Maths Tutor: What Actually Moves a Maths Grade in 2026

The Education Endowment Foundation rates one-to-one tuition at +5 months’ additional progress. Behind that average, maths is the weakest subject in the table at +2 months, against +6 for literacy. Maths is where generic tutoring helps least and where an accurate diagnosis helps most. This page sets out what our maths tutors do in a session, how we pick them, and how to start.

What One-to-One Tuition Actually Moves in Maths — and What It Does Not

Most tutoring pages quote the Education Endowment Foundation’s headline figure and stop there. The EEF’s Teaching and Learning Toolkit rates one-to-one tuition at +5 months’ additional progress on average, drawn from 123 studies, with the strand’s review last updated in July 2021. Read only that line and tutoring looks like a straightforward purchase.

The Toolkit’s own “behind the average” breakdown is less flattering, and it is the part that matters if the subject is maths.

Read the maths row against the headline row. A parent arranging maths tuition for a Year 10 is buying into the two weakest cells in that table at the same time — the subject at +2 and the age group at +4, against a +5 average that pools every subject and every age group together. That is the figure the sector quotes and the figure it rarely breaks down.

We publish that because it changes what good maths tutoring has to look like, not because it argues against it. The Toolkit is explicit that a large share of the studies behind the average were not independently evaluated, and that there is a large amount of unexplained variation between results — which is why the strand is rated moderate rather than secure. An average that pools school-run interventions, including ones delivered by teaching assistants and volunteers working to a structured programme, is not a prediction about a specialist maths teacher working one-to-one with your child. It is a warning that the format alone guarantees nothing.

What the Toolkit does say about the conditions is specific, and we build around it:

• Short and regular beats long and occasional. The EEF puts optimum impact at sessions of about 30 minutes, three to five times a week, over a defined period of up to ten weeks. Most families cannot run that cadence with a private tutor, so the practical translation is a fixed weekly session plus short, specified between-session work — not a two-hour block once a fortnight whenever it suits the diary.
• Tuition has to be linked to the classroom, not parallel to it. The EEF finds impact is likelier when tuition is additional to and explicitly linked with normal lessons. In practice that means the tutor works from your child’s actual scheme of work and actual mock paper, not a generic syllabus.
• Someone has to be monitoring whether it is working. The EEF lists accurately identifying the gap, choosing content against that gap, and monitoring impact and adjusting as the active ingredients. A tutor who cannot tell you in week six what has changed since week one is not running the thing the evidence describes.

One maths-specific data point from the EEF’s own trial programme is worth having alongside the Toolkit average: the second Tutor Trust “Affordable Tutoring” trial recorded +3 months in mathematics under everyday conditions — above the Toolkit’s maths figure, below its headline. That is the honest band to expect from a structured programme, and the reason the rest of this page is about method rather than adjectives.

Source: EEF Teaching and Learning Toolkit, “One to one tuition” strand, and the linked Tutor Trust trial record; retrieved 16 October 2026.

Why a Maths Report Grade Can Look Fine While the Gaps Grow

The question we are asked most often is some version of: the reports have been fine for two years, so why has the mock come back at a 4? Ofsted named the mechanism in its maths subject report, Coordinating mathematical success, published on 13 July 2023 after routine inspections across primary and secondary schools in England. Three of its secondary findings explain almost every case we see.

Internal assessment can give false assurance. Ofsted found that applying GCSE grade thresholds to a school’s own internal assessments gives “false assurance” about what pupils have learned, and that as a consequence some pupils progress through the maths curriculum with significant and growing gaps in knowledge. A 5 on an internal test built from recently taught content is not the same measurement as a 5 on a paper that samples three years of it.

Content gets covered, not learned. Ofsted found that pupils learning maths more slowly than their peers are often rushed through new content in order to complete the course, and that this frequently results in Key Stage 4 repeating content pupils had already studied — but not learned — in Key Stage 3, and sometimes Key Stage 2. If your child has met a topic three times and still cannot do it, the issue is usually not that they need a fourth exposure to the same explanation.

Methods arrive disconnected. Where curriculum planning is weaker, Ofsted found it takes less account of what pupils have learned before and what they will study later, and pupils end up taught a series of disconnected methods and “tricks” that work only in specific circumstances. That is why a student can answer a question phrased the way their teacher phrased it and freeze when the same mathematics arrives inside a worded context.

None of this is a criticism of a particular school. Ofsted is equally clear that maths education in England is broadly healthy and has improved, and that the schools most affected are those struggling to recruit and retain specialist maths teachers. But it does mean a parent cannot read a run of reassuring grades as evidence that nothing is missing, and it is the reason our first session is a diagnosis rather than a lesson. If that is the position you are in, book a free consultation and we will tell you what a maths specialist would look at first.

