Effort is not a plan

Before students start with us, we ask them an open question, with no options to choose from: what do you find hardest about GCSE English? All 108 students wrote something. They wrote about quotations, about analysis, about structure and about running out of time. The phrase “mark scheme” appeared in none of the answers.

Keep that in mind, because we will come back to it. First, picture a student who revises for three hours a night and works through every past paper available. Will that student improve? Very likely, since hard work nearly always pays something back.

The harder question is a different one. How much improvement, and improvement in what? Is the work closing the gaps this particular student has, or is it strengthening what they could already do? And is it building the skills their target grade asks for, given that the skills which carry a student to a grade 6 are different from the skills that carry them to an 8 or a 9?

Nobody can answer those questions without knowing what the student is currently getting wrong, and without that knowledge every hour is a guess. It may be a good guess. Even so, a guess is an expensive way to spend a year.

The calendar, the feeling and the panic

We are not arguing against revision guides, or videos, or hard work, or school teaching. All of them can be useful, and we use some of them ourselves. The problem is rarely the activity itself, but how the activity got chosen.

There are three ways it usually gets chosen, and most families will recognise at least one of them.

The calendar chooses. We move on to the next topic because it is week six, not because week five worked.

The feeling chooses. The work feels productive, so it must be working, and some of it will be. But a student who re-reads a chapter because it is comfortable, rather than because they have found something in it they cannot yet do, is choosing by mood.

The panic chooses. One mock comes back low, so everything changes at once, with a new tutor, a new book and a new method, before anyone has established what actually went wrong.

All three make the same mistake. Something has been decided about what a student does next, without first finding out what that student does not yet understand.

It would be easy, and wrong, to blame teachers for this. A teacher with thirty students in a room and one fixed exam date in June cannot give each student a different amount of time, which is a matter of arithmetic rather than a criticism. Benjamin Bloom made the same point about school systems in 1968. He called the built-in expectation that a third of students will fail “the most wasteful and destructive aspect of the present educational system.”^[1] He was describing a structure rather than a profession, and so are we.

Most students cannot say what is missing

Before students start with us, we also ask them to rate their own understanding of several things. One question asks how well they understand the mark scheme, the document the examiner uses to award their grade.

Of 107 students who answered between June 2025 and June 2026, 82% said they had “a general idea” of the mark scheme or less. Four said they knew it inside out.^[2]

A second, smaller survey asks the same thing in the first lesson of one of our courses. Of 46 students, 85% gave the same answer, and none chose “inside out”. Seventeen of those 46 had also filled in the first form, so the second survey is a consistent check on the finding rather than a separate one, taken at the same point in a student’s journey.^[3]

These responses describe students’ confidence before or at the start of a course. The responses do not measure performance in an exam. Nor do they represent GCSE students in general, because every one of these families had already come to us looking for help.

We also ask students to rate themselves out of five on using quotations, and out of five on exploring ideas. On quotations, 7 students out of 102 chose the highest rating. On exploring ideas, 2 out of 108 did, and 27% rated themselves 4 or 5, against 43% on quotations.

Both numbers are low, so why look twice at them? Because of the gap between them. Students feel more confident about handling quotations than about doing something with the ideas in them, and quotation-handling is the more visible of the two skills. It looks like a technique, whereas exploring ideas does not look like anything until somebody shows you what it is.

The mark scheme is clear about which of the two carries the answer. AQA’s top band in GCSE English Literature, Paper 1, asks for a “critical, exploratory, conceptualised response to task and whole text”, supported by “judicious use of precise references”.^[4] The references are there to support the argument, rather than to be the argument.

We use AQA’s papers as our example in this article. The other boards we checked also ask for well-chosen evidence at their top level. We give each board’s own words in the section on what the marks reward.

Now go back to the open question from the start, where the phrase “mark scheme” appeared in none of the 108 answers. That is an observation, not a proof. It does not establish that no student understood the mark scheme mattered, since they were answering a question about difficulty, and a document you have never worked with does not feel difficult. But set it beside the 82%, and it describes a real situation: the thing most of these students cannot explain is not a thing they are worried about.

