Tutoring is sold with big promises and bigger numbers, so it's worth asking plainly: does it work, how much, and why? The honest answer is genuinely encouraging — one-to-one tutoring is among the best-evidenced things in all of education — but only if you understand what it actually does and what separates effective tutoring from expensive homework supervision. As a tutor, I'd rather you knew the real research than the marketing.
A parent sits with a spreadsheet of tutoring options, every one promising to "transform" their child, several quoting dazzling statistics. Underneath the noise is a simple, fair question that deserves a straight answer: will this actually help, and is it worth it? Having taught maths one-to-one for years, I think the most useful thing I can do is not add to the sales pitch — it's to lay out exactly what the research shows, including the honest caveats, so you can tell the genuine article from the expensive babysitting.
This is the final piece in our series on the research behind maths learning, and in a way it ties the rest together — because, as we'll see, good tutoring is essentially a delivery vehicle for everything the other articles describe. It draws directly on our pieces on curriculum pace, feedback, anxiety and motivation.
Strip away the brochures and the question is always some version of: is this money and time going to make a real difference? It's the right question, and it deserves more than a testimonial. The research can answer it — but only if we're honest about three things at once: that tutoring genuinely is highly effective on average, that the headline "miracle" numbers are overstated, and that the benefit depends enormously on the kind of tutoring, not just the fact of it.
That honesty matters because "tutoring" covers a vast range — from a trained specialist running a structured, diagnostic programme to an older student passively watching a child do their homework. These are not the same product, and they do not produce the same results. So the useful version of the question isn't "does tutoring work?" but "what does effective tutoring do, and how would I recognise it?"
Here is the insight that demystifies the whole thing.
People often imagine tutoring works through some special ingredient a classroom lacks. It doesn't. A good tutor uses the same principles every good teacher knows. What tutoring changes is the delivery: with one child instead of thirty, all those principles can finally be applied fully, simultaneously, and precisely to this learner. Tutoring is what happens when you stop having to average across a class. The pace can bend to one child; the feedback can be instant and personal; the exact gap can be found and fixed; nothing has to move on before it's secure.
A classroom is a bus on a fixed route at a fixed speed: efficient for getting thirty people roughly where they're going, but it can't stop at your door, and if you miss a stop it drives on regardless. Tutoring is a taxi. It goes at your pace, takes your route, waits when you're stuck, and doubles back for the turning you missed. The taxi isn't a magical vehicle — it's an ordinary car shaped entirely around one passenger. That's the whole secret: not a different kind of teaching, but teaching that no longer has to serve thirty people at once.
This reframing is liberating and clarifying. It tells you that tutoring works to the extent that it actually does these things — bends to the pace, diagnoses, gives feedback, secures mastery, builds the relationship. A "tutor" who does none of that, who just sits beside a child re-doing the homework, has the taxi but never turns the engine on. The format is necessary; using it well is what makes it work.
Tutoring has one of the strongest and longest evidence bases in education. Here's the honest picture.
The legend starts with Bloom (1984), whose "two sigma" finding suggested tutored students could outperform around 98% of conventionally taught ones. It's an inspiring result — but best treated as aspirational. More careful reviews since, notably VanLehn (2011), put the average effect of skilled human tutoring at around 0.8 of a standard deviation — a large, genuinely meaningful improvement, but not the mythic two sigma. The realistic number is still excellent; it just doesn't need inflating.
The most comprehensive recent review, by Nickow, Oreopoulos and Quan (2020), pooled the experimental evidence and found tutoring delivers consistent, substantial gains — with an overall effect of roughly 0.37 of a standard deviation across programmes. Tellingly for families, it found maths tutoring tends to be especially effective in the later grades, that trained tutors outperformed untrained ones, and that regular, in-school, well-structured programmes did best. Quality and consistency, not just contact, drive the result.
This isn't a new or fragile finding. Cohen, Kulik and Kulik (1982), reviewing 65 studies, found tutoring improved both achievement and attitudes (and helped the tutors too). And the UK's Education Endowment Foundation, which weighs the hard evidence, rates one-to-one tuition as worth around five additional months' progress, with small-group tuition close behind. Across forty years and very different methods, the verdict is steady: well-delivered tutoring reliably helps.
