A classroom of thirty-two pupils becomes a classroom of twenty-four. The room is quieter. The teacher can see more faces. There are fewer exercise books in the pile. Parents assume the smaller class must now produce better learning.
But what, exactly, changed?
If the teacher gives the same explanation, asks the same three volunteers, marks work with the same delay, groups pupils in the same way and uses the same lesson sequence, the arithmetic of the room changed more than the teaching did. Eight fewer pupils may make the class feel easier to manage, yet the learning mechanism remains almost untouched. Class size matters when the reduced number of learners changes what teachers and pupils can actually do.
That distinction is unusually important because class size is one of education’s most intuitive variables. Parents can count it. Schools can advertise it. Governments can regulate it. Teachers can feel it. Yet a visible input is not the same thing as an instructional mechanism. The Education Endowment Foundation’s current Teaching and Learning Toolkit estimates a small average positive effect—about one month of additional progress—and rates the evidence base as very limited. It also stresses that smaller classes are most likely to matter when teachers use the smaller group to increase interaction, feedback, flexibility and productive learning behaviour. OECD data likewise show that class size and student–teacher ratio are not the same statistic and vary substantially across education systems.
This article therefore owns one narrow job on eduKateSG: how class size can change learning conditions when fewer learners alter attention, visibility, feedback, participation, grouping, behaviour and instructional decisions. It does not own one-to-one tuition, small-group tuition, teaching-assistant deployment, individualised instruction or school staffing policy. Those neighbouring systems sometimes reduce the number of learners one adult handles, but they solve different problems.
The useful question is not, “Is twenty better than thirty?” It is: what becomes possible at twenty that the teacher could not reliably do at thirty, and is the lesson redesigned to use that extra capacity?
The 50-second answer
Smaller classes can help, but the benefit is not created by empty chairs. It is created when lower numbers change the quality of instruction and participation.
- Separate class size from student–teacher ratio. A school may employ many adults while still teaching large classes, or have smaller classes without a low overall staffing ratio.
- Ask which instructional bottleneck the smaller class is meant to relieve. Visibility? Feedback delay? participation? disruption? grouping? practical safety? teacher workload?
- Redesign the lesson around the new capacity. More pupils can answer. More work can be inspected during the lesson. Groups can be reorganised. Feedback can arrive sooner.
- Protect curriculum ambition. Smaller classes should not become slower classes merely because more conversation is possible.
- Watch for threshold effects. A change from thirty to twenty-nine is different from a change from thirty to eighteen.
- Measure what actually changed. Count response opportunities, feedback latency, on-task behaviour, misconceptions caught and quality of pupil explanations—not only enrolment.
- Compare the opportunity cost. Additional teachers and rooms are expensive. The same money might sometimes buy more effective tutoring, professional development or targeted support.
- Avoid making class size a proxy for school quality. A smaller class with weak instruction can still underperform a larger class with excellent instruction.
The shortest principle is: class size changes the ceiling on teacher attention; learning improves only if teaching uses the space created below that ceiling.
1. Class size and student–teacher ratio are not the same thing
Class size is the number of learners following a common course in a particular class. Student–teacher ratio is a system-level calculation dividing students by teaching staff. The two numbers can diverge.
A school might report a low student–teacher ratio because it employs specialist teachers, intervention staff and teachers who do not each run a separate class every period. Yet a pupil may still sit in a class of thirty. Another school may organise smaller classes for a subject while its overall staffing ratio looks less favourable.
This matters because the mechanism lives in the room. If the educational claim is that a teacher can notice more, question more and respond faster, then the relevant number is the number of learners that teacher is actively responsible for during that lesson.
OECD’s Education at a Glance 2025 reports an average primary class size of about twenty-one across OECD countries and about twenty-three at lower-secondary level, while average student–teacher ratios are lower because the statistics measure different things. Parents, journalists and policymakers should not swap one for the other.
2. The first mechanism is an attention budget
A teacher has a finite number of seconds in a lesson. Every explanation, question, glance at a workbook, whispered correction and conversation consumes some of that budget.
