Highlights from today’s Education Select Committee: Questions Answered & Ones Still Needing Answers
11th September 2012
More important were the variability of quality in her answers and the questions that MPs left unasked. Below I review some of the highlights.
Some questions asked of Ofqual and the answers:
Given that TES Evidence shows Edexcel were asked to move grade boundaries, did you also influence other exam boards? < Glenys Stacey explained that AQA and OCR were both in line with expectations, WJEC and Edexcel were asked to reconsider or bring back more evidence. Both WJEC and Edexcel then decided to follow the regulator and change their boundaries.
Were the grade boundaries too harsh in June? No, grades were correct in June but too lenient in January.
Why did it take so long to realise that grades were too lenient in January? Stacey suggested this was due to the low number of candidates in January (14,000) when compared to the much richer picture presented in June once many more candidates were entered.
Schools have been accused of ‘over-marking’, where is the evidence for this? After Stacey arguing first that there wasn’t any specific evidence, then saying that English is very subjective and it is very difficult to ever mark, Amanda Spielman then pointed out that schools had predicted and marked in a way that suggested a12% rise on the previous cohort, this seemed highly unlikely. Spielman was adamant that this sort of change cannot happen from one year to the next – even though a number of MPs tried to point out that having years be outside the average in terms of progress was probably quite likely. Spielman then compared children’s exam results to glaciers: they both move slowly over time.
Why has the Welsh Minister come to a different decision to you and asked for re-grading? Spielman said this was because Wales are in a different political situation where their exam results are not going up every year and “they find this difficult to accept” (Ouch!)
Should Ofqual be investigating itself? Stacey was very clear that Ofqual are the industry regulator of exam boards. Their job is to regulate what the exam boards are doing and ensure consistency. That is precisely what they did when intervening with Edexcel and WJEC and they did so within very tight rules and expectations. However, if there needs to be a bigger investigation she is happy to be part of it.
Would she release all phone calls, emails, memos between Ofqual, the Department and Special Advisors? Yes.
Should there be a re-grading? No.
Questions that still need to be asked and/or answered:
If English cannot be accurately graded when there are 14,000 entries does that mean Latin GCSE – which has less than 10,000 entries – also has inaccurate grades?
At one point Glenys Stacey said that achievement jumped remarkably this year in Economics GCSE but they ‘understood why’. Could we know why? What evidence would you need to see to ‘justify’ a large cohort jump and were the exam boards and schools given the opportunity to present this evidence for the English GCSE?
If grade boundaries are going to change wildly from one cohort to the next, on what should teachers base their predicted grades? (this was a major theme in the Headteacher’s witness session but was unfortunately never addressed with Ofqual)
And finally, why are grades only allowed to rise by 1% between KS2 and KS4? What is the basis of this assumption?
I have not been aware of grade boundaries being linked to KS2 attainment. If they are linked we need to know how and why. It seems there is much less to play for in demonstrating a rise in attainment if there is this degree of fixing, before we start. If there is this degree of fixing there is little point in arguing that standards are too low because (say) ‘only half of 16 year olds get 5 good GCSEs including English and maths’… as that is how the system is set up to be….
We need to see the complete flow of information between OFQUAL and the DFE/Gove to establish the extent of political interference on OFQUAL.
This “fixing” of grade boundaries is applied in all of our exams, starting with KS2. All of the value added analyses have depended on this too.
Exam results generally follow a statistically “normal” distribution, and if they do not, they are fixed to make sure that they do. This is because almost every measure about humans also follows a similar distribution.
The most worrying thing about government policy is that although exam results are fixed to give this pattern, Mr Gove wants all schools to achieve results above average, (totally impossible if you follow this statistical argument). Futhermore, schools not able to achieve this impossibility will be forced to be taken over by Academy Chains and their headteachers fired.
Chris Weadon chief statistician at AQA did try to enlighten the world about some of these issues about exam results, but was silenced by the powers that be. Politics and truth are not good bedfellows!
It is a pity that the teaching profession does not have any mechanism to expose these issues in an apolitical way.
Letters leaked to TES show that the boundaries were pegged to KS2 attainment. The reason appears to be that the leaders at Ofqual feel this is a good way of showing what the cohort is capable of and there is little reason to believe that this capability changes (in terms of its normal distribution) between 11 and 16.
Letters are here: http://www.tes.co.uk/tesassets/images/resources/Ofqual_Edexcel_Documents_2012.pdf