One honest, useful nuance ties to our piece on technology: VanLehn also found that sophisticated computer tutors came surprisingly close to human ones. This isn't bad news for human tutoring — it shows the mechanisms (responsiveness, feedback, the right next step) are doing the heavy lifting, and that a great tutor combines those mechanisms with a human relationship that still gives people the edge, especially for motivation and confidence.
Pull the research together and the "why" is simply the rest of this series, finally deliverable to one child at once.
It bends to the child's pace and secures mastery. A class moves at one speed; a tutor moves at the child's, and doesn't advance until each idea is solid. That directly fixes the pace problem — the single biggest hidden cause of maths struggle — by finding and repairing the missing foundation before building on top of it.
It gives immediate, personalised feedback and diagnosis. A tutor sees exactly where the thinking goes wrong, in real time, and responds to this error — the kind of high-quality, task-focused feedback that moves the needle and that a teacher simply cannot give thirty children at once.
It lowers the fear and rebuilds confidence. One-to-one removes the audience and the clock, the classic triggers of maths anxiety and test anxiety. A calm, private space where mistakes are safe lets productive struggle happen without panic — and steady, visible success rebuilds the competence that drives lasting motivation.
It is built on a relationship. The "relatedness" that sustains motivation, and the trust that lets a child admit "I don't get this," come from a consistent, caring tutor who knows them. That human connection is the part a computer can approximate but not replace — and it's often what turns a reluctant student around.
Tutoring is a global phenomenon, and how countries use it is revealing — including the cautions. Tap through five.
Private tutoring has grown markedly in the UK, and the National Tutoring Programme made structured tutoring a national policy after the pandemic, grounded in the EEF's strong evidence for one-to-one and small-group tuition. The UK case shows tutoring moving from a private luxury toward an evidence-based, mainstream intervention — with quality and tutor training increasingly recognised as what makes or breaks it.
The United States invested heavily in "high-dosage tutoring" — frequent, sustained, small-group or one-to-one sessions — as a leading response to pandemic learning loss, precisely because the evidence (Nickow and colleagues) is so strong for regular, structured programmes. The American experience underlines the research's central caveat: dosage and structure matter; occasional, ad-hoc tutoring delivers far less.
Several East Asian societies have enormous "shadow education" sectors — cram schools and private tutoring at vast scale. The results are mixed and the equity concerns real: heavy tutoring can raise scores but also intensify pressure and widen gaps between families who can and can't afford it. It's a reminder that tutoring is a powerful tool whose social effects depend on how, and for whom, it's used.
Finland leans on strong in-school support and early intervention rather than a large private tutoring market — effectively trying to deliver the benefits of individual attention within the school day. It's a useful counterpoint: the goal isn't tutoring for its own sake, but the personalised, gap-closing attention tutoring provides, which a well-resourced system can partly supply itself.
Across countries, two themes recur. First, the evidence for well-structured tutoring is robust and international. Second, equity is the recurring worry — tutoring can help any child, but if only some can access it, it risks widening gaps. That's a strong argument for free, high-quality practice resources and assessments alongside paid tutoring, so good support isn't gated purely by budget.
The global verdict matches the research: structured, sustained, quality tutoring reliably helps — and the open question is less "does it work?" than "how do we make sure good support reaches the children who need it?" For an individual family, the practical task is simpler: recognise effective tutoring when you see it.
The research is clear that quality is everything, so here's how to tell the taxi-with-the-engine-on from the parked car.
Here's the straight version, from someone who does this for a living: good tutoring is genuinely one of the most effective things you can spend on for a struggling or stuck student — and poor tutoring is one of the easiest ways to waste money and time. The difference isn't the price or the format; it's whether the tutor diagnoses, plans, makes the child do the thinking, and builds confidence. Before committing, ask a prospective tutor how they'll find your child's gaps and what their plan is. A good one will have a clear, specific answer. And there's no shame in starting with a free assessment to see whether tutoring is even the right tool for your child's particular problem — sometimes the honest answer is that a few targeted fixes at home will do.