In a forty-minute lesson with thirty pupils, even an impossible equal division would give each learner only eighty seconds of individual teacher time before whole-class explanation, transitions and administration are counted. Real classrooms are not divided equally. Some pupils need more clarification; some ask questions; some require behavioural redirection; some are missed because they look fine.
Reduce the class to twenty and the budget does not magically become one-to-one tuition. But the teacher has more room to allocate attention strategically.
The key word is allocate. A smaller class creates available attention. Good teaching decides where it should go.
3. Smaller classes can increase visibility
Teaching is partly a detection problem. The teacher needs to see who has started incorrectly, who has copied the procedure without understanding, who has stopped, who is dominating a discussion and who has quietly disappeared from it.
In a larger room, weak signals are easier to miss. A pupil can write three lines using a wrong assumption while the teacher is helping someone else. A hesitant learner can avoid answering for an entire lesson. A pair can drift off-task at the back without changing the room enough to attract attention.
With fewer pupils, the teacher may be able to scan more accurately and revisit each learner more often. But visibility still depends on routines. A teacher who remains at the front for the entire lesson may use little of the advantage.
Smaller classes make observation cheaper. They do not guarantee observation.
4. Feedback latency can shrink
Feedback is more useful when it arrives while the learner can still connect it to the attempt.
Imagine a mathematics class completing multi-step equations. In a larger class, the teacher may discover at the end that six pupils distributed a negative sign incorrectly in the first line and repeated the mistake through ten questions. In a smaller class, the teacher may reach those pupils after question two, correct the representation and prevent eight repetitions of the same error.
The learning gain does not come from “more feedback” in the abstract. It comes from shorter latency between error and useful information.
A smaller class can also make feedback more specific. Instead of writing “check signs” after the lesson, the teacher can ask, “What changed when you multiplied the bracket by negative three?” while the pupil’s reasoning is still visible.
5. Response distribution can widen
Whole-class questioning often hides a participation inequality. A small set of confident pupils volunteer repeatedly while others remain cognitively invisible.
With fewer learners, a teacher can use mini-whiteboards, cold call, brief partner explanation, written responses or structured turns and still inspect the class without the activity becoming unwieldy. The teacher can hear a larger fraction of the room.
This does not mean every question must be answered by every pupil aloud. It means the teacher can design the lesson so that thinking is distributed rather than concentrated.
A smaller class that still relies on three volunteers has not used one of its strongest possible advantages.
6. Discussion quality can improve when speaking time is not monopolised
Discussion is constrained by time. If thirty students each speak for one minute, the lesson is nearly over. This is why “more discussion” is not a sufficient plan for smaller classes.
The useful change is often in the size and structure of discussion groups. A class of eighteen can form six groups of three rather than six groups of five. Each learner has a larger share of the speaking opportunity, and the teacher can listen briefly to more groups.
The teacher can also demand fuller explanations because fewer contributions need to be sampled to understand the room.
But smaller groups can still fail if tasks are vague or one pupil does all the reasoning. Class size creates room for participation; task design determines whether participation is intellectual.
7. Behaviour can become easier to prevent rather than merely correct
Disruption is partly a visibility and probability problem. In a crowded room, small off-task behaviours can spread before the teacher notices. Movement takes longer. Materials create congestion. Transitions involve more people.
Reducing numbers can lower these coordination costs. The teacher may make brief contact before a pupil disengages, circulate without obstruction, and reset a routine with less interruption.
EEF notes that benefits from smaller classes appear linked partly to changes in learning behaviour and classroom interaction. The mechanism is plausible: fewer simultaneous behavioural demands let teachers intervene earlier and more quietly.
Still, class size is not a behaviour programme. Poor routines remain poor routines with twelve pupils.
8. Grouping becomes more flexible
A teacher with fewer pupils can reconfigure groups more quickly and with less monitoring overhead.
Suppose a Science lesson reveals three different misunderstandings. In a large class, creating three temporary groups may require complex movement and leave many pupils waiting. In a smaller class, the teacher can place six pupils around one table for a five-minute explanation while others complete a prepared extension.
