Rethinking assessment by Barnaby Lenon

A group of independent and state school teachers have announced that they do not like GCSEs.  I am glad – it is always good to question the way we do things. So we might help them by laying out a few background thoughts.

1) What are GCSEs for?

You cannot reform GCSEs without answering this question.

One aim is to rank pupils for the benefit of universities, colleges and employers who want to use grades for selection purposes. My hairdresser, for example, only recruits young women with a grade 4 and above in English and maths. Experience has taught him that if you can’t pass English and maths GCSEs you are unlikely to have some of the skills he needs. And these grades act as a measure of general intelligence too.

Another purpose is to determine which courses pupils should go onto next. It’s all very well if you are a highly selective private school, but most schools need to sift out those pupils capable of doing A-levels, which should do BTECs and which should go to a Further Education college to do a lower level vocational course. GCSE grades are a fair and reliable way of doing this.  Every secondary school in England requires a minimum number of GCSE passes if you want to stay on and take A-levels. Most require a very good GCSE grade in the subject concerned if you want to take A-level maths, sciences or modern languages. They do this for one main reason – that years of experience tell us that someone with a weak GCSE maths grade will struggle to pass the A-level.

Another purpose is to accredit the years of work the children have put into these subjects. The average child takes 8 or 9 GCSE subjects and will only continue with two of these into the sixth form.  Do we not want to mark the end of their education in the majority of subjects with some form of certification?

Like it or not, one aim of GCSEs is to judge schools. Given that the tax-payer funds state schools and we believe in giving parents reliable information about their local schools, we need reliable data about school performance. The main indicator here is the Progress 8 score – the progress made by each child from the age of 11 to 16 – and that has to be based on some form of reliable assessment at age 16.

Most teachers would agree that another aim of GCSEs is to motivate pupils to commit knowledge to the long-term memory. I have taught many thousands of teenage boys. They do not generally work hard, they are not especially interested in pleasing their teachers. But many do start to worry when GCSEs approach. They know that their grades will have an impact on their life chances. Every teacher of teenage boys knows that they pick-up speed as GCSEs approach.

But why should they be forced to memorise stuff? Because having information in your long-term memory is very important. You cannot think critically or analytically about a subject if you do not know something about it. And it is simply not true that we ‘forget what we learn at school’. I can still remember most of the poems, most of the science, much of the French I learnt at school.

Too many children are taught things but learn little. It is commitment to the long-term memory that is the most important element. Boys, especially, need to be driven to memorise and exams are that driver.

2) Different types of school have different gripes

Eton and St Paul’s Girls are the most academically selective schools in Europe. All their pupils get incredible GCSE results and all stay on to take A-levels. So of course they do not need GCSE results. But in this respect they are very unusual: on average 60% of pupils move school/college after GCSEs. They must to take with them some indication of their achievement in their first five years at secondary school.

ASCL, on the other hand, represents hundreds of comprehensive schools. For them the problem is ‘the forgotten third’, that is, the third who fail English and maths (and in many cases, other GCSEs). It seems cruel that after all those years studying these subjects (12 years in fact) the bottom third have little to show for it.

In the past twenty years there has been growing concern about the ‘attainment gap’ between disadvantaged pupils and the rest. It is worth noting that reforms to exam systems tend to advantage the already advantaged because they have more resources to make the changes work. Those schools, like Michaela Community School or the Harris, Ark and STAR academies, that have started to get stunning GCSE results with pupils on free school meals, will have to be reassured that their progress will not be undermined by further changes.

 3) The underlying problems with exams

There are four underlying problems:

1. The high-stakes nature of exam results: they determine the future of the children, the futures of the teachers and the future of the schools. Schools with poor Progress 8 scores, for example, get poor Ofsted inspection grades (often initiating a cycle of decline), they are forced to become Academies or forced to move from one Academy chain to another. Head teachers lose their jobs.

This is not the fault of the exams themselves…it is the fault of governments.

Exam results do matter more now than in the past. I taught A-levels at Eton in the 1980s and at that time pupils who did badly went into the army or into a job; there was little stigma attached to those who did not go to university.

That must be because in those days a relatively small proportion of the school population went to university; now half do.

In the past pupils could leave school and go into a job which did not require exam results.  Before 1965 only 20% of the population took exams at age 16 (O-levels). In 1965 the CSE (Certificate of Secondary Education) was introduced for the next 40% who were interested in vocational courses. In 1988 the GCSE began and it was only at that point that that all children took exams at 16.

The economy has changed so that now most jobs require a reasonable degree of numeracy and literacy. What you are supposed to learn at school has come to matter more.

2. The comparable outcomes approach to grading.

In the period 1990 to 2010 we saw rampant grade inflation. This got to the point where mediocre pupils were getting top results. The more selective universities and employers were no longer able to use GCSE or A-level results to choose students so they had to invent their own tests.