The tutoring evidence has lessons for schools too, especially in how they deploy extra support.
Favour structured, sustained, small-group or one-to-one support. The research strongly favours regular, well-organised "high-dosage" tutoring over occasional, ad-hoc help. Schools investing in tutoring get the best return when it's frequent, planned, and delivered by trained tutors working to clear goals.
Use diagnosis to target it. The biggest gains come from finding and fixing specific gaps, so pairing tutoring with good diagnostic assessment — knowing exactly which foundations to repair — multiplies its effect. Tutoring aimed at a precise gap beats tutoring that simply adds general "more maths."
Mind the equity gap. Because privately purchased tutoring can widen advantages, schools and systems that provide structured tutoring and strong free practice resources help ensure that the powerful benefits of individual attention aren't reserved for those who can pay. It's the same equity logic running through the pace and home environment articles.
Whether or not you ever choose tutoring, our practice portal is free: diagnostic, self-paced, with instant feedback and worked solutions. It's our small answer to the equity question — quality maths support that isn't gated behind a budget.
Open the practice portal →"Any tutoring will transform a child — the format is what matters."
Quality and structure decide the outcome. Regular, diagnostic, well-planned tutoring works strongly; passive homework supervision does little. The format is necessary but nowhere near sufficient.
"Tutoring gives a guaranteed 'two sigma' miracle."
The two-sigma figure is aspirational. Realistic estimates put skilled human tutoring around 0.8 of a standard deviation — large and worthwhile, but honest providers quote the real numbers.
"Online tutoring can't match in-person."
The mechanisms that make tutoring work transfer online when it's done well; even computer tutors come close to human ones. A skilled online tutor adds the human relationship on top.
I'll end the series where an honest tutor should: with the evidence rather than the pitch. Tutoring earns its strong reputation — the research, across forty years and many methods, agrees it reliably helps, often a great deal. But it earns that reputation only when it does the real work: finding the gaps, bending to the child's pace, giving feedback, securing each idea, and rebuilding confidence within a relationship the child trusts. That's not a secret formula; it's everything the rest of this series describes, finally delivered to one learner at a time. If you take one thing from all twenty articles, let it be this: the things that help children learn maths are knowable, and you now know them. Whether through a great teacher, a good tutor, or a calmer kitchen table, the levers are real — and they're within reach.
Well-structured tutoring is one of the most evidence-backed interventions in education. A major meta-analysis found it produces consistent, substantial gains, and the EEF rates one-to-one tuition at around five months' extra progress. The caveat: gains depend heavily on quality — regular, structured, diagnostic tutoring by a capable tutor works far better than occasional, passive homework help.
Because a tutor can do what a teacher with thirty children cannot: move at exactly your child's pace, diagnose the precise gap, give immediate personalised feedback, ensure each idea is secure before moving on, and remove the audience and clock that fuel anxiety. Tutoring isn't a different magic — it's a delivery system for everything we know helps, focused on one child.
That figure comes from Bloom's famous 1984 result and is best treated as aspirational. More recent reviews put the average effect of human tutoring at around 0.8 of a standard deviation — still a large, meaningful improvement, just not the mythic two sigma. Honest providers quote the realistic numbers; the genuine benefit is substantial without exaggeration.
Effective tutoring is regular and sustained (not a one-off cram), diagnostic (it finds the specific gaps rather than re-doing tonight's homework), structured around a plan, active (the child does the thinking, with feedback), and built on a good relationship. Ineffective tutoring is passive homework supervision with no diagnosis or plan. The format alone guarantees nothing; how it's used is everything.
Yes, when done well. The mechanisms that make tutoring work — one pace, immediate feedback, diagnosis, a strong relationship, mastery before moving on — all transfer to a well-run online session. Research on intelligent tutoring systems even found sophisticated computer tutors nearly as effective as human ones; a skilled online tutor combines that responsiveness with the human relationship that still gives people the edge.
The free assessment is an honest place to start: we diagnose where the real gaps are, tell you plainly whether tutoring is the right tool, and leave you with a clear plan either way. No pressure, no obligation.
Book the free assessment