This is where class size interacts with formative assessment. Evidence becomes useful only when the teacher can act on it.
Smaller classes can reduce the logistical cost of acting on classroom evidence. They do not remove the need for a coherent curriculum or accurate diagnosis.
9. Practical subjects can gain from safer, denser supervision
The value of lower numbers may be different in laboratory Science, Design and Technology, Physical Education, performance subjects or vocational workshops.
When learners handle equipment, move through space or perform multi-step practical procedures, the teacher’s supervision burden is not only academic. Safety, setup, materials and sequencing consume attention.
A smaller group may make more hands-on participation possible because the teacher can supervise more closely. This is a different mechanism from the one operating in a lecture-heavy subject.
Class-size effects should therefore not be treated as a universal constant. The instructional environment changes the value of an additional empty seat.
10. Pacing can become more responsive
Large classes encourage teachers to select a pace that is survivable for the middle of the distribution. Frequent individual adjustment is expensive.
With fewer learners, the teacher can detect whether “most pupils are ready” more accurately and can give a short intervention to a subset without stopping everyone. This can make pacing more responsive.
Yet there is a danger: smaller classes sometimes become conversationally comfortable and slower. The teacher gives every question a long discussion, finishes fewer examples and quietly reduces curriculum exposure.
The objective is not slowness. It is better control of pace—slowing when evidence demands it and moving when understanding is secure.
11. Smaller classes should change lesson architecture, not only teacher mood
Teachers often report that smaller classes feel easier. That matters for workload and stress, but feeling easier is not automatically a pupil learning mechanism.
A school should decide before reducing class size what teachers will do differently. More live checking? More short writing? More practical work? More targeted feedback? More pupil explanation? More flexible groups?
Without an architectural change, the intervention can become expensive comfort.
This is why EEF’s warning is central: reductions are most likely to matter when they permit teaching to change. The causal chain should be visible.
Fewer pupils → different opportunities → different teacher moves → different pupil activity → better learning.
If the middle links are missing, the first link cannot carry the whole argument.
12. Thresholds matter more than slogans
The difference between thirty and twenty-nine pupils is real but small. The difference between thirty and eighteen may alter the room.
EEF’s current synthesis notes that many studies examine reductions of roughly eight to ten pupils and that small reductions are less likely to be cost-effective. It also notes that identifiable effects may emerge when classes become smaller than around twenty, although this is not a universal law.
This is a threshold problem. Certain teaching moves become feasible only after enough capacity is released.
For example, conferencing with every pupil for two minutes is sixty minutes in a class of thirty and thirty-six minutes in a class of eighteen. The second may fit across a double period; the first may not.
The educational question is not “smaller?” but “small enough to change the feasible set of instruction?”
13. Age changes the mechanism
Younger pupils often require more help with routines, materials, attention and communication. This can make additional teacher attention particularly valuable.
Older students may operate more independently, but class size can still matter in essay conferencing, laboratory work, oral discussion, performance feedback or advanced problem solving.
A class-size policy that assumes the same effect in every year level ignores developmental and subject differences.
Schools should therefore ask where reduced numbers solve a real instructional bottleneck. A blanket reduction everywhere may be less efficient than targeted reductions in phases or subjects where interaction intensity matters most.
14. Subject changes the mechanism too
EEF reports a somewhat larger average impact in mathematics than reading within its synthesis, but the evidence is limited and heterogeneous. That should not be turned into a universal rule that mathematics “needs” small classes.
Instead, consider task structure. In mathematics, rapid error diagnosis and guided practice may make additional interaction especially valuable. In writing, conferencing and feedback can be time intensive. In reading, grouping and discussion structure matter. In practical subjects, supervision matters.
Class size has no single instructional pathway. It modifies the cost of many pathways.
The correct implementation question is: which high-value teaching move is currently too expensive to perform at this group size?
15. Disadvantaged learners may benefit differently
EEF notes international evidence suggesting smaller classes can have positive effects for socioeconomically disadvantaged pupils, with some evidence of larger benefits in primary settings.