In 2010 Ofqual and Michael Gove decided to put a stop to this. They did so by fixing the proportion of pupils getting each grade in each subject. So if last year 65% of children got a grade 4 or better in English Language GCSE, next year the same proportion would get a 4 or better.

The difficulty with this approach is that if the ability of the cohort improved, perhaps because they were better taught, 35% would still fail. That seems unfair.

In order to meet this criticism Ofqual introduced National Reference Tests in English and maths A large sample of pupils take the tests in the March before their GCSEs. The standard of this test is the same every year and, if the average mark creeps up, Ofqual allows the number getting each GCSE pass grade to inflate.

But this still leaves the ASCL criticism – that far too many pupils fail their GCSEs. Most have learnt something at school, but that does not seem to be acknowledged.

This is partly because society has designated grade 4 GCSE as the pass and every grade below 4 as fail. When GCSE grades were introduced there was no mention of pass or fail. So how people will interpret and use exam results must be an extremely important part of any attempt to reform the system.

3. Rigidity in the system

A number of things have happened which have made the exam system much less flexible than in the past. When I was a head of department the (then) Oxford and Cambridge Board allowed me to write my own O-level Geography syllabus and change it every year to reflect the interests of the staff. That was flexibility.

After 1988 the state started to determine what should actually be taught in schools. The National Curriculum was relaunched in 2011, followed by new GCSE and A-level specifications.  Because Ofqual is rightly concerned about fairness, the three English exam boards had to produce syllabuses based on these specifications so in many respects they are all similar to each other.

Schools that tried to break away from this rigidity found it hard. The International Baccalaureate, an alternative to A-levels, has struggled because of the additional expense; state school funding at sixth form level has been cut. The other alternative to A-levels, the Cambridge Pre-U, was scrapped in 2020.

After 2012 both independent and state schools opted for the international GCSE, offered by Edexcel and Cambridge Assessment. But because of concerns that this qualification would undermine the Gove reformed GCSEs, funding for state schools to offer the international GCSE was withdrawn. Most independent schools carried on with the iGCSE but they, too, came under pressure to switch to the reformed GCSEs because Ofqual and Department for Education research suggested that the grading standards of the iGCSE were different from GCSE (harder and easier depending on the subject and exam board).

Because of the high-stakes nature of GCSEs, Ofqual has understandably wanted to ensure a level playing field for all candidates. This has meant less use of teacher-assessed coursework, fewer opportunities for exam boards to differ from each other and a general antipathy towards the iGCSE.  Modular exams have gone, replaced by terminal exams. Early sitting of GCSEs by able pupils has been discouraged.

Finally, it is worth reminding ourselves that both GCSEs and A-levels have only just been reformed. Neither Ofqual nor the current government ministers are going to be in the mood to think again. Not yet.

4. Exam grade unreliability

Ofqual’s own research measures unreliability in this way: if you give an exam script to ten experienced markers, to what extent do they come up with the same mark? The answer is – very often in the case of maths, sciences, modern languages; these exams are reliable. But often not in the case of essay-based subjective subjects like English and history.

This would not matter hugely were it not for the fact that increasing numbers of universities ask for an A or A* grade in your A-levels. Slipping to a B grade in your history A-level will prevent you taking up your place at Oxford, which is galling if a different marker might well have given you an A.

4) Other complaints made by the Rethinking Assessment group

In their open letter of 27 September there were a number of criticisms made of GCSEs, some of which have been mentioned above. Let us consider the rest.

1. Exams are stressful and pupils are suffering mental health problems as a result.

It is true that recognition of mental health problems in teenagers has grown in recent years.  This is partly because mental health has got worse and partly because we are paying much more attention to it.

It is not true, as the letter says, that ‘thirty or more GCSEs are taken in one month’. The average number of GCSEs taken is eight and no-one takes more than twelve. If you take ten GCSEs you will have to sit 20-25 exam papers, all quite short. But this is a lot.

The problem is that if you reduce the number of papers you are reducing the size of the sample of the syllabus you are examining and that makes the exam less reliable.

Exams are stressful and of course that is one of their merits – the stress is what motivates the pretty large number of unmotivated pupils to do some work. But there may be ways of de-stressing the system and we will come to that.

2. The current exams do not measure some important things.

True – there are plenty of important personal qualities not measured by exams, such as the ability to be a team player or a creative thinker. The question is – can these things be measured and if so, are we sure we want to? I would not myself want pupils formally assessed for their ability to play football or perform in a school play: the knowledge that they were being judged as part of a high-stakes system would reduce the enjoyment. But these things need not be ignored either. I would expect them to be highlighted in any reference than a school might write.

3. We are too exam-orientated.

True – but this must be partly the fault of parents and schools. It has not become harder to ‘get into university’; it has become easier. Teachers and parents who stress the importance of exam grades the whole time are causing the stress they are now complaining about.

