? Whenever I’ve been involved in University examinations processes when emergency changes were required we have always implemented a `no detriment to the students’ policy. It’s the obvious fair thing to do.

Oh, and you might ask why universities don’t show some humanity and accept students whose grades have been reduced. The answer to that is simple. If they do, they will go down in the league tables. And for many senior managers that’s all that matters.

*which means, of course, (indirect) racial profiling too.