There are plausible reasons. Learners who have had fewer educational opportunities outside school may benefit more from additional interaction, feedback and teacher attention. A smaller class may also make it easier to detect missing prerequisite knowledge before it compounds.
But class size should not become a substitute for curriculum quality. A disadvantaged learner does not need a diluted curriculum in a cosy room. They need strong teaching with barriers detected early.
Equity is not achieved by simply lowering a number. It is achieved when extra capacity is used to increase access to ambitious learning.
16. Teacher workload can fall—or simply change shape
Fewer books to mark is an obvious workload effect. Less administrative complexity and fewer parent communications may also help.
But high-quality small-class teaching can use the saved capacity for more responsive feedback, richer tasks and closer planning. Workload does not always vanish; some of it is reinvested.
This creates a leadership challenge. If teachers are told, “Your class is smaller, so you can do everything in greater detail,” the benefit can be converted into unsustainable expectations.
Class-size policy should specify which work is meant to improve and which work should not expand indefinitely. Capacity needs a destination.
17. Opportunity cost is not an optional footnote
Reducing class size is expensive because it usually requires more teachers and sometimes more rooms. EEF rates the implementation cost as very high.
That means the comparison is not “smaller classes versus nothing.” It is smaller classes versus what else the same money could fund: structured small-group tuition, teaching-assistant interventions, teacher professional development, curriculum materials, attendance support, specialist services or reduced teacher workload.
A school may rationally choose smaller classes for reasons broader than measured attainment—safety, inclusion, wellbeing, parent preference. But the trade-off should be explicit.
Educational resources are finite. A visible intervention should not escape comparison merely because it is intuitively attractive.
18. Class size can be confused with school size
Parents sometimes infer that a small school means small classes. It may not.
A small school can have limited staffing and combine year groups. A large school can have enough staffing flexibility to run smaller classes in selected subjects.
School size also changes peer networks, subject choice, specialist provision and administration. Those are separate mechanisms.
When evaluating a school, ask for actual class sizes in the relevant year and subject, not a general impression that the school is “small.”
19. Small-group tuition is not the same intervention
A class of eighteen is still a class. Small-group tuition typically means two to five learners receiving focused teaching targeted to diagnosed needs.
The two approaches can share mechanisms—greater interaction, easier diagnosis, faster feedback—but the intensity is different.
This is why a school facing one narrow learning gap may get more value from targeted small-group tuition than from reducing every class by three pupils.
Class size is a structural allocation decision. Small-group tuition is a targeted instructional intervention. Keeping the canonical boundary clear helps schools spend money on the problem they actually have.
20. One-to-one tuition is further along the intensity spectrum
If the learning bottleneck is highly individual—an entrenched decoding problem, a specific mathematics misconception, a complex writing pattern—one-to-one tuition may create a level of diagnostic interaction that even a class of fifteen cannot.
EEF’s toolkit finds stronger average effects for one-to-one tuition than for reducing class size, though the costs and implementation demands differ.
This does not make one-to-one “better” in every situation. It highlights the ratio mechanism: attention becomes more targeted as group size falls.
The decision should match the breadth of the need. Whole-class size reductions treat everyone. Tuition concentrates resources on selected learners.
21. Teacher expertise can dominate the size effect
A highly skilled teacher in a larger class may outperform a weaker teacher in a small one.
This sounds obvious but is often forgotten when parents compare schools. Class size is one condition within an instructional system. Curriculum coherence, subject knowledge, explanations, questioning, assessment, relationships and classroom management still matter.
Smaller classes can amplify good teaching because the teacher has more opportunities to deploy it. They can also amplify unproductive habits: excessive teacher talk, low challenge or over-helping.
The right mental model is multiplicative rather than substitutive. Class size changes the operating environment. Teacher expertise determines how effectively that environment is used.
22. The value of a smaller class depends on what the teacher does with silence
One hidden benefit of fewer learners is reduced waiting time. The teacher can circulate faster, but the class may also become quiet enough that learners expect the teacher to rescue every difficulty immediately.
