Why weren't teachers/schools more realistic with predicted grades?

[nervousnelly8]

Genuinely looking to understand how the A Level results seem to have gone so wrong. I don't work in education and DS is too young to be impacted, so hoping that those more informed might have some input!

Everywhere I see, people are calling for teacher assessed grades to be used. But if this happened, wouldn't the results be way out of line with history, rendering them useless as a form of comparison with other year groups? I understand that the model that has been used appears flawed when considering individuals, but does aggregate performance not also matter? Why wouldn't schools and teachers have been sensible in their predictions relative to previous cohorts so that their assessment could be used reliably?

Not really an AIBU I suppose, I'm sure IABU for seemingly missing the point completely!

Do you think we should have guessed which ones might have missed a question or which might have drawn the short straw with regards to examiners who haven't set foot in a classroom in years? We've been careful with grading but we can only predict what we know they're capable of.

But it must be obvious which centres had over predicted-the one where no one achieved less than an A comes to mind! Ofqual have had the predictions for 5 months. Why not moderate those rather than a blanket application of a flawed algorithm.

Elton did get downgraded. Head not happy
twitter.com/lewis_goodall/status/1295105129331466242?s=21

The bulk of schools are entirely fair and higher group results just represent the more accurate reflection of the ability of the nervous / misreaders. The dodgy taking a chance ones are the small independents, IME, where years of moderating show these centres are inept or Tryjng g to game the system hoping moderators don’t download grade all the way. I would imagine they have submitted way above with actual results too.

@HoneysuckIejasmine

I think 2020 results are always going to be viewed as inaccurate compared to other years, so they should have just allowed CAGs. Employers will always use a pinch of salt when looking at 2020 cohort. Remove the uni place caps to allow them to take all they offered to. Clearing allowed as usual as places will open up when offers are declined.

Then we can all move on.

If you're uni bound, a lot of employers won't look at your A levels anyway, once you have a degree.

The uni cap isn't just funding though: do the universities have the physical capacity if more students than expected turn up?

Note that if not as many as expected for the grades they needed the university has the option to drop the entry requirement so where they haven't done that then courses are probably full (I saw on the news that Oxford has done so)

Predicted grades predict what a student could get not necessarily what they will get. So if I have four students capable of an A they could all get As or there could be other factors, tricky question, personal circumstances, off day, dodgy marking, that means they miss that grade. It’s not some kind of mad cheating grade inflation.

You may have 5 students all acheving a solid 6 in class work, homework and mocks throughout. You will have evidence they are working at a 6 and therefore CAG is 6.

On the day it's possible to get any number of combinations of actual results

1 or 2 may have pulled it out the bag and got a 7.

1 or 2 may have been ill, panicked, performed badly and got a 5.

There's is no way of knowing which of these students it would be.

Likewise your solid 7's could get 6's or your solid 5's could get 6's.

This system already assumed a student would perform at their level and not do better on the day.
It's totally unrealistic to just let allow an algorithm to predict which students wouldn't have done well and have them marked lower than their actual ability because they didn't take the exam.

Yep. Another ‘let’s blame the teachers’ without fully understanding what has gone on. Thanks OP

Could you explain how, for example, a student who sat two full exam papers in a school hall under exam conditions at the beginning of March, submitted a full piece of coursework before lockdown and the schools shut worth 20% of the course. Achieved roughly a C grade (using the 2019 grade boundaries) in both. All topic tests since the start of Year 12, plus end of last summer exams and homework were all never below a C and plenty B and A grades at times. Gained a 6 at GCSE in the same subject. Was predicted a grade C by me (over inflation my arse) and OCR have given him a U grade.
Can you explain that please?

(I can. A victim of the ranking system and a large cohort of pupils which means he was up against another 7 students all working at a C grade. By the way, none of them, even the one ranked 1, was awarded a C by the exam board. Luckily I have the mock exams in hard copy and the coursework to present to the exam board but even then, they keep moving the goal posts)

Please Mumsnet. Stop blanket blaming the teachers on this one. Please.

I believe I got my grades spot on. Hours of work with other dept members got us the grades for GCSEs. We believe we have it right for our cohort this year based on what modules and NEAs they had already sat.

@HoneysuckIejasmine

Also this is something that could have been tackled over the summer. If a school hasn't over inflated (by comparison of whole school data Vs CAGs), then accept the CAGs. If there's a big difference, ask the school to evidence it.

As it is, this algorithm was only 60% accurate. That's appalling.

Yes this is exactly what they should have done. It's also difficult to compare this year's cohort to a previous one because they were the first to take the full 1-9 GCSEs weren't they so shouldn't be compared to a group who took different exams. As for the 60% accuracy, I would say that was pure luck! They didn't use the CAGs in the algorithm so I don't think they should be claiming accuracy. It was bound to match some of the time.

