How to decide on a grade?

I had an interesting exchange with Stuart Lock (@StuartLock) on Twitter before the GCSE grades came out. He was asking me whether I thought criteria or norm referencing was the best way to assess pupils under an examination system. This is an old chestnut in education. O-Levels were based on norm referencing and SATs were originally conceived of as criteria referenced when the National Curriculum started in 1988. The reformed GCSEs under Michael Gove implied a return to norm referencing as the mode of assessing performance nationally. I have previously written an explicit defence of norm referencing as the fairest system of accounting for pupil performance (It is well worth reading Daisy Christodoulou and Dylan William on this topic). Being judged against the performance of your peers represents a much fairer judgement than arbitrary external standards as the pressure to perform just produces a levelling down to the necessary criteria. The clearest example of this was GCSE coursework which became an abomination. Starting from a well intentioned effort to get pupils to produce work that demonstrated their initiative and problem solving skills, it descended into farce with thousands of pupils producing the same basic scripts effectively repeated year after year to match the criteria and guarantee ludicrously high marks. Thankfully, Ofqual reformed the examinations to rid us of this and other mechanisms for gaming the system, like modular examinations that can be repeated two or three times to ensure ever better grades.
All of this makes sense. However, norm referencing with its relative scale of judgement does nothing to enhance the standard of education on offer to pupils. Even if you make the examinations tougher it makes no difference to the expected standard for a grade. If the top 20% of the cohort only achieve 50% in an examination then the standard for the top grade drops to 50%. Relative success does not mean the standard of teaching and pupil knowledge and understanding improves. What is needed is a conversation about what the standard should be that we aim for in an educated child for each particular subject. This is exactly what the examination boards used to do.
I remember going to examiners meetings a long time ago where the chief examiner for a paper would explain what the standard was for a C grade or an A grade. We argued for some time about whether we could accept particular responses. The whole time we had to go back to the notion that we were trying to educate pupils to do the subject and be capable of taking it further. It wasn’t an exercise in increasing the numbers reaching the pass mark. It was about finding out what that cohort couldn’t do and what their teachers had failed to get across that year. The examiners then knew what they had to re-examine the following year to make sure the standard improved. This kind of sense of authority came from people who worked as part of a team of examiners for years. Most were either active teachers or retired with many years teaching experience. They knew what the standard was they were defending.
Things have changed. Few of this breed of examiners remain. The examinations are dictated to by external requirements. The pressure on examinations boards is to guarantee the maximum pass rates for the least effort. Despite the current reforms, we are still sold examinations in terms of how easy they are, rather than the standard they uphold for a subject.
For this reason, even with the new GCSE specifications and their tougher questioning of pupils – one hopes – unless we are explicit about the standard we wish pupils to achieve to reach a pass grade be it a 4 or 5, then we are just going to see a new round of schools clamouring for ways to game the system.
Worse still, if Ofqual sticks to its comparability of the number reaching the new pass grade and the old pass grade, then nothing matters. We might as well not have bothered with the reforms in the first place. All we are left with is a sifting exercise that attempts to filter off pupils into the same sized slices of the pie. It matters not one jot how well they do in the examination of their knowledge of the subject. Instead, it is just how well they do in relation to everyone else in the school system in that cohort. We might as well use a colouring in competition or the time to solve a crossword to sift them.
What we need is a community of educators capable of upholding the standard of what an educated child should be able to achieve in every subject. How many teachers and head teachers would be prepared to do that?

David Perks

Comments 2

  1. A few thoughts – you’re better to deal with this than me, but I’ve been following it.

    If we look at Maths, the standout oddity this year in terms of grade boundaries, the obvious answer to your question is that the Government has an idea of the standard that they want 16 year olds to aspire to – the new curriculum, as outlined in official materials, textbooks, the questions posed in mock papers and the actual papers they took.

    This turned out to be very challenging for students at every level, even the top end (to a lesser degree that the lower ends perhaps). So, they had to radically adjust the grade boundaries this year as it would have been shockingly unfair if they hadn’t. It seems to me that there are two key issues now: will Government and exam boards persist with the standards demanded this year; and do they need to modify the examinations if they do? By the latter I mean to query whether it is sustainable or indeed useful to have a pass at around 18% and for that matter the equivalent of the old A at around 52%. In time students might improve and cope with the new material, but it will take time I think.

    In the meantime, perhaps, if the level demanded by the curriculum is to be maintained, the exams need to be modified. For example, somewhat akin to how the A* works at A level, slightly easier papers overall (than this year) could include a final third that was significantly more demanding. A 9 would be awarded for the same level overall as the 8 grade, but combined with a particularly high score on the final thirds (averaged over all the papers) – cf 80% overall and 90% average on C3 and C4 for the A* at A level.

  2. Helpful. My recent experience of my daughter was how schooled she was in the language of criteria for getting grades, boundaries etc. Art, for example, was totally debased. Contrast that with the online English lit teacher she found – Mr Bruff – to help her ‘know’ how to pass. He turned the tables, getting his students to focus on 9. For the first time in a long time, she managed to move beyond obsession with a formula and started enjoying the subject – and to her surprise, a 9 it was!

Leave a Reply