Why your tutor and your child's school will not give the same grade

A parent wrote to us recently, and he was not happy.

His child’s school had just given her a 4 in English Language and a 7 in Literature. We had told him, a few weeks earlier, that she was working somewhere between a 7 and a 9. His conclusion was reasonable, and he put it plainly: there must be something wrong with the way you mark.

It is a fair challenge, and it deserves a proper answer rather than a defensive one. So here is the answer we gave him, written out in full, because a lot of families hit exactly this moment and nobody explains it to them.

The first thing worth establishing

When two grades disagree, the useful question is not which one is wrong. It is what each of them was measuring.

That sounds like a technicality. It is the whole problem. The instinct is to trust the number that came from the institution. But a grade on an English essay is not a reading from a thermometer. It is one trained person’s judgement of a piece of writing, made against a set of criteria, on a particular day.

We know how much that judgement varies, because the exam regulator has measured it. Ofqual, which oversees exams in England, studied how often a script gets the grade a senior examiner would give it. For a mathematics GCSE, the probability was 0.96. For an English language and literature qualification, it was 0.52.^[1] In other words: in English, a second marker reaches the same grade as the first roughly half the time.

That is not an accusation against teachers. It is simply what marking essays is like. Nearly every script lands within one grade of the “definitive” grade — Ofqual puts that figure above 95 per cent for all subjects.^[1] But within that one-grade window, honest, qualified people disagree, all the time. Two markers looking at the same essay are not going to see everything the same way, and a parent should not expect them to.

So when a school says 4 and a tutor says 7, the first thing to establish is not who is wrong. It is what each of them was actually measuring.

What we mark against, and what we do not

We do not mark a student against another person’s opinion of them. We mark against what we believe, after a lot of work, to be the most accurate understanding of what the examiners reward. That is the whole of it.

In practice it means we teach the writing element by element. A paragraph of analysis has parts: the topic sentence that states the argument, the evidence, the inference, the close analysis, the effect on the reader, the writer’s purpose. Each part is taught, practised and marked on its own, against the criteria the exam board publishes. When a student follows those instructions properly and does the work, we expect them to score highly, because they are building the exact thing the mark scheme describes.

A school mock is a different measurement, and it is worth being honest about how. A mock is a snapshot of one morning. It is often marked quickly, by a teacher with a hundred other scripts to get through, against expectations that may be the school’s own. We have written elsewhere about what a single disappointing grade actually measures, and why a piece of work produced at home with time and help is not the same thing as work produced alone under the clock. Both are real. They are just not the same number.

Structure versus time

There is a deeper reason the two grades diverge, and it is about how the two places organise the year.

A school runs on time. The calendar decides what is taught when, and the mock lands on a date, whatever has been covered by then.

We run on structure. We break Language and Literature down into topics and teach them one at a time, thoroughly, and then we interleave them so that the earlier ones do not fade. A student who is with us in the autumn will have gone deep on some parts of the course and not yet touched others.

Now imagine the school sets a mock on a topic we have not reached yet. There is a good chance the student will not score their best on it. That does not mean our grade was wrong. It means the mock measured a topic we had not taught, on a day we did not choose. The same student, on the topics we had covered, is producing work at the level we said.

We are not going to change that order because a mock fell on a bad week. The order is the method, and the method is what produces the result in the summer.

The student who was given two 5s

One case is worth telling in full, because it shows all of this happening at once.

A student of ours, Ayaan, came home from his mocks with two 5s. His mother was upset, understandably, and asked whether I would meet his school teacher to talk it through.

I told her honestly that I did not think the meeting would help her son, and I explained why rather than being polite about it. Not because the teacher was wrong. Because the meeting could not change what he needed to do next, and one mock was not going to change how we teach. We would keep teaching the way we teach, because the method is built on evidence rather than on any one result. That is the whole reason a single mock does not move us. We have seen a student sit mocks at 4 and 5 and finish the year with an 8 and a 9. We have seen another sit two 7s and finish with a 9 and an 8. Not every student travels that far, and we would never promise it. But when a student knows what to do in the exam, knows what the examiners are rewarding, and has put the work into the writing itself, a mock grade is far more often a starting point than a limit. What we could do was show them exactly where he was strong and where he was not.

Then he did something sensible. He asked his school for a second opinion on the same work. The second marker gave him an 8. So on one piece of writing, one teacher said 5, another said 8, and we had him between 7 and 8.

In the summer he sat the real exams and came out with a 9 in one paper and an 8 in the other. His message afterwards said something we have quoted since: “I was just sticking to the plan that we discussed.”^[2]

Nobody in that story was dishonest. Three people looked at the same student and saw three different things, because that is what happens with essays. The number that settled it was the only one that was ever going to: the one from the examiner.

The word “mock” means what it says

This is the part parents find hardest, and we understand why. A mock result arrives in an envelope, with a grade on it, from the school. It looks exactly like the real thing.

But it is not the real thing. It is called a mock because it is a rehearsal. Nobody applies to college or university with mock grades. A mock matters as information, and that is a real use: it shows you what to fix while there is still time to fix it. What it does not do is follow your child anywhere.

We are focused on the exam, because the exam is what your child will actually use. That focus has a cost: we cannot promise that every mock will go well, and we will not bend our teaching to make a December number look better if that means a worse June. We have written before about why a better mock can sometimes mean a worse GCSE. The short version is that the things which lift a mock quickly are not always the things that lift the real exam.

We are direct about this because the method has a record, and that record is the only reason we hold the line.

Twenty-five or so students sat their GCSEs with us this summer. These are the fifteen whose complete results we hold.

Across the two subjects, 17 of those 30 grades were a 7, 8 or 9, and half of all the grades awarded were an 8 or a 9. Ten of the fifteen students averaged a grade 7 or better. Four came away with at least one grade 9, in a year when a 9 went to 2.4 per cent of English Language entries and 3.2 per cent of English Literature entries nationally.^[3]

One of those four had once been on a 3 and a 4. Another had scored a 4 and a 5 in her mocks. Their mocks did not predict that. Their work did.

What to do when the two numbers disagree

None of this means you should ignore a school grade. It means you should treat it as one reading, and find out what it was reading.

1. Ask the school what the mark was for. A breakdown by question, and if possible the script itself. A 4 that comes from one unanswered question is a different problem from a 4 spread across every answer.
2. Ask which topics the mock covered, and check them against what your child has actually been taught so far, by us or by anyone. A low mark on untaught material is not a verdict.
3. Put the two pieces of work side by side. The mock script and a recent piece marked by us. If they look like two different writers, the difference is usually time, pressure and topic. That is a training problem, and it can be trained.
4. Ask your tutor to show you the criteria. Any tutor placing your child at a 7 should be able to point at the exam board’s own descriptors and show you where the work meets them. If they cannot, that number is an opinion. If they can, it is evidence.

We can be wrong. Our judgement is a judgement too, and we hold it to the same test we ask of everyone else: show the criteria, show the work, keep the record. What we will not do is adopt someone else’s number in place of our own reading of your child’s work. That would look like humility. In practice it would mean giving up our evidence, experience and best judgment to fit someone else’s opinion.

If you would like a proper picture of where your child stands — the specific skills, measured against the examiner’s criteria, not a single grade — our assessment takes about two minutes to start.

Sources