The same 3-pax class size should not produce the same teaching design in English, Mathematics and Science. The group format is only the container. The subject determines what the tutor needs to observe, what students should produce and what counts as successful transfer.

This page has one job: show Punggol parents how a 3-pax small-group lesson should operate differently across English, Mathematics and Science rather than treating “small group” as a complete teaching method.

For current subject programmes, use the English, Mathematics and Science routes. This legacy page owns the cross-subject lesson-design comparison.

---

What 3-pax gives every subject

• more student turns;
• closer visibility of working and explanation;
• faster feedback;
• peer comparison without a large-room effect;
• room to differentiate prompt level; and
• short periods of independent work while the tutor focuses elsewhere.

Those are format advantages. The subject determines how they should be used.

English: maximise language production

English tuition needs students to produce and interpret language. A useful 3-pax English lesson should include substantial reading, explanation, writing, revision and oral response.

• each student explains evidence for a comprehension answer;
• students compare two possible word choices;
• writing plans are discussed but final writing remains individual;
• grammar corrections are explained, not merely filled in;
• oral responses receive genuine follow-up questions.

The tutor should be listening for precision, organisation and language control.

Mathematics: maximise visible reasoning

Mathematics tuition needs the tutor to see how the student represents and transforms quantitative relationships.

• students show working before the final answer hides the route;
• different methods can be compared for validity and efficiency;
• the tutor identifies the first invalid line;
• one condition can be changed to test understanding;
• students estimate and check answers.

The tutor should be looking for concept, representation, selection, execution and checking failures.

Science: maximise explanation from evidence

Science tuition needs students to connect observations, diagrams and data to scientific concepts and mechanisms.

• students explain why an observation occurs;
• the group compares two scientific explanations;
• variables in an investigation are identified;
• diagrams are annotated to show processes;
• a setup can be changed and students predict the outcome.

The tutor should distinguish fact recall from genuine scientific reasoning.

The wrong use of a small group

A three-student class can still be low-information if all three complete identical worksheets while the tutor waits.

• English becomes silent worksheet completion.
• Mathematics becomes answer checking instead of reasoning diagnosis.
• Science becomes fact memorisation without explanation.

The group size does not fix poor task design.

Different feedback loops

The feedback loop should match the subject.

• English: interpret → express → revise → reuse.
• Mathematics: represent → select → solve → check → vary.
• Science: observe → recall concept → explain mechanism → test with new evidence.

All three include correction and transfer, but the middle of the loop differs.

Different forms of peer learning

Peer comparison should also be subject-specific.

• English: compare interpretations, wording and organisation.
• Mathematics: compare methods and efficiency.
• Science: compare explanations and strength of evidence.

Students should learn from difference without simply copying the strongest student.

Different types of independence

• English independence: can the student read, plan, answer and revise without line-by-line prompting?
• Mathematics independence: can the student identify the relationship and method without being told the chapter?
• Science independence: can the student select the relevant concept and explain a new observation?

A 3-pax class should steadily move each student towards the relevant form of independence.

Different ways to differentiate the same shared task

Differentiation does not require three unrelated lessons.

• English: same text, different levels of evidence prompting.
• Mathematics: same concept, different complexity or number of steps.
• Science: same diagram, different depth of explanation or inquiry question.

The shared core preserves group discussion; the adjusted demand preserves fit.

Current 2026 examination context

For 2026 PSLE candidates, SEAB lists English Language 0001, Mathematics 0008 and Science 0009. The three assessments differ substantially in content and task, which is another reason one generic tuition method is insufficient.

Official reference: SEAB PSLE formats examined in 2026.

How a mixed family programme should use one 3-pax philosophy

eduKatePunggol’s current model is up to three students, typically for 1.5 hours. Across subjects, the shared philosophy can remain consistent:

1. diagnose the actual weak link;
2. teach clearly;
3. make student reasoning visible;
4. correct the high-leverage failure;
5. retest after support is reduced; and
6. transfer into school or exam conditions.

The subject-specific lesson design sits inside that common framework.

When 3-pax is a good fit

• the student benefits from close feedback but can work independently for short intervals;
• peer comparison improves reasoning rather than distracts;
• the child’s level is compatible with the group;
• the tutor can differentiate within the shared task; or
• parents want a class where errors and thinking are easier to observe.

When another format may be better

• the learner requires specialised support;
• continuous one-to-one redirection is necessary;
• the student is far outside the group’s level;
• peer presence causes persistent distraction or anxiety; or
• the learner is already highly independent and only needs occasional specialist consultation.

Small group is a format choice, not a universal quality label.

Parent checklist by subject

• English: how much language does my child actually produce?
• Mathematics: can the tutor see where the working first goes wrong?
• Science: does my child explain mechanisms or only memorise answers?
• All: is tutor support decreasing as competence grows?

Progress receipts

• English answers become more precise and independently revised;
• Mathematics working becomes more valid and self-checked;
• Science explanations connect evidence to mechanism more consistently;
• peer comparison improves thinking without copying; and
• students increasingly perform outside the tutor’s immediate support.