The First Session: Finding Where the Chain Actually Broke

Ofsted’s maths research review splits mathematical knowledge three ways, and the split is the most useful diagnostic tool we have: declarative knowledge (facts, concepts and formulae), procedural knowledge (methods, procedures and algorithms), and conditional knowledge (the strategies that combine facts and methods to reason and solve problems).

Almost every parent describes the problem in one sentence — “she can’t do algebra” — and that sentence can mean any of the three. They have different fixes, and the wrong fix wastes a term:

• A declarative gap looks like hesitation on things that should be instant: times tables, square numbers, the value of a negative exponent. The fix is spaced retrieval of a short, specified fact set, not re-teaching.
• A procedural gap looks like a method half-remembered and applied inconsistently. The fix is a fully worked example followed by faded practice — the tutor does step one, then the student does more of each successive question until none of it is scaffolded.
• A conditional gap is the one most often missed, because the student can do every step correctly once told which step to use. The fix is not more practice. It is sorting problems by type until the student can say what kind of question it is before touching it.

What a backward trace looks like in practice. Take a typical case. A Year 10 arrives because she “can’t do simultaneous equations”. For the first twenty minutes a good tutor will usually not teach simultaneous equations at all. They will ask her to solve 3x + 4 = 7, then 3x + 4 = 2x + 7, then ask what happens when you subtract one true statement from another. If the first is fluent and the second is not, the gap is not simultaneous equations: she is reading the equals sign as “and the answer is” rather than as a balance, which is a Key Stage 2 idea carried forward for five years. Teaching elimination on top of that produces correct answers in the lesson and nothing three weeks later. Spending one session on the balance idea fixes simultaneous equations, rearranging formulae and iteration at the same time, because all three sit on it.

This is also what the research recommends. The EEF’s guidance report Improving Mathematics in Key Stages 2 and 3, published on 3 November 2017, makes “use assessment to build upon pupils’ existing knowledge and understanding” the first of its eight recommendations, and says teachers must not only address misconceptions but understand why pupils persist with errors — knowledge of common misconceptions being invaluable in planning. A first session that opens a textbook at the topic the parent named has skipped the recommendation entirely.

The Misconceptions We See Most — and What Fixing Each One Involves

Many students arrive believing they are simply “not a maths person”. In almost every case what they have is a specific, nameable misconception nobody has addressed. These six come up most often:

• Negative numbers. Sign errors when multiplying or subtracting negatives, especially under time pressure. Fixed by separating the two rules that get conflated — the rule for operating on negatives and the rule for two signs meeting — and drilling them apart before putting them back together.
• Fractions and percentages. Confusing percentage change with percentage of, or mishandling mixed numbers. Fixed by forcing every percentage question through a single multiplier representation, so reverse percentages stop being a separate topic to remember.
• Algebra. Expanding double brackets incorrectly, and treating a letter as a label for an object rather than a value. The second one is the deeper problem and the harder to spot, because it produces right answers until the question needs the letter to vary.
• Ratio and proportion. Applying additive thinking to a multiplicative problem — one of the most common errors in GCSE papers, and almost always invisible to the student. Fixed by making them state, before calculating, whether the relationship is “add the same” or “times the same”.
• Trigonometry. Knowing the formulae but not which applies, particularly once triangles stop being right-angled. This is a conditional-knowledge gap, so the work is question sorting, not formula practice.
• Showing working. Jumping to an answer with nothing written down, which costs method marks even when the answer is right. Examiners award marks for correct mathematical steps; a blank page with a wrong answer earns none of them. This is pure technique and usually the fastest marks on the page to recover.

A skilled tutor spots these quickly. Rather than correcting the answer, they ask the question that exposes the belief underneath it and replace it with something that holds.

What a Term of Maths Tutoring Is Actually Structured Like

A programme that matches the conditions the EEF describes looks roughly like this over a twelve-week term. The shape changes with the student; the sequence does not.

• Week one — diagnosis. A backward trace as described above, plus one full past paper from the board and tier your child is actually entered for, marked question by question against the mark scheme rather than totalled. The output is a written list of named gaps, sorted into declarative, procedural and conditional.
• Weeks two to five — prerequisites. The earliest gaps first, because everything later rests on them. This is the stage that feels counter-intuitive to parents, since a Year 11 may spend a session on fraction arithmetic. It is also the stage that makes the rest work.
• Weeks six to nine — current content, linked to school. The tutor works from the class scheme of work so sessions reinforce rather than compete with lessons. This is the EEF’s “explicitly linked” condition, and it is the one most commonly dropped by tutors working to their own plan.
• Weeks ten to twelve — exam technique under timed conditions. Whole papers to time, then a line-by-line review of where marks were lost: method not shown, question misread, time mismanaged, or genuinely not known. Only the last of those is a maths problem, and it is usually the smallest category.
• Throughout — re-measurement. The week-one paper is re-sat late in the programme. Parents get the comparison, not an adjective.

