DfE Data Cruncher predicts number of students who will get straight 9s

[noblegiraffe]

His guess is.... 2

Not 2%,

2 kids in the whole country will get all 9s in their GCSEs.

So that's the new challenge for the MN boaster.

Ofqual reckon 0 kids will manage it. They clearly haven't met any MNetters' kids.

twitter.com/timleunig/status/845699774754017280

no more than a 1 -2 grade error

I'm pretty sure parents would be unimpressed with an A grade kid being given Cs, or a B grade kid being given A*s and being told 'oh, that's expected, we can't be that accurate'.

I think that it is fine to say 'My DD is at a superselective school, so I know that her grades will be in the top few % and her teachers say the same, though they may be out by a grade or two'.

It is not fine to say that the same predictions can or should be able to be made by other teachers in other types of school - so to imply that noble is wrong, or worse than your DD's teachers, when she says that she cannot make such predictions, is clearly not fine. She is just working in very different circumstances, in which making predictions is HUGELY more difficult.

So a superselective, typically, will draw its children from a wide area, and can be pretty sure that they are in the top 10% of ability compared even to a wider peer group. However, a comprehensive draws from a much smaller area, and can have much less confidence in how the spread of the ability of its children reflects the situation nationwide. So for example, I can say that i know the (primary) school I teach in has an ability profile skewed towards the more able, towards the nationwide picture - but exactly how much, and exactly where the 'nationwide average;' is within our cohort is pretty much impossible given the small sample size.

So for a teacher like noble, where in her, say, 150-250 pupil cohort the 'true pass mark' or even 'the true top 7% nationally' lies is well-nigh impossible, and certainly very greatly more difficult than for a school that can state with certainty that virtually all its pupils are in the top 10% nationwide.

Bobo, I think we established earlier that the French system doesn't work quite as you have represented it - of course the marks a child gets don't matter if actually what matters is how a school writes a report?

why noble should decry the predictions of 9s as lies and nonsense

I didn't. I said it was probably fair to say that a kid who gets 100% in mock exams will get a 9.

This isn't the same as grading internal tests and this isn't the same as saying that teachers can tell where the 9 cut-off is and where the kids who will be an 8 or a 7 are.

Equally, though, noble, even if goodbye's DD's teachers (in a superselective) all stick a pin in a chart and say 'I predict a 9' they are statistically much more likely to be right than they would be in a full spectrum comprehensive! It's not because they are 'better at precting', it;'s just that they have a very small spectrum of grades to choose randomly from!

I don't know what 'precting' is! Predicting....

I'm not talking about the French system. I'm talking about the fact that far too much time is wasted by parents/teachers/students when marking schemes are complex and changing. There is a lot to be said for simplicity.

Just to be clear - there are hardly any words on French reports.

can'tkeepawayforever but you see I haven't said what you've said I've said!

noble objected strongly (very strongly) to any predictions for the new exams whatsoever, in any context, and said any teacher, in any context, who is making predictions is a liar and talking nonsense. I defended the giving of prediction in our school's context. noble has repeatedly objected to that.

Bobo, so what do the reports from schools that go to universities say, and how do universities judge them against one another?

Now who is misrepresenting what people are saying, goodbye?

Goodbye,

I didn't notice you saying 'although grading work is relatively easy in my DD's school, I can see that it would absolutely be invalid in 93%+ of other schools, and thus it would be an utterly unreasonable thing to do nationwide'? I would also say that even in your context, the absolute assurance of a 9 is obviously untrustworthy, though 'in this context, we can be fairly sure that's 8is-9ish' is possible.

Bobo,

I can set a test than every single one of my class get 20/20 on. Or that all get below 10 out of 20. Or I can set tests that I think are of equal difficulty, but in the event find than in one, the mean score is 13 and in another it is 17.

So unless the scores out of whatever are set in some kind of context, absolute scores are meaningless.

If 10, 20 or 100 is the national grading currency, schools are held to account.

noble I'm not misrepresenting - check back.

can'tkeepawayforever I said what I said. If you think I should have phrased it differently - not my problem.

noble how many times do I need to emphasize that no-one seems to be taking predictions as assurances. I certainly never have done. Over the years the school has always ringed predictions around with qualifications to do with commitment etc. Please stick to what people have said.

Yes Bobo I agree completely (again). Keep it simple.

The universities (and other HE) receive grades on reports, the (minimal) comments and an opinion on the chosen course of study from the student's Head.

can'tkeepawayforever you do notice that you and noble are directly at odds on the grading even at superselectives. Incidentally, the school is not grading internal tests only on a very range at the moment - there's a fairish spread. Nor are predictions 8/9, they're framed as single grades. You'd have to be pretty dim to think predictions = certainty. Boys especially are warned about this at school - always have been.

I rather like the idea of grading with emojis....

twitter.com/drrobmoody/status/846833384357838850

That should read on a very narrow range.

We have emoji grading from the English teacher

noble I'm not misrepresenting - check back.

I don't need to check back because I can see from your latest post that you are still conflating giving GCSE predictions (something that is necessary for Y11) with giving gcse grades for performance on internal tests (totally unnecessary and pretty much nonsense).

Bobo, so the 10, 20 and 100 are exactly the same as 1-9, or A* to G - they have (or should have) nationally agreed meanings and norms and are standardised?

Then - except for nomenclature, they are exactly the same as the current 9-point scale or the old alphabetical scale - they are not some mythically different and better method of grading. Grades on French reports = grades on English exam results, as far as I understand? Are they externally moderated, or only internally-marked by schools?

Noble, the confusion is interesting. I mean, when we had O-levels, and also had internal grading of pieces of work A-, B+, C- at school, although they both used letters of the alphabet, NOBODY linked the two. Nobody expected there to be a link between an A- piece of work and a low A in O-level, and there wasn't.

So it COULD be that Goodbye's school is using a 1-9 scale on interbal tests completely co-incidentally, and intending no link whatever to be made to GCSEs - and like the old A- = a good piece of work, that would be fine. It is the fact that goodbye and other parents have assumed that there IS a link between the two grades that is creating the issue.

The point I am trying to make is not about the French system. It's about the benefits of having a longstanding, transparent, universal marking system that everyone understands and doesn't waste time fretting about!

It's about the benefits of having a longstanding, transparent, universal marking system that everyone understands and doesn't waste time fretting about!

Please tell that to our various Education Ministers who cannot help but meddle as soon as they're appointed.

Indeed, titchy. People are very resistant to change in France but, tbh, the education system is delightfully traditional in many respects. Whole swathes of "issues" don't exist here because no minister has invented them yet!