To think that the government did this on purpose

[therhubarbbrothers]

The A level results fiasco seemed to penalise bright children from disadvantaged schools. Aibu to think the government knew that this would happen but saw it as acceptable ?

They were not “40% inflated”. They were around 12% above last year... and that doesn’t necessarily mean inflated. As I’ve already pointed out, Ofqual say they can’t be sure students get the right grade even after their exams have all been marked because more than one grade might be a fair reflection of their performance. In my subject analysis suggests 42% of students are given the wrong grade in the exams each year.

The algorithm can’t tell you who has inflated and who hasn’t. Nobody can, because nobody outside of the individual school ever looked at any evidence of the performance of their students this year. The idea that I should be able to say with certainty who would have got a C or a B, two months before they were due to sit the exam, without a set of exam papers to base it on is laughable. 2 out of every 5 students I teach normally gets given the wrong grade by the exam board so that makes it tricky to predict. All this talk of teachers inflating grades shows a total lack of understanding of the system.

@therhubarbbrothers

The A level results fiasco seemed to penalise bright children from disadvantaged schools. Aibu to think the government knew that this would happen but saw it as acceptable ?

Reverse snobbery

That’s batshit, the scores were forty percent over inflated

The batshit bit is that you keep insisting that the scores were 40% over inflated, a claim that has absolutely no basis in reality.
Think about it - the Headteacher of Eton would not have written a strongly worded letter if something was not well wrong. He is only one of many, many senior professionals who protested vehemently.

@Bluntness100

What should have occured is teachers told to assess grades but keep it within a given percentage of the average of previous years Performance.

My group this year were better better than last years.

Why should I be trying to fit them in to a percentage that has nothing to do with them?

I half wondered if they did in fact do it on purpose to increase competition and student panic surrounding university places....to ensure every student snapped up uni places offered to them to minimise the number wanting to defer to next year when COVID impact on their freshers year will hopefully be less

Yes the DOE is volunteering in a desperate bid to get some brownie points before he's judged for his past ignorant comments.

the Duke of Edinburgh ??1

AQA have just sent an update through saying that they are all set and the results will be the official ones and on time.

sorry wrong thread

Nah it isn't deliberate. Their incompetence is entirely accidental.

Does anyone know why Ofqual won’t reveal the algorithm used. Can the government make them release it for public scrutiny since the work was paid for by taxpayers?

@AndromedaPerseus

Does anyone know why Ofqual won’t reveal the algorithm used. Can the government make them release it for public scrutiny since the work was paid for by taxpayers?

An IT expert and father tried to unpick what he could - and predicted problems to no interest in 2017

www.theguardian.com/education/2020/aug/14/punishment-by-statistics-the-father-who-foresaw-a-level-algorithm-flaws

Using what scant information Ofqual had made public on its methodology, and an understanding of statistics gained from his PhD in physics, he analysed the 2017-2019 A-level data for Matthew Arnold school.

@Bluntness100

Pease, I’m not sure I agree with you, because the point I am making is even with the drop the grades were still over inflated v previous years, that’s The point people are missing.

Even with the alogorithm moderated results those results were still over all much higher than ever before. So it is simoly a question of How over inflated. At no stage were the over all results down year on year or even equal to previous years. Both sets of results are over inflated.

So the point remains it is still a very small minority of students who were Penalised, who were the outliers, because the overall results were still inflated by the government.

And now we have a situation where the cags are being blamed, we are already seeing the comments on here about it,

The issue is an exam performance can’t be questioned, but anything else, when it’s not what a child wants or needs and is adrift from the initial prediction parents are kicking off. The only way to further stop the complaints is to not use either the cags or the algorithm and use the initial predictions and even then folks would complain they were wrong.

I don't always agree with you @Bluntness100 but with this comment I totally do. The problem with this debacle is the existing over inflation of grades PLUS the media hysteria fueled by parents, candidates and an absolutely ineffective DfE ministerial team. Who caved. Rather than acknowledge that no system will ever suit all individuals. I work in this field (competently). Not sure I want to for much longer.

@WhyNotMe40

I thought the FFT fed into the algorithm though - which takes into account postcode and so socioeconomic status?

I’ve now looked at how the algorithm works and it doesn’t use FFT nor any socioeconomic data.

Disadvantaged schools weren’t ‘targeted’ because of their disadvantage. They may have had features that meant that they were downgraded more such as poor historic attainment data but the algorithm didn’t know what type of school they were, just the SATs/GCSE results profile and the prior achievement at GCSE/A-level.

Thanks Noble. I admire your awesomeness 😀

Fundamentally there were two major problems :

The first was that the parameters given by the DOE to Ofqual prioritised a lack of grade inflation above all else, including fairness to the individual

The lack of prioritisation of the individual led to a situation aptly described by a poster on another thread , or maybe this one , as being like a dry cleaner who has given all the clothes back so is judged a success but actually he has given them all to the wrong people so Sally has Janets skirt and Janet has Steve’s trousers and so on. So his customers are livid but he’s happy.This is I think @Bluntness100 point that overall we shouldn’t worry if the wrong people got into medical school or to selective universities or were left to get a job instead because as long as the overall grade inflation is kept down that is a price worth paying to preserve the system.

The second issue, and in my opinion the real reason why they have caved, is quite simply that they had not put in place a robust , or even functioning , system of appeal. There was no system of redress for the individual and limited for the school. In fact until the ‘mocks’ announcement the only grounds for appeal were administrative error . This meant that a huge number of people were left feeling utterly powerless and Ofqual knew that they would simply not be able to process the appeals.