Continue to current subject programmes

For current class information, use the English, Mathematics and Science routes.

---

About this rebuilt legacy page

The original page bundled English, Mathematics and Science inside one generic small-group pitch. This 2026 update keeps the multi-subject URL but gives it a useful comparative role: showing why the same 3-pax format must be run differently according to the subject.

---

One group size, three different learning systems

A three-student class creates the same basic opportunity in every subject: the tutor can see more individual work, hear more explanations and return feedback quickly. But the information the tutor needs is different.

English reveals itself through language production and interpretation. Mathematics reveals itself through representation, method selection and working. Science reveals itself through models, evidence and causal explanation. A good 3-pax programme should therefore feel recognisably different from subject to subject.

English 3-pax: the learner must produce language

English cannot be taught well if students spend most of the lesson silently completing worksheets. The tutor needs to hear what the child thinks, see how ideas become sentences and observe where meaning breaks during reading or writing.

• Read aloud or explain selected passages where useful.
• Justify comprehension evidence.
• Generate vocabulary in context.
• Plan and draft independently.
• Revise after feedback.
• Respond orally to follow-up questions.

The central English question is: how much useful language is each learner producing and revising?

Mathematics 3-pax: the learner must make reasoning visible

A final answer can hide a weak method or lucky guess. Mathematics tuition should inspect the path: how the child represented the problem, why a method was chosen, where the first invalid line occurred and whether checking was available.

• Show working.
• Explain method choice.
• Compare alternative valid routes.
• Estimate before exact calculation where useful.
• Identify the first broken step.
• Retest with changed numbers or context.

The central Mathematics question is: where did the mathematical model or execution first diverge?

Science 3-pax: the learner must connect model to evidence

Science answers can look fluent while hiding memorised keywords. The tutor needs to know whether the child can explain the concept, identify the evidence and make a claim that fits the evidence.

• Predict before seeing the answer.
• Explain the concept in plain language.
• Read graphs, tables and diagrams accurately.
• Identify variables and controls.
• Compare competing explanations.
• Revise the model when evidence disagrees.

The central Science question is: does the learner understand why the evidence supports the explanation?

The same tutor attention should produce different outputs

In English, tutor attention should produce clearer language and more independent revision. In Mathematics, it should produce more valid reasoning and more precise checking. In Science, it should produce stronger concept-evidence chains.

If every subject lesson looks like “teacher explains, students complete worksheet, teacher checks”, the small group is underused.

The feedback loop is the real 3-pax advantage

Small size matters when feedback arrives while the learner can still reconstruct the reasoning. The tutor sees the error, identifies the mechanism, gives the smallest useful correction and requires a fresh attempt.

The loop should repeat often enough that old errors become less frequent and prompts reduce.

English feedback latency

If a student writes an unsupported inference, waiting several weeks to address it allows the pattern to repeat. In a small group, the tutor can ask for evidence, hear the explanation and require a rewritten answer immediately.

The same applies to tense shifts, sentence boundaries, paragraph drift and forced vocabulary.

Mathematics feedback latency

A sign error at line two can corrupt the entire solution. The tutor should identify the first divergence rather than explain the final wrong answer only.

Fast correction is especially useful when the wrong algebraic habit could otherwise be rehearsed several more times.

Science feedback latency

A misconception can hide underneath correct terminology. The tutor should expose it early through prediction, drawing or explanation before the student memorises another model answer over the top.

The feedback changes the model, not just the sentence.

English peer learning: compare interpretations, not copy answers

Three students can read the same passage and select different evidence. Discussion helps them compare which interpretation is best supported.

Private first attempts are important. Otherwise the strongest student’s answer becomes the group answer and the tutor loses diagnostic visibility.

Mathematics peer learning: compare methods

One student may use a bar model, another an equation, another a unitary method. The tutor can compare validity, efficiency and checking without requiring everyone to use one route.

The final fresh question should still be solved independently.

Science peer learning: compare predictions and evidence

Students can make predictions privately, then compare why their models differ. The evidence from an experiment, graph or diagram becomes the shared basis for revision.

Peer discussion should not become a vote on scientific truth. The tutor calibrates the final explanation.

English differentiation: same text, different prompt depth

One student may need help locating evidence, another needs answer precision, another is ready to analyse tone or argument. The text can remain shared while the question depth changes.

This keeps discussion coherent without pretending every learner is at the same state.

Mathematics differentiation: same concept, different complexity

One learner may use a visual scaffold, another solve standard questions independently, another compare two methods or tackle a changed condition.

The shared concept keeps the class connected; complexity and prompt depth vary.

Science differentiation: same phenomenon, different reasoning depth

One learner identifies the observation, another explains the mechanism, another evaluates the experimental design or predicts a changed condition.

Depth can increase without moving students into unrelated topics.

English independence means less language scaffolding

Sentence frames, vocabulary prompts, evidence questions and planning boxes should reduce over time. The student should still produce coherent English when the support disappears.