It is worth acknowledging the fact the Michael Gove, after 2010, greatly reduced the number and frequency of exams sat by our children by scrapping modules, module re-sits and January exams.

5) So what might be done?

Things I would not do:

I would avoid replacing written exams with too much teacher assessed coursework. In England we cut back on teacher-assessed coursework after 2012 and there were good reasons for that:

  • teachers were under a great deal of pressure to ensure that their pupils got good results. So they, and parents, gave much undue assistance. The system was unfair because pupils with pushy parents and teachers did best.
  • many teachers and pupils found that the continuous assessment through coursework was stressful and burdensome.
  • the coursework in subjects like maths and science was formulaic and dull.  It did little to generate motivation or real knowledge of the subject.

Years of experience tell us that teacher-assessed coursework sounds better than it is.

Things we could do:

Independent schools are free to choose what they do. Some, such as Bedales, Sevenoaks and St. Edward’s Oxford, have already moved away from GCSEs. The important thing is to persuade universities – will you accept our alternative qualification? – and parents. But such schools need to be careful because parents are quite conservative and some choose independent schools precisely because of their good GCSE and A-level results. Many schools which have embarked on the International Baccalaureate have found it hard to build demand.

Nevertheless, we could well move to a less rigid system than ‘GCSEs or nothing’, starting with those subjects that find the GCSE particularly limiting.

Many English teachers are unhappy with the GCSE, maths and science teachers less so. If different schools used different types of assessment for English at age 16, universities would almost certainly not object.

In recent years central government has driven the exam system in a way that was unthinkable before 1990. Department for Education funding rules, DfE School Performance tables and Ofsted inspection criteria are the levers which have been used, with great success. So little can change in state schools without flexibility from governments.

We should encourage schools to move away from an undue focus on exams. Schools like Eton spend almost as much time on non-examined activities as teaching for GCSEs and A-levels.  The recent Ofsted inspection reforms point in the right direction in this respect. Schools might devise their own diplomas to recognise achievement in things like sport, drama, public speaking and community service. Both the Duke of Edinburgh Award Scheme and the International Baccalaureate already do this.

We should begin a national programme to ensure every pupil aged 11 and above has access to a computer, keyboard and internet access at home. Covid-19 has shown that these things are a basic necessity for secondary-age pupils. If we had that, we would be able to contemplate exams done online and exams done using a keyboard rather than a pen.

We could certainly contemplate increasing the teacher-assessment element in some GCSEs but to do that we would need to train teachers in the ethics of coursework marking, stop head teachers leaning on their staff to push up marks, and appoint exam officers in schools who have real authority and status. Schools could work in clusters to moderate each other’s work.

The problem of ‘the forgotten third’ has been caused by the fact that is has proved impossible to reconcile the competing claims of those who want to:

  • offer the same exam to all ability levels
  • and stretch the most able (hence the introduction of grade A* and then grade 9 at GCSE)
  • and award something useful to the bottom third
  • and have a system which can be understood by universities, colleges and employers

To deal with the problem of ‘the forgotten third’ we could require all pupils to take functional skills tests of basic literacy and numeracy when they are ready to do so. A much higher proportion of the forgotten third would then get a qualification. After all, employers are interested in ability to write and do basic maths, not how many poems you have analysed or how proficient you are at geometry and algebra.

Some will argue for vocational alternatives to GCSEs for those who are less academic.

But there remains a strong argument against this. Past experience suggests that if schools are allowed to push pupils in a particular direction at the age of 13 or 14 (for example, vocational alternatives to GCSEs) they often get it wrong. We know that able pupils from disadvantaged backgrounds were routinely denied access to the EBacc GCSE subjects in favour of vocational alternatives, qualifications that Alison Wolf found to be worse than useless (Wolf Report, 2011). If an exam has no status in the public realm it turns into a negative qualification – it signals you as being unintelligent.

So the only way vocational alternatives can be made to work is if the course is so good that it leads onto other good vocational courses which in turn lead to good jobs. There is no point in pleading ‘parity of esteem’ for vocational courses. Evidence that they lead to a decent job is needed.

So long as English and history teachers want to use essays to examine students (as opposed, say, to multiple choice questions) there will be a degree of mark and grade unreliability. The solution might be to publish alongside the grade a measure of how secure the grade actually is and how close the student was to a higher or lower grade boundary. Then the users of these qualifications might, if we are lucky, use the grades more intelligently.

There is nothing new about plans to scrap GCSEs. But the formation of this latest pressure group is a useful reminder that there is no system that can please everyone. Reforms take a decade to put in place, so now is not a bad time to start thinking afresh.

Barnaby Lenon

Dean of Education, University of Buckingham

Leave a Reply

Your email address will not be published. Required fields are marked *

Create a website or blog at

Up ↑

%d bloggers like this: