The GCSE results and the need for uncertainty
Most people in the UK will be aware of the recent GCSE results in England, Wales and Northern Ireland (but not in Scotland, where there is no such thing as the GCSE):
Results fell by a modest 0.4 percentage points across the board but there was intense consternation about a deeper drop in English results focused in particular schools. The share of entries graded at C or above fell by 1.5 percentage points in English year-on-year, from 65.4% to 63.9%. Results in maths and science have also fallen, against a backdrop of an explicit order from the exams regulator to curb grade inflation – and promises from politicians to increase rigour.
But the focus of teachers’ anger is on the shifting of the grade boundary for English, between candidates who took exams in winter and those who took papers in summer.
Robert Robson, principal of the Samuel Whitbread academy in Shefford, Bedfordshire, said: “According to our calculations if you did the foundation paper in English in January and got 43 marks you would have received a C grade, while this summer you would have to get 53 marks to get a C grade. The most significant effect is on the C/D borderline. We have 50 students who would usually have got a C that have got a D.”
Although it’s of course terribly unfair and upsetting to the students who feel they’ve been deprived of the results they thought were rightfully theirs, and although I do have some sympathy for the view that Gove should create a new exam (with a new name) rather than making the exams harder every year, I must say I can see the need for some uncertainty in the system.
The thing is that if the schools know exactly how many marks you need to get a C grade, and if the papers don’t change much from one year to another, then it becomes very tempting to teach to the exam rather than actually teaching things that are useful to know. This effect is exacerbated by the fact that schools are rated by the proportion of pupils achieving at least a C, so they’ll redirect a large part of the resources on pupils that fluctuate between a C and a D — or, to express it in marks, the pupils who are are expected to get between 35 and 43 marks and can be pulled up over the 43 threshold.
This is why it’s useful not to have a specific number of marks needed to get a specific grade. If the schools know that the number of marks needed to get a C can be anything between 35 and 50, depending on the actual paper, they cannot concentrate all their resources on coaching a small number of students.
As I’ve argued before, I think grades should be awarded based on percentages: Once all papers have been graded in the country, a computer should work out the A/B borderline so that 10% of the students get an A, the B/C borderline so that 25% of the students get a B, and so on. This would remove grade inflation overnight.