This can create dependency. A student raises a hand at the first sign of friction because the teacher is now more available.
Good small-class teaching protects productive struggle. Extra teacher capacity should increase the quality of intervention, not simply the frequency of intervention.
Sometimes the best use of extra attention is noticing that a learner can continue without help.
23. Small classes can support richer oral language when turns are designed well
In language-heavy subjects, fewer learners can create more room for elaborated responses, follow-up questions and feedback on explanation.
A teacher can ask a pupil to defend a claim, invite another pupil to challenge it and return to the first speaker. That kind of sequence consumes time that is harder to distribute in a very large class.
Yet rich talk still needs content. Smaller classes are not automatically dialogic classes. If the questions remain shallow, the conversation simply becomes more evenly shallow.
24. Practical feedback can become more immediate
In art, music, PE, laboratory work or technical education, feedback often depends on observing performance in real time.
A smaller group allows the instructor to see more attempts before the task has moved on. A violin teacher hears more phrases. A PE teacher sees more movement patterns. A laboratory teacher catches more procedural errors.
This is a supervision-and-feedback mechanism that differs from essay marking after class.
Subject-specific analysis matters because the value of lower numbers is not evenly distributed across pedagogies.
25. Inclusion needs can change the effective group size
Two classes with twenty-four pupils can impose very different demands depending on the support needs, communication profiles and independence of the learners.
A numerical class-size policy can therefore miss the functional workload inside the room.
Some settings may need additional adult support, specialised expertise or environmental adaptation rather than a simple universal reduction.
Do not turn this into deficit accounting about children. The point is that resources should follow the actual instructional and access demands.
26. Space per learner can matter independently of number
Twenty pupils in a cramped room can be harder to supervise and move than twenty-four in a well-designed space.
Furniture, sightlines, acoustics, equipment and movement routes change how class size is experienced.
When leaders attribute improvement to lower numbers, examine whether space and room organisation changed too.
Class size is a social ratio embedded in a physical environment.
27. Technology can reduce some large-class visibility problems without changing class size
Digital polling, shared documents and learning platforms can gather evidence from many students quickly.
They do not create the same relationship or feedback capacity as lower numbers, but they can make some thinking visible at scale.
Schools facing fixed structural class sizes should therefore ask which mechanisms can be approximated through routines and tools. The alternative to structural reform is not helplessness.
28. Teaching assistants change adult density, not necessarily class size
Adding a teaching assistant lowers the number of pupils per adult in some activities, yet the class remains the same size and coordination becomes a new task.
If the additional adult simply hovers around low-attaining pupils or performs administrative work, the instructional mechanism differs from reducing the class.
The separate eduKateSG owner on teaching-assistant interventions should govern that design problem.
Do not report “more adults” and “smaller classes” as equivalent interventions.
29. Split classes can create curriculum-coordination costs
A school may divide one large class into two smaller groups. Now two teachers need aligned curriculum, pacing, assessment and expectations.
If one group receives a different instructional sequence, comparisons become difficult and pupil mobility can create friction.
Structural change creates new coordination work. This is manageable, but it should be planned rather than ignored.
30. Teacher allocation can erase or amplify the benefit
If the smallest groups receive the least experienced teachers, some potential benefit may be offset. If leaders strategically place strong teachers with groups that need intensive interaction, the effect could be amplified.
EEF specifically notes teacher allocation as one mechanism that can influence outcomes under setting and other group structures.
Class-size decisions therefore interact with staffing quality, not merely staffing quantity.
31. Schools should avoid involuntary ability sorting disguised as class-size policy
Sometimes small classes are created mainly for lower-attaining pupils while larger classes remain for others. That is no longer a pure class-size intervention; it is also an attainment-grouping intervention.
Expectations, peer composition and curriculum may change simultaneously.
When evaluating effects, name all moving parts. Otherwise the institution may attribute a grouping effect—positive or negative—to size alone.
