Why One-on-One Tutoring Outperforms Group Classes: The Evidence for Personalised Learning

In 1984, educational psychologist Benjamin Bloom published a paper that has been quietly reshaping the conversation about education ever since. The paper, titled “The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring,” reported something that surprised even its author. Students receiving one-on-one tutoring consistently performed two standard deviations above students receiving conventional classroom instruction on the same content. In practical terms, this means that the average tutored student outperformed approximately 98% of students who received only classroom teaching.

Bloom called it a problem because the implication was uncomfortable: if tutoring is this significantly more effective than classroom instruction, and if classroom instruction is what most students receive, then most students are learning in suboptimal conditions. The practical and economic challenges of providing every student with individual instruction mean that the gap between what is possible and what is delivered has remained wide for four decades. But for parents and students making individual decisions, the findings are instructive and actionable.

What Exactly Is the 2-Sigma Effect and Why Does It Matter?

A two standard deviation advantage is a large effect by any measure in educational research. To contextualise it: a student who scores at the 50th percentile in a group receiving conventional instruction would, if they received one-on-one tutoring, be expected to perform at the 98th percentile of that same population. Bloom’s results have been replicated and extended by subsequent researchers, though the effect size found in later studies typically ranges from 0.4 to 2.0 standard deviations depending on methodology, subject, and the quality of the tutoring.

Even the lower end of this range is substantial. An effect size of 0.4 in educational research is considered large. The average effect of class size reduction, a perennially debated and expensive educational intervention, is approximately 0.2. Private tutoring, even when accounting for variation in quality, consistently produces larger effects than almost any other educational intervention studied.

The Core Mechanism: Adaptive Instruction

The most important reason tutoring outperforms group instruction is not that tutors are better teachers than classroom teachers. Many classroom teachers are extraordinary. The difference is structural: a classroom teacher with 25 students cannot adapt the pace, depth and direction of a lesson to match each student’s current understanding. A tutor working with one student can, and does, in every session.

In a classroom, a teacher makes a single instructional decision: this is the pace and level at which I will teach this topic today. Students who are ahead of that pace are unchallenged and disengage. Students who are behind that pace fall further behind. Students who are exactly at that pace learn well. In a typical classroom of mixed ability, the proportion of students in that third category is often surprisingly small.

A tutor, working with a single student, makes continuous instructional decisions in real time. If a student understands a concept faster than expected, the tutor accelerates, deepens, or connects it to the next topic. If a student struggles, the tutor slows, tries a different explanation, breaks the concept into smaller steps, or diagnoses the underlying misconception before it compounds. This responsiveness is the core mechanism of the tutoring advantage and it cannot be replicated in a group setting.

Immediate and Specific Feedback

A second major mechanism is the immediacy and specificity of feedback in one-on-one instruction. In a classroom, a student who makes an error in written work may not receive feedback for days. When the corrected work returns, the student may no longer clearly recall their reasoning at the moment of writing. The feedback therefore arrives too late to connect productively with the original error.

In a tutoring session, errors are identified and addressed in real time. When a student applies an incorrect formula, misinterprets a question, or makes a logical error in an argument, the tutor responds immediately. This immediate feedback loop is far more effective at correcting errors and reinforcing correct understanding than delayed written marking. The student still remembers their thinking when the correction is given, which makes it possible to address the root cause of the error rather than simply noting that it was wrong.

Bloom’s own analysis identified feedback and corrective intervention as one of the two primary variables (alongside student time on task) that explained the tutoring advantage. Research since then has extended this picture: not just that feedback is given, but that feedback is acted upon within the same learning session, is the critical element. Tutoring sessions naturally create this cycle in a way that group instruction structurally cannot.

Addressing Gaps Before They Compound

Mathematics education offers a clear illustration of a broader principle: much of what students struggle with in secondary school is not a current-year problem but the accumulated result of gaps in earlier years that were never closed. A student who does not fully grasp fractions in primary school will struggle with algebra. A student with gaps in basic algebra will struggle with calculus. The difficulty appears to be with the current content; the root cause is earlier, unclosed.

Classroom teaching cannot efficiently diagnose and address these foundational gaps because it must progress through the current year’s curriculum. A tutor can. In the early sessions of a tutoring relationship, a skilled tutor will identify where a student’s actual understanding begins and ends, often revealing that the apparent difficulty with year-eleven content is rooted in year-eight understanding that was never secure. Addressing that root gap produces rapid improvement not just in the current topic but in the student’s general confidence with the subject.

This diagnostic work is among the most valuable things a tutor provides, and it is genuinely invisible in group instruction. Students often believe they are “bad at maths” or “not a language person” when in reality they are dealing with specific, identifiable, addressable gaps. The attribution of difficulty to fixed ability, rather than to specific missing knowledge, is one of the most damaging beliefs a student can hold about their academic potential. Effective tutoring frequently dismantles this belief by demonstrating, concretely, that the gaps can be closed.

Motivation and the Relationship Factor

Bloom’s original framing of the tutoring advantage focused primarily on cognitive variables: feedback, pacing, time on task. Subsequent research has given increasing attention to the motivational and relational dimensions of tutoring, which are not separate from the learning outcome but integral to it.

Students who work with a tutor they trust are more willing to ask questions they would be embarrassed to ask in a classroom. They are more honest about what they do not understand. They are more likely to persist with difficult problems rather than giving up. These behaviours matter enormously for learning outcomes and they are enabled by the private, supportive nature of one-on-one sessions in a way that classroom dynamics cannot consistently replicate.

The relationship that develops over a sustained tutoring programme also enables a kind of long-term goal alignment that is difficult to achieve in any other educational context. A tutor who has worked with a student for six months knows not just their academic profile but their particular motivational triggers, their tendency to rush or to overthink, the contexts in which their confidence is stronger or weaker. This accumulated knowledge allows increasingly precise and effective intervention.

Is Tutoring Right for Every Student?

Tutoring is not the right response to every educational challenge. For students who are performing well, finding their classroom instruction engaging and making good progress, the marginal benefit of tutoring is lower. The resource is most valuable where there is a specific gap to close, a high-stakes examination to prepare for, or a motivational challenge that one-on-one attention can address.

The quality of tutoring also matters enormously. The research showing large tutoring effects almost uniformly involves skilled, trained tutors with genuine subject knowledge and pedagogical awareness. An untrained university student reading from the same textbook a student already has at home is unlikely to replicate a two-sigma effect. The characteristics of effective tutoring described throughout this article, diagnostic assessment, adaptive pacing, immediate feedback, relationship quality, and genuine expertise, are not automatically present simply because the instruction is one-on-one. They require a deliberate, professional approach.

Making the Most of Your Sessions

Students and parents who understand what makes tutoring effective can contribute to those conditions. Come to each session with specific questions about content you found difficult since the last time. Attempt problems independently before the session so that the tutor can observe your approach and identify where it breaks down. Do the work assigned between sessions. Communicate honestly when something is still not clear rather than nodding in acknowledgement. The tutoring relationship is a collaboration, and students who participate actively in it extract significantly more value than those who attend passively.

At MP STUDIO, we build these principles into every programme. We begin with a diagnostic assessment, design sessions around the specific gaps that assessment reveals, and review progress regularly. Our goal is not to be present at every lesson forever but to give students the understanding and confidence to work independently, with our support in the background as and when it is needed. That is what effective tutoring looks like in practice.