It seems to be accepted that the downgrading of students grades is because of teachers giving over generous predictions. But is it?
First of all any teacher who consistently over predicts will be censured. But more important the algorithm is primarily based on the past results of schools, not on individual predicted grades.
There was a discussion about this algorithm on the BBC News at lunchtime and iirc the algorithm operates in 3 stages, first it looks at the past record of school in that subject and then it tries to match the range of grades this year, with the range of grades in the past
Only then do predictions come into play, but not as a grade, only as a ranking, so that if E was at the bottom of the range in previous years, the bottom grade this year has to also be an E
So as I understand it, a pupil could theoretically be set for an A grade, but if everyone else in their class was predicted an A*, they would have to get the E.
The algorithm does not consider current data, it only produces data that reflects the past
The presenter, Mark Mardell, said, "isn't that an inherent bias towards the status quo, you could say an algorithm says that the last two out of three prime ministers went to Eton so that should be what happens in the future"
This monumental injustice is not down to teachers, its down the data that this government chooses to look at.
(Frankly, it would probably be more accurate if teacher predictions were used instead of this algorithm. It would certainly be fairer.)