Statistics tutor: what the 2026 exam data actually shows

GCSE Statistics drew 38,770 UK entries in summer 2026 and 22.6% of them were awarded grade 7 or above — a better rate than GCSE Mathematics managed. This page sets out what the published data shows about who sits the qualification, when they sit it, what the papers demand, and exactly where statistics is examined inside A-level Mathematics on each board.

How big is GCSE Statistics, and how selective is it?

Start with the size of the thing. In the summer 2026 series the Joint Council for Qualifications recorded 912,410 UK entries for GCSE Mathematics and 38,770 for GCSE Statistics across all ages. Statistics is a specialist option that a minority of schools offer, not a second maths GCSE that everybody sits. Any tutor who talks about it as though it were universal has not looked at the entry data.

The results run the other way. Across all ages, 22.6% of GCSE Statistics entries were awarded grade 7 or above and 72.3% reached grade 4 or above. The equivalent figures for GCSE Mathematics were 17.0% and 57.3%. On paper the smaller subject is the higher-performing one at both key grades, and by a clear margin at each.

Nothing in that comparison says statistics is an easier subject. It says the two cohorts are not comparable. Almost every sixteen-year-old in the country sits GCSE Mathematics; GCSE Statistics is entered set by set, usually from groups already working confidently on the higher tier, and usually alongside maths rather than instead of it. A self-selected cohort produces a better distribution. That is a fact about entry policy, not evidence about difficulty.

JCQ splits its GCSE tables into four age bands — 15 and under, 16, 17 and over, and all ages — and mixing them is the standard error on this data. The table below keeps them apart. Every figure is all-UK, male and female, from the summer 2026 series.

The two middle rows are the interesting ones. Of 38,770 entries, 11,296 came from candidates aged 15 or under — 29% of the whole subject — and that group was awarded grade 7 or above at 32.9%. The 26,780 candidates aged 16 managed 18.4%. More than 82% of the younger group reached grade 4 or above. The early cohort is not keeping pace with the main one. It is comfortably ahead of it.

For a family planning GCSE tuition, that one comparison is more useful than any general advice about the subject, because it suggests the deadline may fall a year earlier than assumed. The next section deals with that directly.

Why do so many candidates sit the exam in Year 10?

Two independent sources agree on early entry, which is what makes it safe to plan around. JCQ counts by age across the whole UK; Ofqual counts by year group, England only, rounded to the nearest five, at provisional entry stage. Both show the same shape, and Ofqual shows it in the form a parent can act on.

Year 10 accounts for 11,060 of 39,360 English entries against 27,150 in Year 11. A further 610 entries came from Year 9 and below, with 350 in Year 12 and 195 in Year 13 and above. Whatever the school prospectus implies, a large minority of this qualification is sat before the GCSE year proper.

Ofqual's total is 39,360 and JCQ's is 38,770. That is not a discrepancy and neither number is wrong. Ofqual is counting England, by year group, rounded, before the series; JCQ is counting the whole UK, by age, after results. Use Ofqual for the shape of the entry and JCQ for outcomes, and never split the difference between them.

The practical consequence is a compressed timetable. A Year 10 entry means the statistical enquiry cycle, the sampling methods, the standardised scores and the correlation work all have to be secure while the student is still part-way through the maths GCSE course, and in a year that already carries mock exams in everything else. Families who assume statistics is a Year 11 subject routinely start support twelve months late.

It also changes what good teaching looks like. A Year 10 candidate has met less algebra and less graphical work than a Year 11 candidate, so the sequencing has to lean on what they already have rather than on what the textbook assumes they have. That is the practical difference between a subject specialist and a general maths tutor picking the subject up.

Is GCSE Statistics growing, or does it just feel that way?

It is growing, and the growth is measurable rather than anecdotal. Ofqual publishes provisional entries for England as a five-year series, so the trend can be read from one source without stitching anything together.

