Pupils in Wales will not sit GCSEs, AS-Levels or A-Levels next year. Instead, their grades will be decided through coursework, to be completed in classrooms but set and assessed externally.
The SNP government in Scotland has taken a similar approach, having announced that Nationals (the Scottish equivalent of GCSEs) will not take place but will be replaced by teacher-led assessment, moderated by exam centres, but assessed within schools. (A “sample of work” will be taken by assessment centres, rather than this year’s attempt at moderation based on previous years’ work.)
In Northern Ireland, the power-sharing executive has yet to reach a conclusion one way or the other, but it has not ruled out cancelling exams. In England, the Conservative government has thus far insisted exams will take place – but as we know, that is no guarantee.
All four governments are facing the same question as governments around the world: how do you make sure grades are fair when you can’t control which schools may be disrupted by outbreaks of coronavirus among students or staff?
So, how do the four approaches differ – and which is the best practice at present?
It’s worth remembering that in the United Kingdom, qualifications have three roles: their first role is effectively to act as a token. Six “standard” passes (that is, above a C or 4) get you into an average sixth form, nine A*s or 9s get you into the post-GCSE qualification lane of your choice. With the exception of a handful of technical qualifications – of the “you must have this qualification to start fiddling with those wires or unscrewing that pipe” variety – your qualifications at school are useful only as a form of currency. They don’t really mean very much on their own, but what counts here is comparison within the cohort: my B grade in A-Level English means that your A-grade gets you into Manchester ahead of me, and so on.
There is one very important exception in our system: that passing grade at GCSE or Nationals in English and Maths. Without those qualifications, people are shut out from all but a handful of jobs, and their value broadly lies in the fact that if you are an employer, you can say with reasonable confidence that someone with a pass in English and Maths can happily and effectively work behind a till, read a delivery manifest, or do any other job that requires basic literacy and numeracy. What matters here is the value of the qualification in absolute terms: it doesn’t really matter if you and I both have a 4 in Maths, what matters is whether an employer can trust that our 4 is a reliable indicator of basic numeracy.
But their third role is in assessing the performance of teachers and schools. This is one reason why I favour having more, lower-stakes exams. (One problem at the moment, particularly in England, Scotland and Northern Ireland, is that children have relatively few moments of external assessment, but the ones they have are incredibly high-stakes, both for the pupils and for their schools.) The problem is that while exams in their present form deliver “whole school” accountability, they are pretty poor tools for identifying when something has gone wrong.
By 2023, a child in England will undergo a “baseline assessment” right at the start of their time in primary school. But, with the exception of a numeracy assessment at age nine, they won’t have another set of all-round assessments until their Year 6 Sats aged 11 – and then they will have no centrally assessed exams of any kind until Year 11, when they sit their GCSEs.
The problem here is obvious when you think about it. If a child does very poorly in their baseline assessment, does very well in their numeracy assessment, well in their Year 6 Sats, and then very poorly in their GCSEs, what has gone wrong? Who is to blame? We can say their primary school was clearly doing something right, but our ability to identify best practice or good models of teaching is limited other than pointing at a primary school and saying: “Well, they clearly had the right idea.” But was the problem the teaching in Year 7? Year 8? Year 9? Or did something go wrong solely in Years 10 and 11 at GCSE? And because GCSEs in England are now largely an “all or nothing” set of final exams, by the time we have identified a problem in a secondary school cohort it is, bluntly, too late for that cohort.
In addition, because those exams are “all or nothing” for pupils, teachers and school leaders, they are unavoidably going to be emotionally fraught events. Lower-stakes, higher-frequency exams (provided you avoid the disruptive process of constant retakes) means you can identify problems quickly, creates a more granular picture of what is going well and badly within schools, and avoids the high-pressure environment of the current GCSE system. It also has the insurance process that, in the event of unexpected disruption to exams – which happened on a nationwide scale thanks to the pandemic, but happens at a human level due to family bereavement, sickness and other unexpected events – means you can produce a reasonably fair assessment based on the previous work pupils have completed.
Why not just scrap external assessment and have continuous, teacher-led assessment? Well, because the advantage of external assessment is, frankly, teachers are human like the rest of us. Sometimes they will feel sympathy for a child who has worked hard but has not managed to hit the level required to pass; other times they will take against a child. Sometimes they will get their assessment wrong. External assessment is a good bulwark against that human error. In the rare exceptions where a teacher is actively malign, external assessment is a safeguard against that, too. Whether you do that via exams or by externally assessed coursework, what really matters is the role of an external invigilator or mediator.
Of course, the other question is whether one of those two systems gives better security for employers. Can I, as an employer, look at a candidate with a C or a 4 in Maths who has had their coursework externally assessed, and one who has had their exam results externally assessed, and feel confident that both candidates are equally qualified? If I can’t, then the qualification is meaningless, and it provides nothing of value to pupils.
The reason to believe there is a difference comes back to the three roles of exams in the United Kingdom: because teachers themselves are assessed on the basis of their students’ performances, and because the stakes of not passing English and Maths are so high, from both a self-interested perspective and a pupil well-being perspective, teachers are heavily incentivised to do all they can to help their pupils do well in classroom assessment, which means coursework may not accurately reflect the abilities of those pupils themselves.
I think this is probably true, and I think it is likely as a result that we will see higher grades in Wales in 2021 than we have in previous years. That said, I think it is also probably true that the high-stakes nature of exams means that, as it stands, some people who ought to have that passing grade in English and Maths do not get it, and that the “true” figure as far as the pass rate is concerned probably exists somewhere between the pass rate in Wales in 2019, under the old system, and the pass rate in Wales in 2021.
The trouble in England is that the pre-existing unfairness of exams – that something can just go wrong on the day – is turned up to 11 because some pupils’ schooling will not be disrupted by coronavirus, and some pupils’ will. Some students will have to self-isolate and will miss out. Others will see their teachers have to self-isolate and will end up with patchier teaching as a result. Either way, using “all or nothing” exams rather than continuous assessment is going to produce greater unfairness than in a normal school year.
The system the Scottish government has produced looks to me to be the worst of all worlds. Children taking their Highers and Advanced Highers (AS-Levels and A-Levels in Wales and England) will still face the unfairness experienced by pupils in England. Employers will still have to contend with a qualification that is different from the old pass in Nationals. But because the assessment is not external, pupils do not have the same ability to change their trajectory and defy any of the negative expectations of their teachers.
But it is Wales that will face a particular challenge. I think it will be hard to return to a system of centrally assessed exams, because if it produces roughly similar outcomes to 2019, why have exams? And if it produces more passes, it is hard to switch back to a system that produces worse outcomes for individual pupils. One way or another, I think the new Welsh system is here to stay.
So how, then, can Wales keep the benefits of centrally assessed coursework without the drawbacks? One way around this is to decouple that passing grade in English and Maths from the system of awarding GCSEs: seeing as it is as essential to getting a job as a driving licence is to becoming a delivery driver, it makes sense to have a basic “this would-be employee is literate and numerate” certificate that anyone in Wales can take at any time, whether at school, at another educational institution, in a Job Centre or in prison. I suspect the life-altering consequences of not getting that pass in Maths and English is the single biggest reason there is grade inflation, and if you removed this, you would remove one of the major inflationary pressures in our grade system. And even if it didn’t, it would in any case remove the single biggest problem created by grade inflation.