How much does tutoring actually help?

The headline number, and what it means

The Education Endowment Foundation's Teaching and Learning Toolkit, which synthesises evidence across education interventions, reports that one-to-one tuition delivers approximately five additional months' progress on average compared with similar pupils who did not receive it.

"Months of progress" is a translation of effect size into a unit that school leaders can reason about. It is an average across many studies, not a promise about a particular child. Averages of this kind conceal wide variation: some studies find much larger effects, some find close to none, and the difference usually lies in how the tuition was delivered rather than whether it happened.

Age changes the answer

The same evidence finds that studies in primary schools tend to show greater impact — around +6 months — than those in secondary schools, at around +4 months. That gap is worth taking seriously when deciding where to spend money.

A plausible reading is that primary-age gaps are narrower and more tractable: a specific missing skill can be identified and rebuilt. By secondary, difficulties are more often cumulative, entangled with motivation, and harder to resolve inside a weekly hour. It does not mean secondary tutoring fails — +4 months is a large effect by the standards of educational interventions — but it argues against expecting a transformation late in a course.

Short, frequent, time-limited beats long and open-ended

This is the finding most commonly ignored in practice. The EEF reports optimum impact from short, regular sessions of about 30 minutes, three to five times a week, sustained over a set period of up to ten weeks.

Compare that with the standard commercial arrangement: one 60-minute session per week, continuing indefinitely. That pattern is convenient for scheduling and for recurring revenue, and it is close to the opposite of what the evidence recommends on every dimension — longer sessions, less frequent, with no defined endpoint.

You will not always be able to buy the evidence-based pattern; few private tutors sell 30-minute slots four times a week. But you can move toward it: shorter and more frequent where possible, and above all a defined block with a stated goal rather than a standing appointment that renews by inertia.

• Ask for a fixed block — eight to ten weeks — with a review at the end
• Prefer two shorter sessions a week over one long one where the budget is equal
• Agree in advance what "this worked" will look like
• Treat indefinite continuation as a decision, not a default

The two-sigma claim, and why it is not a promise

Discussions of tutoring frequently invoke Benjamin Bloom's "2 sigma problem". In 1984 Bloom reported that students taught one-to-one using mastery learning performed around two standard deviations better than conventionally taught peers — in his words, "the average tutored student was above 98% of the students in the control class", with roughly 90% of tutored students reaching levels matched by only the top 20% of untutored ones.

Two standard deviations is an enormous effect, and it is routinely quoted in tutoring marketing without its conditions. The result did not come from one-to-one attention alone: it depended on mastery learning, in which students must reach a high threshold — Bloom used around 90% on assessments — on prerequisite material before advancing, with corrective support until they do. Almost no commercial tutoring arrangement is structured that way.

Treat two sigma as a demonstration of what individualised instruction can achieve under a tightly specified method, not as a forecast for a weekly session. If a tutor cites it at you, the reasonable follow-up is to ask which part of their own practice implements mastery learning.

What the evidence says about who benefits

Tutoring is disproportionately purchased by households that can afford it, and this is a documented equity problem rather than an incidental one. Research on "shadow education" — the global market of private supplementary tutoring — finds access correlates strongly with family wealth, and a 2012 Asian Development Bank and University of Hong Kong study warned that private tutoring can "maintain and exacerbate social inequalities" and divert household income into an unregulated industry.

That literature also documents the scale involved: roughly 90% of elementary students in South Korea and about 85% of senior secondary students in Hong Kong participate in some form of shadow education, with national markets valued in the billions — around $17.3 billion in South Korea in 2010 and approximately $70 billion in India by 2017.

The relevance for an individual family is modest but real: you are buying into a large, lightly regulated market where the quality distribution is wide and price is a poor proxy for effectiveness. That is precisely why verification of the individual tutor matters more here than in sectors with professional licensing.