32. Parent perception can influence enrolment decisions more than evidence warrants
“Maximum class size 12” is easy to market. “Teachers use whole-class response systems, rapid feedback and coherent curriculum” is harder to visualise.
Parents should value a small class as one potentially useful condition, not as proof of quality.
Ask what the school does with the ratio. A premium price for a small class should purchase better instructional possibilities, not simply fewer desks.
33. The smallest class is not always the most socially rich
Peers provide models, discussion partners, friendship possibilities and diverse perspectives. Extremely small classes can limit group formation or subject availability.
This is not an argument for large classes. It is a reminder that class size has more than one social effect.
The optimum depends on the educational job, resources and community context. There is no universal number that maximises every outcome.
34. School leaders should model thresholds, not averages alone
An average class size of twenty-four can hide one class of thirty-two and another of sixteen.
Staffing models should inspect distributions, peak sizes and where high interaction is educationally valuable.
Averages are useful for budgeting and misleading for lived classroom experience.
Ask how many pupils are actually in each relevant teaching group.
35. Research averages hide implementation differences
Studies of class-size reduction vary in the size of the reduction, age groups, teacher practices, duration and surrounding policies.
This heterogeneity is one reason EEF rates the evidence security as very limited despite a sizable literature.
Schools should resist copying the average effect as a promise. The research estimate is a starting prior, not a local guarantee.
36. The mechanism should be declared before the budget is spent
If leaders say the goal is faster feedback, define what faster means and how practice will change. If the goal is practical safety, specify the tasks affected. If the goal is participation, observe response distribution.
A declared mechanism creates an evaluation target.
Without one, class-size reduction becomes a high-cost intervention whose success is judged mainly by satisfaction.
37. Implementation should include teacher planning time
Teachers may know in principle that smaller classes allow richer feedback but need time to redesign routines, tasks and grouping.
Professional learning can focus on exploiting the new capacity: conferencing, all-pupil response, flexible grouping, practical supervision and timely feedback.
Do not assume a structural change automatically rewrites pedagogical habits built for larger groups.
38. The counterfactual matters
If a school spends heavily on smaller classes, ask what would otherwise have been funded.
Perhaps the alternative is one-to-one tutoring for struggling readers, additional specialist staff, better curriculum resources or teacher planning time.
The correct decision depends on local constraints and goals. High cost means the comparison set matters.
Class size should win the investment only when its mechanism fits the problem well enough.
39. Smaller classes can be chosen for legitimate non-attainment reasons
A school might value wellbeing, teacher workload, community feel, safety or inclusion even if measured academic effects are modest.
That is a legitimate policy choice if stated honestly.
The error is to justify every preference with inflated claims about test scores. Education contains values as well as effect sizes.
40. The endpoint is better learning conditions used well
The strongest conclusion is modest and practical. Class size changes the operating conditions of teaching. It can release attention, visibility, time and flexibility. Those resources then have to be converted into instructional activity.
The empty chair is not the mechanism. What teachers and pupils can now do is the mechanism.
Worked case 1 — Thirty-two becomes twenty-four, but the lesson stays the same
A school invests heavily to reduce Year 7 mathematics classes from thirty-two to twenty-four. Teachers are pleased. Behaviour improves slightly. Yet attainment barely moves.
Lesson observation shows why. Teachers still explain for twenty minutes, ask volunteers for answers, assign the same worksheet and collect books for marking after class. The smaller group reduced pressure but did not change learning activity.
The department redesigns lessons. Every pupil now shows answers on mini-whiteboards at key points. Teachers circulate during the first three independent problems. Misconceptions trigger temporary groups. Feedback happens before pupils complete the full set.
The second change—not the first—activates the instructional capacity created by the smaller class.
Worked case 2 — A small writing class uses the time badly
A writing programme advertises twelve students per class. Parents expect intensive feedback.
The teacher reads each student’s full draft during the lesson while the other eleven wait quietly. The small class has created a teacher bottleneck rather than solved one.