Between sessions, work is short and specified: two or three named question types, not “do some revision”. The EEF’s cadence finding is the reason — frequency and regularity carry more of the effect than session length does.

How We Choose Maths Tutors — and What to Check If You Hire One Yourself

We accept about 3% of applicants, and the stage that decides it is not the subject test. It is the trial lesson: a supervised live session with a real student, assessed on pacing, rapport, explanation quality and confidence. A maths graduate who cannot hear the misconception underneath a wrong answer is no use to a struggling student, and that only shows up when you watch them teach. Every tutor then has an enhanced DBS certificate verified and two references checked independently before any placement, and lesson recordings and parent feedback are reviewed after they join.

If you are considering hiring a tutor privately instead, three things from the Disclosure and Barring Service’s own guidance are worth knowing, because they are not obvious and they are where most parents are left exposed.

• You cannot request a DBS check on a self-employed tutor yourself. Government guidance is explicit: where you hire a self-employed person to deliver services to your family, you cannot request the check. They have to request it themselves, and you have to check their certificate. A marketplace listing that says “DBS checked” is a claim about a document you have not seen.
• There are four levels, and the word “DBS” on its own tells you nothing about which one. For one-to-one work with a child, only the fourth row below answers the question parents are actually asking.
• A DBS check has no official expiry date. That is the government’s own wording. A certificate is accurate on the day it was issued and says nothing about any day since. If a tutor is signed up to the DBS Update Service you can confirm online that the certificate is still current — which is the only version of “DBS checked” that carries real information.

So the three questions worth asking any private maths tutor are: may I see the certificate, what level is it and when was it issued, and are you on the Update Service. A tutor who is used to working properly will not find any of that awkward.

Source: GOV.UK, “Check someone’s criminal record” — types of check, when to repeat a check, and hiring a self-employed person; retrieved 16 October 2026.

What Changes Beyond the Mark

The first thing that shifts is usually not the grade. It is that a student stops treating maths as a subject that happens to them. When someone finally explains why a method works rather than only how to execute it, the subject becomes something they can reason about, and that is what makes the next unfamiliar question survivable.

In practical terms, tutoring closes named topic gaps, builds the habit of writing a method down, and puts exam-style questions in front of a student often enough that the format stops costing marks on its own. The marks that come back first are usually technique marks — method written down, question read properly, paper finished — because those are recoverable in weeks where missing content is not.

Exam anxiety falls for an unglamorous reason: familiarity. A question type seen many times in a low-stakes setting is not threatening when it appears in an exam hall. We are rated 4.8 out of 5 on Trustpilot across 59 reviews, and the pattern in them is consistently about a specific tutor and a specific child rather than a system.

Where to Find the Specification Detail for Your Child’s Board

This page deliberately does not carry paper structures, topic weightings, tier decision tables or grade boundary marks. Boundaries in particular are reset after every exam series, and a figure published on a service page goes stale the first August after it is written. We keep all of it on four pages that are maintained against the awarding bodies’ own documents, so there is one current copy rather than several drifting ones:

• GCSE Maths — the qualification itself: which of the five GCSE Maths qualifications your child is on, which paper is non-calculator, the total mark, how the 9 to 1 scale works, what is on the formulae sheet, and what happens if a student does not reach grade 4.
• AQA GCSE Maths (8300) — AQA’s own June 2026 grade boundaries, the 8300F and 8300H entry codes, which formulae AQA gives and which must be memorised, and AQA’s published key dates for the next series.
• Edexcel GCSE Maths (1MA1) — assessment objective and topic weightings, the full topic-weighting table, Pearson’s June 2026 subject and notional component boundaries, and where a 1MA1 grade leads.
• OCR GCSE Maths (J560) — assessment objective weightings by tier, J560 boundaries, the three content layers, and OCR’s Check In diagnostic tests.

If you are not sure which board your child sits, it is printed on their exam timetable and on the front of every past paper their school sets. Our tutors confirm it before the first session either way, because the differences are large enough to change what a programme prioritises. Broader stage guides sit on our GCSE tutoring and A-Level tuition pages. If you already know the board and want a maths tutor who teaches it, book a free consultation and we will tell you what the first four weeks would cover.

Frequently Asked Questions

Maths is a subject where the right support at the right time genuinely changes outcomes. Whether your child needs to rebuild the foundations or sharpen technique before final exams, a tutor who can hear the misconception underneath a wrong answer is what makes the difference.