Nobody can judge work without a standard

What is the name for the skill that is missing here? It is metacognition, one of the better-evidenced ideas in education. Metacognition means thinking about your own thinking: planning how you will approach a task, monitoring how it is going while you do it, and evaluating how it went afterwards. The Education Endowment Foundation’s review of 355 studies puts strategies that teach it among the strongest available, at around eight months of additional progress.^[5]

For a GCSE English student, it comes down to five questions they should be able to ask about their own work:

What do I actually understand here? Where does this piece of writing fall short? What evidence am I basing that judgement on? What should I practise next? And how will I know if it has improved?

The third question is the one that does the work, and a student cannot answer it alone, because a judgement needs something to judge against. So we teach students to hold their own paragraph beside the criteria and mark it. The aim is not to turn them into examiners. A student who has compared their writing with the standard has something better than a feeling about it, because they can point at the sentence that is missing.

Confidence ratings like the ones above tell us what students believe about themselves. They become far more useful when you put them next to the student’s actual work, because then you can see how accurate those beliefs are. Take a student who rates themselves 4 out of 5 and writes a grade 5 paragraph. What that student needs is to see the distance, once and clearly, far more than they need extra hours.

The marks reward chosen evidence, not more of it

The belief we meet most often in lessons is that more quotations mean more marks, so the obvious way to improve a paragraph is to add one.

The document says otherwise. In AQA’s GCSE English Literature mark scheme, the top level asks for “judicious use of precise references to support interpretation(s)”. In English Language Paper 1, it asks the student to select “a range of judicious textual detail”.^[4]

Judicious means well-judged, or chosen with good sense. It is AQA’s word for choosing the right evidence rather than gathering a lot of it. In AQA’s documents, the word appears only in the level descriptors, which describe what each grade looks like, and never in the assessment objectives themselves. No AQA descriptor in either subject says anything about how many quotations an answer should contain.

Other boards use different words for the same idea. Here is how each board we checked describes evidence at the top level.

Pearson Edexcel GCSE. English Literature Paper 1: “Discerning references are an integral part of the response”. English Language Paper 1, Question 4: “The selection of references is apt and discriminating”.^[14]

Eduqas GCSE. English Literature Component 2: “pertinent, direct references”. English Language Component 1, Question 2: “a well-chosen range of evidence”.^[15]

OCR GCSE. English Literature Component 1: “Textual references and quotations are precise, pertinent and skilfully interwoven”. English Language Component 1, Question 4: “apt, skilfully selected and integrated textual references”.^[16]

Pearson Edexcel International GCSE. English Literature: “Discriminating use of relevant examples in support”. English Language A, Paper 1, Question 4: “The selection of references is discriminating”.^[17]

Cambridge IGCSE. Literature in English, Paper 1: “well-selected reference to the text skilfully and with flair”. First Language English, Paper 1, Question 3(d), in the specimen paper for 2027: “precisely selected words/phrases”.^[18]

We have not checked CCEA’s mark schemes for this article.

Discerning and discriminating both describe evidence chosen with good judgement. None of the top-level wordings above gives a number of quotations.

A student who adds a fourth quotation to a paragraph has usually made that paragraph worse. A student who takes three words from one quotation and works out what they are doing has usually made it better. A student rarely discovers this by working harder, because it has to be shown to them and then practised. We have written separately about how the mark scheme’s own vocabulary misleads students who have read it, which is a related problem.

A detail earns its place by illuminating the whole

Here is what teaching from evidence changes in an actual lesson.

Ask a student what a metaphor is and you will usually get this answer: describing something without using “like” or “as”. The answer is correct, but close to useless, because it defines a metaphor by what it is not. A student holding that definition can identify a metaphor correctly and still have nothing to say about it.

Professor William Dowling of Rutgers University opens a class on Browning’s My Last Duchess differently. A metaphor, he tells his students, is “a counterlogical use of language”. It is “a statement that makes no sense whatsoever on the literal level but makes all the sense in the world on a higher level.”^[6] Buildings cannot talk, so “the White House announced today” is nonsense when read literally, yet nobody pauses over it. You go straight to the higher level and understand it.