Now we are left with total chaos, but arguably we should consider as a society whether A levels should be a measure of achievement or a system to manage university admissions because it’s clear that when A levels represent the working grade as opposed to the exam grade there is a conflict. Also had universities been given the heads up in July they could have been adjusting admissions criteria etc , the fact that they have been kept completely in the dark has seriously compounded the problem .

I think there was a third, slightly lesser problem, which was that some pupils have confused CAGs with their UCAS predicted grades. It seems that there were students who were saying ‘I was predicted AAA and I’ve been given BBC’ who didn’t realise that their CAGs were actually BBB making it appear in the media as if the system was even more capricious than it actually was.

Noble that seems quite sad , but I think you are right plus a lot of parents especially those who are doing this for the first time have not understood the system. Also the media have confused predicted grades with CAG , I lost count of how many times I have shouted at the radio about this over the last week. There have definitely been times , one was on the BBC after the climb down where they were trying to make out a student had been disadvantaged but actually it turned out his CAG was too low anyway .

This has been exacerbated though by how many schools have still not released the CAG to pupils. Ds had his on Friday morning but some schools are apparently still clinging on for dear life to the grades so the pupils really don’t know where the problem lies.

@noblegiraffe You would hope that pupils understand that UCAS predicted grades are not necessarily what their teachers realistically expect them to achieve but certainly many parents etc will not have appreciated that. I can’t remember what the figure is but the number of pupils achieving their UCAS grades is very low each year.

Of course many pupils always get into one of their preferred choices even with dropped grades - as long as it is not one of the very competitive courses.

I feel so sorry for universities having to sort this out now with very little time.

As I’m sure others have pointed out, it would have been far better to iron out local & individual inconsistencies earlier on in the process. What a mess.

Dd was told the cag for one of her subjects before she left which May have been a bit naughty but it did mean that we knew immediately that she was a ‘grade out’

Then i got a bit panicky and have spent the last few days going ‘are you sure s/he said that? You didn’t misunderstand’

What would be funny is if her cag for the other subject meant that she had again ‘lost a grade’ but we didnt know (i doubt very much thats the caseL i think both her apparent cags are fair based on the work she has done this year if she isn’t doing an exam)

Rufus that CAG you were told in March may not even be the CAG that was sent off at the end of May.

In my school, teachers entered their assessed grades onto SIMs in March and after that they went through a moderation process by the HOD who had a view of the whole year group for that subject and after that they were scrutinised by SLT who had a view of all the subject results and did even more fiddling to fit with the historic data.

March CAGs and May CAGs are not the same thing, as a lot of work went on in the meantime!

Have a listen to More or Less on R4 - it's on now but should be available to listen to later - they had a whole segment on this and it explained very well how the system worked, and how some students were disadvantaged.

The over inflation was at its worst in those subjects taught almost exclusively by private schools - Latin and Classics saw a rise in A/As of over 10%*

Too simplistic without proper analysis - 30% of entries for Latin/Classics were from state schools (the Independent newspaper) - undoubtedly with small cohorts. Eton entered 38 students for Latin last year.

@noblegiraffe

Ofqual had nothing to do with Gavin Williamson’s announcement last Tuesday that students could pick the best of their CAG, their calculated grade or a valid mock result. You can tell that from the blatantly pissed off announcement they made once he said it, saying they’d need some time to make it workable.

That’s all on Gavin Williamson. He should resign for that bit of chaos at the least, even if you think he should have understood nothing about the algorithm.

However: teachers had the students in mind when assigning their grades. Ofqual had the stats in mind when assigning theirs. It was the DfE’s job, led by Williamson, to stand in between the two and make sure that the result was fair to both students and the stats. They appear to have made no effort to do this and when they realised their fuck-up, dropped the ball at the last minute and made things much worse.

Totally agree.

As Noblegiraffe says, the algorithm doesn't use FFT or any other socioeconomic data. The problem is that Ofqual went down a route that tried to get too precise and, as a result, was unfair in a number of ways. They should have simply looked at the average points score for the school and moderated the grades as necessary to bring this year broadly into line with what would be expected based on the previous performance of the school and the pupils - that is what I thought they were doing. This would have used both the predicted grades and the rank order to produce the final grades. Whilst that would still have meant some pupils getting a higher or lower grade than they would have achieved had the exams taken place, it would have been a lot fairer than what has actually happened.

The problem is that Ofqual tried to predict the actual grade profile that the school would achieve, i.e. how many pupils would achieve an A, how many an A and so on. They then ignored the predicted grades and used the order in which pupils had been ranked by the school to determine the grades awarded. So, for example, if the algorithm said your school would have two pupils with an A, the two pupils ranked top would get an A* and no-one else would, regardless of predicted grades. This was unfair in several ways.

It couldn't handle outliers. So, for example, if you were a high performing pupil, on course to achieve an A*, but you attended a poorly performing school, the algorithm might say that you couldn't possibly get higher than a B. Equally, at the other end of the scale, if the algorithm predicted that two pupils would get a U, the bottom two pupils would be awarded a U even if they were actually on course for a C.

They knew that their statistical model would break down if there were too few students. They dealt with that by simply awarding the predicted grades without any form of moderation when a school had a small number of students entering a particular subject. This favoured independent schools as they tend to have small classes. It also favoured students in less popular subjects - if you studied classical Greek, for example, you were likely to get your predicted grade regardless of the school you attended.

I do not understand why they didn't keep it simple and use the average points scores. That would have brought predicted grades fully into the process for everyone and would have been much fairer in my view.