Andrew Taylor, AQA Head of Curriculum for Maths, discusses grade boundaries for the November 2017 GCSE Maths exams.
I thought it would be useful to look at the grade boundaries for the November series, because there have been comments in the press and on social media that some boundaries across exam boards seem quite low – and are lower than June’s.
First, some background about the differences between June and November.
November boundaries are difficult to set for a couple of reasons. The cohort of students sitting is very different to the summer and we have to use a different approach to statistically predict the likely performance based on prior attainment.
Also, the cohort is much smaller so the predictions are less robust than in the summer, particularly at the Higher tier where we may be looking at a few hundred students rather than hundreds of thousands. All this becomes even trickier in the first sitting of a very different examination.
If the exam papers are consistent in demand with the summer, there is no reason why the boundaries should not also be similar to meet the very different expected outcomes for students and maintain standards.
However, if the papers prove to be very different in difficulty from the summer, then the predictions may suggest very different boundaries. This can make awarding a challenge, particularly at grades where the statistical evidence is less strong.
Fortunately, though, I am privileged to work with exam writers who have achieved remarkable consistency across the 42 specimen, practice and live papers they have written for the 9–1 GCSE and that consistency can be seen in the boundaries for November 2017.
So, how did we do compared with last summer?
The below table shows the boundaries for this November with difference from the summer in brackets.
At first glance they are unremarkable. Five have gone up, five down and one is unchanged. All the changes are in single figures and equate to no more than a mark or two on each paper, which is well within the variation we are used to. Further, the boundaries going up for the top grade and down at the bottom spreads the marks out a little more which is generally seen as a good thing.
So, why the concern?
In the summer, the concern was that the grade 4 and 5 Higher tier boundaries for all exam boards were low. As Ofqual has explained, half the marks available on the higher tier have to be targeted at grades 7, 8 and 9, so it’s to be expected that the boundaries for the lowest grades available on that tier will be fairly low. But those boundaries being lower in November across boards briefly re-ignited the debate.
As I described after the summer exams, the grade 4 and 5 boundaries are set to ensure that they represent similar performance at the boundaries across the two tiers. So, this November, common question performance for students achieving 127 marks on the Foundation tier was the same as students who achieved only 40 marks on the Higher tier. This suggests that Foundation papers were marginally easier for grade 4 and 5 students than in the summer, while the Higher papers were a little more difficult.
We set boundaries by using all the information and experience at our disposal to be fair to all candidates. If we hadn’t set a lower boundary for a 4, about 70 young people would not have received that grade when their performance on common questions showed that they deserved it, and that would have beeen unfair.
As always, our grade boundaries reflect our very best effort to be fair to all students and they represent a consistent and credible approach to the examination of mathematics at GCSE.