Then he gives his students the part that turns a definition into a method. Interpreting a metaphor means deciding what is relevant and discarding the rest. Take Spenser’s line about a beautiful woman: she is the rose, the glory of the day. The beauty of a rose is relevant, and so are its fragrance and its freshness. That roses need fertilising, get eaten by beetles and have to be watered is all irrelevant. And the reason the line works, Dowling says, is that “the resemblances between the X and the Y have become so overwhelming that the speaker has stopped seeing the differences. And thus you see how the speaker sees the world.”

That definition gives a student something the first one never could. A metaphor stops being a label to spot and becomes a way into a mind. And the business of keeping what is relevant and throwing away the rest is the same skill the examiner called judicious: one idea, appearing in two documents that never mention each other.

The same move works on a whole text. One of the extracts on an AQA GCSE English Language Paper 1 specimen paper comes from City of the Beasts by Isabel Allende. A boy called Alex wakes in the night. His mother is seriously ill, and outside there is a storm.

Question 2 on that paper is worth eight marks: “How does the writer use language here to describe the effects of the storm?”^[7]

Most students answer it as a question about weather. They write about “lashing” and “roaring” and “furious”, name the techniques correctly, and score in the middle. Nothing they have written is wrong, but the answer is thin, because they have treated the storm as scenery.

What if the storm is more than scenery? What if it is what is happening inside a boy whose mother might die? Read that way, the wind is doing a job in that passage, and the job is Alex.

Question 4 on the same paper is worth twenty marks. It asks students how far they agree that “Alex is struggling to cope with his mother’s illness.” The board has put the storm and the mother two questions apart, on one paper, for twenty-eight marks between them. A student who reads the storm as weather answers Question 2 thinly and arrives at Question 4 with nothing built. A student who reads the storm as Alex answers both.

Other boards set a similar pair of questions on one fiction extract. Pearson Edexcel GCSE English Language Paper 1 asks how the writer uses language and structure in Question 3, for 6 marks. Question 4 then asks students to evaluate the same extract, for 15 marks.^[14] Eduqas GCSE English Language Component 1 asks what impression the writer creates of a character in Question 2, for 5 marks. Question 5 asks how far students agree with a statement about the same extract, for 10 marks.^[15] We have not compared this pairing on the OCR, Cambridge IGCSE, Edexcel International GCSE or CCEA papers.

The same move works far beyond one paper. Ask what the fog and the locked doors are doing in Jekyll and Hyde. That strange, dark, half-dreamlike London is there to do more than set a scene, because it is a Gothic setting doing the novel’s real work: showing a divided mind from the outside.

So the teachable move is to ask of any detail what is this doing for the whole text? That is a better rule than “weather means feelings”, which is sometimes false. The same question unlocks a character. Most main characters change, and asking “who changes most?” is the quickest way to find the one the story belongs to. Some main characters do not change at all. Sherlock Holmes ends every case the same man he began it, and where that happens it is usually because the story is about him changing the world around him instead. Either answer opens an essay.

A mock is for finding what is missing

Families are put through more unnecessary distress here than anywhere else in the two years.

No exam board sets, marks or moderates a mock as standard. The school chooses the paper, runs it when the timetable allows, and marks it in-house. There is no national standard, which is why a grade 5 at one school is not the same as a grade 5 at the school down the road. A sixth-form place rests on real results, not mock results. Ashton Sixth Form College states it plainly: “We can only enrol you based on the grades we see on GCSE results day.”^[8]

Does that mean mocks are meaningless? No, and the claim that “mocks don’t mean anything” is false too. Mocks feed the predicted grade that gets a student their conditional offer in the first place, the offer they then have to meet in August. They are the main evidence a school uses when deciding which tier a student sits in subjects that have tiers. And where a student needs extra time in the real exams, the rules require the school to produce mock papers showing that extra time was used.^[9] Most parents have never been told that last one.

What a mock is genuinely for is none of those things. A mock is the last inexpensive opportunity to find out what a student does not know while there is still time to fix it.