That is a rise of 74% in four years, from 22,615 in summer 2022 to 39,360 in summer 2026. The denominator is what makes the claim honest: across all subjects, GCSE entries rose 1.1% in a year, from 5,777,020 in summer 2025 to 5,840,185 in summer 2026. GCSE Statistics rose 11.9% year on year on Ofqual's England figures and 11.2% on JCQ's UK figures. Either way it is growing several times faster than the exam system around it.

Growth on that scale usually means schools adding the subject to an existing option block rather than thousands of individual students discovering it. It also means the supply of teachers with statistics-specific experience has not grown at the same rate, which is the ordinary reason a parent ends up searching for a statistics tutor rather than a maths one.

None of this makes the qualification a better choice than it was, and it would be dishonest to suggest that it does. What the numbers support is narrower and more useful: statistics is no longer a curiosity on a timetable, and a student sitting it now belongs to a national cohort of nearly forty thousand, with published boundaries and published outcomes to be measured against.

Where does statistics sit inside A-level Mathematics?

Every A-level Mathematics student in the country studies statistics, whether they chose to or not, because the content is not the board's decision to make. AQA states it plainly in its own specification: "The subject content for A-level Mathematics is set out by the Department for Education (DfE) and is common across all exam boards." What each board does decide is where that content is assessed — and the three major boards have answered that question three different ways.

AQA is the cleanest arrangement for a student who wants to isolate the subject. Sections K to O — statistical sampling, data presentation and interpretation, probability, statistical distributions and statistical hypothesis testing — are examined on Paper 3 and nowhere else, while mechanics sits on Paper 2 alongside vectors. If you are on AQA 7357, one whole paper of the three is the statistics paper.

Pearson Edexcel does the opposite. Paper 3 of 9MA0 is a single 100-mark paper split into Section A statistics, topics 1 to 5, and Section B mechanics, topics 6 to 9. Neglecting one section damages a paper you need the other section for as well. Edexcel also publishes a large data set which, in Pearson's words, "will support the assessment of Statistics in Paper 3: Statistics and Mechanics", and students are required to become familiar with it in advance. That familiarity is worth marks and cannot be acquired in the final fortnight.

OCR's A-level Mathematics A takes a third route: component 02 pairs pure mathematics with statistics, and component 03 pairs pure with mechanics. An OCR student therefore cannot revise statistics as a self-contained block the way an AQA student can, because it shares a paper with pure content and the pacing of the paper reflects that. Which of the three arrangements applies is the first thing worth establishing about any A-level maths tuition, and it is the question the tutor marketplaces ranking for this subject never ask.

Which A-level Maths paper was hardest to score on in June 2026?

AQA publishes notional component boundaries alongside its subject boundaries. They are illustrative rather than awarded grades — candidates are graded on the total, never paper by paper — but they do show what each paper demanded in a given series, and in June 2026 they point one way.

The statistics paper sat above the pure paper at every grade on the scale. An A on Paper 3 needed 76 marks of 100 while the same grade on Paper 1 needed 71 — 5 marks more for the same letter — and the gap runs in the same direction at A*, B, C, D and E.

For context, the qualification as a whole is scaled to 300 marks, and these were AQA's subject boundaries for the same series.

This is the strongest single argument for treating statistics as its own body of work rather than as the tail end of the maths course. It is also the sort of evidence tutoring sites do not tend to look at, because notional component boundaries sit further down the same document as the headline figures and are explicitly labelled as being for information only. Read them with that caveat attached: one series is not a trend, and a higher boundary can mean an accessible paper as easily as a punishing one.

What the boundaries do establish is that the statistics paper is not the soft option inside A-level Mathematics that students often expect it to be. Plenty arrive at the first session describing the statistics content as "the easy third" of the course, on the basis that it involves fewer manipulations and more reading. AQA's own numbers for this summer say otherwise, and a student who has budgeted their revision on that assumption is the student who loses marks on Paper 3.

What has happened to AQA GCSE Statistics grade boundaries?

AQA published the June 2026 GCSE boundaries on 20 August 2026. Set against the two previous summers, the Higher tier tells a consistent story, and it is not the story this page used to tell.