The programme changes the routine. Students first complete a short self-check. The teacher then conducts three-minute conferences focused on one high-leverage feature: argument, evidence, paragraph logic or sentence control. While waiting, students revise a specified section rather than sitting idle.
The small class becomes effective because scarce teacher attention is scheduled around a clear instructional job.
Worked case 3 — A large class with strong response systems
A primary Science teacher has thirty-one pupils. The school cannot reduce numbers.
Instead, she improves visibility: mini-whiteboards, turn-and-talk with assigned roles, quick checks at transition points, structured circulation and short exit tasks. She cannot create the attention budget of a class of eighteen, but she reduces the amount of invisible learning.
This case matters because class size is not destiny. Schools facing fixed staffing can still redesign how evidence is collected and attention distributed.
The existence of a structural constraint should not become an excuse to ignore instructional design.
Worked case 4 — A targeted reduction beats a blanket reduction
A school has funds to hire two additional teachers. One option is to reduce every lower-secondary class by about two pupils. Another is to create much smaller groups for first-year mathematics and literacy transition modules where diagnostic data show large gaps.
Leaders model the likely mechanism and costs. A reduction of two pupils is unlikely to change normal lesson architecture. The targeted groups, however, can move from thirty to fifteen for specific periods.
The school chooses the targeted model and evaluates response opportunities, feedback latency, curriculum progress and learner transfer back to ordinary classes.
The decision is not ideological. It follows the size change that is large enough to alter instruction.
Practical route for teachers
If your class becomes smaller, do not simply enjoy the relief and continue as before. Choose two or three instructional moves that were previously too expensive.
You might increase whole-class response checks, schedule brief conferences, circulate earlier, use temporary diagnostic groups, hear more pupil explanations or shorten feedback delay. Keep the curriculum goal fixed and monitor whether the new routines increase thinking rather than teacher attention for its own sake.
If your class remains large, steal the mechanism where possible. Use response systems that make all pupils visible. Build routines that reduce transition cost. Pre-plan where you will circulate. Use peer structures that preserve accountability. A structural limitation does not eliminate instructional agency.
Practical route for parents and families
Do not judge a school by class size alone.
Ask what teachers can do because classes are smaller. How is feedback given? How are quiet pupils made visible? Are misconceptions caught during lessons? Are groups flexible? Is the curriculum ambitious? How stable are teachers?
A class of sixteen is not automatically personalised; a class of twenty-eight is not automatically anonymous.
If your child is struggling, ask whether the problem requires a different class size or a more targeted intervention. Sometimes a short period of high-quality small-group or one-to-one teaching is a better answer than moving to a school with slightly smaller ordinary classes.
Practical route for learners
A smaller class gives you more access to the teacher, but access still has to be used.
Ask questions earlier. Show your working. Volunteer explanations that reveal your reasoning, not only final answers. Use feedback before the next attempt rather than storing it in a folder. If you understand, use the extra interaction to test whether you can explain, compare and transfer.
In a larger class, make your thinking visible through written responses, questions and deliberate help-seeking. Do not wait for the teacher to infer confusion from silence.
Practical route for school leaders
Before changing class size, write the mechanism on one page.
Which year groups or subjects? What size change? Which teacher behaviours should change? Which pupil behaviours should change? What will be measured? What is the opportunity cost? What happens if recruitment leaves vacancies?
Support teachers to redesign lessons rather than assuming adaptation will occur automatically. Protect the saved capacity from being consumed by new administrative demands. Compare the cost with targeted alternatives.
Most importantly, do not report “average class size fell” as if that were the educational outcome. Report whether the instructional system changed.
Common failure modes
- Reducing a class from thirty to twenty-eight and expecting a large attainment effect.
- Treating student–teacher ratio as if it were actual class size.
- Keeping the same teacher-talk-heavy lesson architecture after numbers fall.
- Letting the same confident pupils dominate discussion.
- Slowing curriculum coverage because every contribution now receives extended attention.
- Using smaller classes as a marketing proxy for teaching quality.
- Ignoring the opportunity cost of additional staffing and rooms.
- Assuming all subjects and ages benefit through the same mechanism.