The obvious thing to write at the top of a marked mock is a grade. The research says something uncomfortable about that grade. In a study reported by Dylan Wiliam, one group of students was given only comments on their work, one group only marks, and one group both. The comments-only group scored around 30% higher on their next piece of work. The marks-only group gained nothing, and the group given both marks and comments also gained nothing. As Wiliam puts it: “if you write careful diagnostic comments on a student’s work, and then put a score or grade on it, you are wasting your time.”^[10]

So the number does not simply fail to help. It appears to cancel the help sitting beside it. Ofsted’s review of English says something similar in flatter language: grade descriptors “created for summative purposes may not function effectively to provide formative feedback”.^[11] A grade tells you where a student finished, but it tells nobody what to do on Monday.

A disappointing mock in November is information, and it only becomes a prediction if nothing changes. We have also written about why a better mock can mean a worse GCSE.

Hold the standard, vary the time

Our process is not complicated to describe. We find out what a student can currently do, we identify what specifically is missing, and we choose practice that targets that gap. Then we mark the next piece of work and use it to decide what comes next. Where something is not yet right, the student does not move past it, because they are taught the correction and write it again.

Barak Rosenshine, summarising decades of classroom research for teachers, put the principle in one sentence: “Only after the students have mastered the first step do teachers proceed to the next step.”^[12]

The formal name is mastery learning, and the phrase needs care, because it is used for two different things in this country. Many parents will have met “teaching for mastery” in primary school maths, where the whole class is deliberately kept together, and we mean something else. We mean the older idea from Bloom, in which the standard stays fixed and the time is allowed to vary. One student needs three attempts at a paragraph structure, while another needs one, and both leave with it.

The Education Endowment Foundation rates mastery learning at around five months of additional progress and rates the strength of that evidence as low. It notes that the effects are largest in primary maths, and it warns that the approach is “much less effective when pupils work at their own pace”.^[13]

That warning tells you which version to look for. Mastery learning works when a person marks the work, teaches the correction and holds the standard, which a pile of videos and a login cannot do.

Our results, and their limits

Fifteen of our students sat GCSE English in summer 2026. Their average grade across both subjects was 6.87. Two-thirds averaged a grade 7 or above, and half of all the grades they were awarded were 8s or 9s.

Several finished several grades above their mocks. One student who joined us on a grade 3 and a grade 4 finished with a 9 and an 8. One student’s Language grade came out one grade lower than his mock.

There are two limits on those numbers. Fifteen students is a small group, and the mock grades several of them are measured against were recorded from memory rather than from a document, so we treat the individual movements as indicative rather than exact.

The survey figures earlier in this article have a limit of their own. We ask those questions when a student arrives, and we have not asked the same questions again at the end, so we can show you what students believed about themselves on the way in, but not the change.

Why are our figures the only ones here? Because there is no national data on how much students typically move between mocks and the real exam, since mocks are school-set and school-marked and nobody collects them. Confident numbers on this circulate online, and every one we traced led back to a tutoring company’s own marketing.

Three questions to ask this week

You do not need to be able to mark an essay to hold a process to account. When a piece of work comes back, from us, from school or from anyone, three questions do most of the work.

1. What specifically was weak in this? Ask about the weakness itself rather than the grade. “Her topic sentences state the topic instead of making a point” is an answer, while “she needs to go deeper” is not.

2. What practice targets that weakness? There should be a direct line from the weakness to the next task. If the answer is “more past papers”, ask what the last past paper told anybody.

3. How will we know it worked? There should be a next piece of work, and somebody should be looking at that exact thing in it.

If a teacher or a tutor can answer those three questions about your child, they are working from evidence. If they cannot, somebody is guessing, and your child is paying for the guess in hours.

Understanding what English rewards is a little like understanding a sport before you watch it. Cricket looks slow and strange until somebody explains what the fielders are for and how a wicket is taken, and then the same match becomes interesting. Nothing about the game changed. What changed is that you could see what everyone was trying to do.

That kind of seeing is what we are trying to give a fifteen-year-old who has decided they are simply not good at English. Usually they are not bad at it at all. Usually nobody has shown them what the marks are for.

Sources