Grade 9 has fallen from 135 marks out of 160 in June 2024 to 129 in June 2025 and 114 in June 2026: a 21-mark fall across two series, with every other Higher grade moving down alongside it. This page previously described the movement as a post-pandemic climb back towards normal, which was a reasonable reading when 2024 was the newest data available. It is not what the series shows now, and that paragraph has been rewritten rather than quietly left in place.

Foundation tier, which is capped at grade 5, moved the same way over the same period.

A falling boundary is a statement about the paper, not about the students. Boundaries are set after marking so that a grade means roughly the same thing from year to year; when a paper is harder, the marks needed come down. The honest reading of 135, then 129, then 114 is that the recent papers have been more demanding than the 2024 paper — not that grade 9 has become easier to earn. A Higher candidate aiming at grade 7 needed 83 marks of 160 this summer; a Foundation candidate needed 92 for the grade 5 the tier allows.

The practical use of a series like this is mock marking. A raw mock score means very little by itself, but the same score set against three published boundary sets tells you whether a target grade is realistic and how much of the shortfall is missing content rather than exam technique. Nobody can use last year's boundaries to predict next year's, and any tutor who claims to is guessing. What the range can do is tell a family what a mark is currently worth.

Earlier versions of this table also carried 2019, 2022 and 2023 rows. Those have been removed, because they could not be verified against AQA's currently published boundary documents in this review. Three verified series are more use than six rows of uncertain provenance.

What the AQA 8382 papers ask — and the two things they never ask

AQA GCSE Statistics 8382 is examined in two papers. Each runs 1 hour 45 minutes, is tiered Higher and Foundation, carries 80 marks and is worth 50% of the GCSE. Each one contains multiple-choice items, short-answer questions and a dedicated Statistical Enquiry Cycle question. That mix is worth knowing before the first mock, because multiple-choice statistics rewards a different kind of preparation entirely: there is no method mark to fall back on, and a half-remembered definition costs the whole item.

The Statistical Enquiry Cycle question is the one where students most often lose marks for reasons that have nothing to do with statistics. It asks them to work through a problem the way a statistician would — pose the question, plan the collection, handle the data, analyse it, then conclude — and it rewards structure and carefully hedged, evidenced statements rather than a single computed answer. Students who have only ever practised calculation questions tend to answer it far too briefly.

The specification is also unusually prescriptive about vocabulary, which is a gift to anyone preparing for it properly. Correlation is marked against fixed numeric bands: values of 0.6 or above, or -0.6 or below, are treated as strong; 0.2 up to but not including 0.6, and the negative equivalent, are treated as weak; and values between but not including -0.2 and 0.2 count as no correlation. AQA adds that "use of the word 'moderate' for correlation will not be required". A student who learned that word in another subject will use it here and get nothing for it.

Two things the specification explicitly excludes are worth stating, because tutors from a university statistics background teach them by reflex. AQA's notes state that "formal use of null hypothesis will not be required", and separately that formal tests of significance will not be required. GCSE Statistics stops short of formal inference. Teaching a fifteen-year-old to run a significance test is not bad mathematics, but it is not on this paper, and the hours it takes are hours not spent on the sampling and enquiry-cycle work that is.

The version quoted throughout this section is AQA specification 1.2, dated June 2026. Specifications are revised more often than most parents expect, so the version stamp matters more than it looks; a resource written against an older version can be teaching content that has since moved or gone.

AQA or Edexcel for GCSE Statistics? Same syllabus, different exam room

Content is not the differentiator here. The Department for Education's subject content document states that it "sets out the knowledge, understanding and skills common to all specifications in GCSE statistics". Whichever board the school chose, the syllabus is the syllabus, and a resource built for one board teaches the same material as a resource built for the other.

The difference is time. Both qualifications are examined over 160 marks in two papers, but AQA allows 1 hour 45 minutes a paper and Pearson Edexcel allows 1 hour 30 minutes — half an hour less exam time overall for an identical mark total. This page used to call the two specifications close in structure. On content that is fair. On timing it is not, and pacing is precisely the thing that timed practice fixes.