- Creating smaller classes but increasing teacher administrative work until the capacity disappears.
- Confusing whole-class reduction with targeted small-group tuition.
- Measuring only final test scores and never checking whether participation, feedback or visibility changed.
- Treating a threshold estimate such as “below twenty” as a universal law.
- Assuming disadvantaged pupils need lower expectations in smaller rooms.
- Cutting class size without training teachers to exploit greater interaction.
- Declaring the intervention successful because the room feels calmer.
Frequently asked questions
Are smaller classes always better?
No. On average, smaller classes show a small positive effect, but the evidence is limited and the benefit depends heavily on whether teaching and pupil behaviour change.
What does EEF currently estimate?
EEF’s current Toolkit estimates roughly one month of additional progress on average, rates implementation cost as very high and evidence security as very limited.
Is reducing thirty pupils to twenty-five enough?
It may improve the experience of teaching, but EEF notes that small reductions are less likely to be cost-effective. Larger changes are more likely to alter what teachers can actually do.
Is a student–teacher ratio of 14:1 the same as a class of fourteen?
No. Student–teacher ratio counts teaching staff across a system. Actual classes can be larger or smaller.
Why might mathematics benefit more?
EEF’s synthesis reports a somewhat larger average effect in mathematics, but evidence is heterogeneous. Rapid checking and correction may be one plausible mechanism, not a settled universal explanation.
Do smaller classes help behaviour?
They can make monitoring, movement and early redirection easier, and EEF notes behavioural changes as part of the likely pathway. Strong routines are still necessary.
Should schools reduce every class equally?
Not necessarily. Targeting year groups, subjects or learner groups where interaction intensity is a genuine bottleneck may be more efficient.
What should parents look for?
Look for evidence of interaction quality: pupils responding, teachers circulating, timely feedback, flexible grouping and ambitious curriculum work.
Is small-group tuition just a smaller class?
No. Small-group tuition is a targeted intervention, commonly two to five learners, designed around diagnosed needs. A class-size reduction is a structural change affecting the ordinary class.
Can large classes still work well?
Yes. Strong routines, all-pupil response systems, formative assessment and careful task design can make learning visible even when structural ratios cannot change.
Sources and further reading
- Education Endowment Foundation, Reducing class size, Teaching and Learning Toolkit, review updated October 2025 and current in 2026: https://educationendowmentfoundation.org.uk/education-evidence/teaching-learning-toolkit/reducing-class-size
- OECD, Education at a Glance 2025: How do student-teacher ratios and class sizes vary across education levels up to upper secondary education?: https://www.oecd.org/en/publications/2025/09/education-at-a-glance-2025_c58fc9ae/full-report/how-do-student-teacher-ratios-and-class-sizes-vary-across-education-levels-up-to-upper-secondary-education_76b87b21.html
- Education Endowment Foundation, Small group tuition: https://educationendowmentfoundation.org.uk/education-evidence/teaching-learning-toolkit/small-group-tuition
- Education Endowment Foundation, One to one tuition: https://educationendowmentfoundation.org.uk/education-evidence/teaching-learning-toolkit/one-to-one-tuition
Continue exploring on eduKateSG
- How Small-Group Tuition Works for Performance | Visibility, Diagnosis and Fast Correction
- How Checks for Understanding Work | Why Evidence Must Change the Next Teaching Move
- How Instructional Pacing Works
- How Within-Class Attainment Grouping Works
- How Teaching Assistant Interventions Work
- How X Works Hub
The final idea
Class size is one of education’s most visible numbers and one of its easiest to misunderstand.
Fewer pupils can release something valuable: attention, visibility, time, feedback, flexibility and calmer coordination. But released capacity is not learning. It becomes learning only when teachers convert it into better instructional moves and pupils convert those moves into more successful thinking and practice.
So the useful comparison is not “large class versus small class.” It is what can this teacher now do, with these learners, at this size, that was previously too difficult or too slow—and is that new possibility actually being used?
When the answer is clear, class size becomes an instructional variable rather than a slogan.