Teachers are human. Some schools will have put in hours looking at cags, some will dash them off on the back of an envelope whike witching East Enders. It has never mattered before, so no surprise that some won't have taken much effort on it.
Some people are natural optimists. "Well he might gets a B" and others want to see evidence.
Present the same piece if work to a room full of teachers and they won't give the same grade.
You also have, do you want to be the teacher who gives a candidate B, knowing that they need an A for their dream uni place. There's a 5% chance of them getting it, so why not?
Or the one who might just scrape a 4 in maths.

Then you will get some that will try it on. You have the example above if a teacher who said they were put under pressure to over predict. And some will do it because they want to look good.

There’s some articles you can read about it.... stating it isn’t really the teachers faults. It’s also based upon location etc the teachers put in their grade but because of the government’s algorithm based on the actual school’a performance it can cause lower grades. So students in a good grammar school in a better location over all may receive better grades than a high performing student in a crappy comprehensive. Which is very unfair.

I don't think anyone is blaming the teachers. Blaming a system where parents and children have their hopes raised by over-optimistic UCAS predictions, yes. Lots of the children interviewed in TV and radio saying they were predicted X, Y and Z are talking about UCAS predictions, which are notoriously unreliable.

The system this year clearly hasn't worked. I do remember the time (in decades past) where children were told that mock grades would be used if, for example, the exam papers were lost or they were in hospital, or similar. So perhaps a more formalised system of mocks under exam conditions (which both my DC did, in fact) should be in place. And UCAS grades could be taken from these too.

Predicted grades predict what a student could get not necessarily what they will get. So if I have four students capable of an A they could all get As or there could be other factors, tricky question, personal circumstances, off day, dodgy marking, that means they miss that grade. It’s not some kind of mad cheating grade inflation.

We were not very explicitly not to predict what they could get but what they were most likely to get, bearing in mind historical results and the ranking!

Very concerned for our students now!

I don't think anyone is blaming the teachers.

You have read the thread title, yes?

It's not the difference between the UCAS prediction ( everyone knows that these are aspirational ones) and the allocated grade that is the issue, it is the difference between the CAG and the allocated grade that is. In the majority of cases the CAG was completely ignored and just the rankings were used.

a high performing student in a crappy comprehensive.

It has to be an unusually high-performing student in a “crappy comprehensive” and I agree it’s very unfair.

Such situations should be dealt with by exception - it would not be ok simply to accept all of that school’s CAGs if they were improbably optimistic.

It's totally unrealistic to just let allow an algorithm to predict which students wouldn't have done well

The algorithm didn’t do this. Teachers did, via their rankings.

All in all I think it's very easy for the government and ofqual to blame the teachers and this is the crux of the matter.
They had nearly 5 months to sort this out, if they had really seen over predictions from schools they should and could have tackled that matter with individual schools but that would mean alot more hard work on their behalf so let's downgrade massively and unfairly and then blame the teachers🙄.The world lives to bash teachers.

That’s nonsense about Eton by the way. They also had grade changes down.

With the best will in the world, this just isn't something that can be predicted.

How do you account for the students who lose it on the day? How do you decide who on the grade boundary will pull it out of the bag and who won't? How do you account for the questions suiting some students and not others? It's impossible.

Children need to do actual exams to be evaluated. That's the bottom line.

@LaurieMarlow

With the best will in the world, this just isn't something that can be predicted.

How do you account for the students who lose it on the day? How do you decide who on the grade boundary will pull it out of the bag and who won't? How do you account for the questions suiting some students and not others? It's impossible.

Children need to do actual exams to be evaluated. That's the bottom line.

There’s an argument that this year’s ranking exercise was fairer. At least this way, results were based on pupils’ performance over 18 months, and teachers’ knowledge of their ability, rather than on who had an off day or forgot to turn over the last page of the exam.

@HipTightOnions

It's totally unrealistic to just let allow an algorithm to predict which students wouldn't have done well

The algorithm didn’t do this. Teachers did, via their rankings.

Not true. The algorithm was written by people and programmers who chose to manipulate those rankings in a particular way that involves a lot more. If it was as you say, I wouldn’t have had the outcomes in my school that I did.

We were! Ours were totally in line with our three year average. Still hammered. Our results do not make sense with the information provided by Ofqual. If anything we were harsh.

A further problem is that teachers tend to be positive for university applications and think they can get ABB when realistically they are BCC candidates on a good day.

Though students know their UCAS predicted they will not necessarily know what was submitted by the school on the CAG.

@GravityFalls is right - schools have different types of predictions.