In practice an Edexcel candidate needs to rehearse against the clock earlier and more often, and needs a firm personal rule about when to abandon a question and come back to it. An AQA candidate has more room, which only helps if they know they have it and do not rush the extended items. Either way the past papers have to come from the board the school actually entered: the same content, asked in a different house style, at a different speed.

Neither JCQ nor Ofqual publishes GCSE entries broken down by awarding body, so claims about which board most schools use for statistics cannot be sourced from published data. This page used to make one. It has been dropped rather than repeated, and the only reliable way to find out is to ask the school which specification it has entered. Beyond that, the work is the same shape as the rest of our subject tuition: diagnose against the specification, mark against published boundaries, rehearse under exam conditions.

Is a standalone A-level in Statistics worth taking?

It exists, it is small, and its grade profile is not the one families expect. JCQ recorded 1,140 UK entries for A-level Statistics across all ages in summer 2026, against 115,380 for A-level Mathematics — roughly 101 maths candidates for every statistics candidate. This is a genuinely niche qualification, and that is a neutral fact rather than a criticism.

The top-grade rates are the point. A* went to 1.6% of A-level Statistics entries and A*–A to 12.5%, against 17.7% and 42.5% in A-level Mathematics. At the other end of the scale, 92.7% of statistics entries were awarded A*–E. A qualification with a low A* rate and a high pass rate behaves like a subject that is accessible to enter and hard to top out in.

Cohort explains most of that too. A-level Mathematics attracts a self-selecting group of strong mathematicians in very large numbers; standalone A-level Statistics is more often taken as a fourth subject, or by students whose route into data is applied rather than pure. The 18-year-old band tells the same story as the all-ages band, which is a good sign that the pattern is real: 1,019 entries, 1.5% at A* and 12.5% at A*–A.

The route itself is Pearson Edexcel 9ST0. It is examined in three papers — Paper 1 Data and Probability, Paper 2 Statistical Inference and Paper 3 Statistics in Practice — each 2 hours, each 80 marks, each a third of the qualification, 240 marks in total. Paper 2 is where the formal inference that GCSE Statistics explicitly excludes finally arrives, which makes the step up from the GCSE larger than the shared subject name suggests.

An earlier version of this page named Edexcel 9ST0 in its comparison table and then told readers in its FAQ that standalone A-level Statistics was offered by some exam boards. Both now say the same thing. If a student is weighing this against other sixth-form options, the honest summary is that it suits someone who wants applied data work and does not need the pure mathematics, that the national cohort they will be measured against is very small, and that the universities on their list should be asked directly how they treat it.

How we plan statistics tuition

Everything above exists because it changes what a first session looks like. We start by establishing three things: which qualification the student is actually entered for, which board, and which year group they will sit it in. Those three answers fix the syllabus, the paper timings and the deadline, and they are misremembered often enough that we never take them on trust.

From there the work is diagnostic rather than generic. A GCSE Statistics student is marked against AQA or Pearson Edexcel past papers with the published boundaries applied, so a mock turns into a grade estimate rather than a bare percentage, and the gap to the target grade is expressed in marks on specific question types. An A-level student is placed against the right paper for their board: sections K to O on AQA 7357, Section A of Paper 3 on Edexcel 9MA0, component 02 on OCR H240 — including, on Edexcel, structured work with the large data set, which is examinable and reliably neglected.

Our tutors are subject specialists who teach these specifications regularly, including tutors who studied mathematics and statistics at leading UK universities. What matters more than any badge is that the person sitting with your child knows that "moderate" is not a permitted correlation description on AQA 8382, knows that the enquiry-cycle question is marked on structure, and knows which paper carries the statistics on their board.

We are equally clear about what tutoring cannot do. It cannot predict boundaries, it cannot turn a Year 10 entry into a Year 11 entry, and it does not replace the school's own scheme of work. What it can do is close the distance between what a student currently knows and what the specification asks of them, on the timetable the school has already set.

Frequently Asked Questions