The final proof is a fresh passage or prompt completed with minimal tutor intervention.

Mathematics independence means less method prompting

The student should increasingly identify the unknown, represent the relationship, choose a method and check the answer without the tutor naming the chapter or first step.

Mixed-topic work is especially useful for this because chapter cues disappear.

Science independence means less concept prompting

The learner should be able to inspect evidence, select the relevant scientific model and write a justified claim without the tutor asking “Which keyword?” first.

Changed-context transfer is the strongest test.

Different subjects need different kinds of productive struggle

In English, productive struggle may mean searching for evidence or revising a sentence. In Mathematics, it may mean choosing a representation or checking the last valid line. In Science, it may mean reconciling a prediction with contradictory evidence.

The tutor should know when to wait and when direct instruction is more efficient.

Different subjects need different stopping rules

English revision can become endless if the child keeps polishing every sentence. Mathematics checking can become excessive if the student repeats correct calculations. Science answers can become bloated when the learner writes every related fact.

Each subject needs a definition of “complete enough”.

English stopping rule

The response is relevant, complete, supported, grammatically stable enough and checked for the learner’s personal high-risk errors.

Mathematics stopping rule

The method is valid, working is clear enough, the answer matches the required form and one relevant plausibility or personal check has been completed.

Science stopping rule

The concept is correct, the evidence is relevant, the causal link is explicit enough and the claim does not exceed what the question supports.

Different subjects need different homework

English homework may be reading, vocabulary retrieval, one revised paragraph or a fresh comprehension response. Mathematics homework may be a focused error-family set or mixed retrieval. Science homework may be explanation, graph interpretation or changed-condition prediction.

One generic homework philosophy across subjects creates unnecessary load.

Different subjects need different error ledgers

• English: reading, vocabulary, inference, evidence, grammar, planning, paragraph control, timing.
• Mathematics: concept, representation, algebra, strategy, execution, units, timing, checking.
• Science: concept, evidence, misconception, causal explanation, claim calibration, variable reasoning, timing.

The categories make parent updates and tutor planning more precise.

The ledgers should share one field: prompt depth

Across all three subjects, support should fade. Record whether the learner needed a full model, strong cue, light cue or no prompt.

This creates one common measure of independence even though the subjects differ.

The ledgers should share one field: delayed retest

An immediate correction is not enough in any subject. Return later with a fresh passage, problem or Science context.

Durability is cross-subject even though the content differs.

The ledgers should share one field: school transfer

Tuition improvement should appear in the environment where the tutor is absent. Marked school work is therefore useful external evidence for English, Mathematics and Science.

A 3-pax English lesson example

1. Private reading and annotation.
2. Individual answer attempt.
3. Compare evidence.
4. Tutor diagnoses each error family.
5. Students rewrite independently.
6. Fresh transfer question.

The group is used for language contrast; the proof remains individual.

A 3-pax Mathematics lesson example

1. Private problem representation.
2. Show working.
3. Compare methods where useful.
4. Tutor identifies first broken step.
5. Focused repair.
6. Changed-number or changed-context retest.

The group is used for method comparison; the proof remains individual.

A 3-pax Science lesson example

1. Private prediction.
2. Explain the model.
3. Inspect shared evidence.
4. Compare explanations.
5. Tutor calibrates misconception and claim.
6. Changed-condition prediction.

The group is used for competing models; the proof remains individual.

What the tutor should hear in English

The tutor should hear whether the student understands the text, can justify an inference, can explain a word choice and can revise a sentence after feedback.

Silence for most of the lesson means a major diagnostic channel is being lost.

What the tutor should see in Mathematics

The tutor should see representation, notation, line-by-line transformation and checking. A page with only final answers removes much of the diagnostic value of small-group visibility.

What the tutor should challenge in Science

The tutor should challenge unsupported claims, keyword-only explanations, unfair experimental conclusions and models that fail when one condition changes.

The group should not be run at one uniform pace

One learner may need more processing time, another may need a harder transfer task, another may need direct repair. Shared teaching can be brief; individual work can branch temporarily.

Group fit depends on whether these differences remain manageable without turning into three unrelated lessons.

The fastest student should not control the lesson

Private first attempts and extension tasks prevent a fast student from pulling the group forward before others have produced evidence of understanding.

The slowest student should not permanently control the lesson either

If one learner consistently requires a different curriculum or much more rebuilding, the group composition should be reviewed.

Small group should allow differentiation, not permanent mismatch.

Peer correction should be calibrated differently by subject

In English, peers can question evidence or wording. In Mathematics, they can identify a method difference. In Science, they can challenge whether evidence supports a claim.

The tutor remains responsible for final accuracy.

The group should create short independence windows

While the tutor works with one student, the other two should have meaningful tasks that test retrieval, transfer or checking. These moments can be valuable because the learner cannot immediately ask for rescue.

Productive waiting is an advantage of small group when the task is calibrated correctly.

The group should make progress visible

Students can notice that a peer now solves independently, explains more precisely or no longer needs a particular scaffold. The tutor should frame this as strategy and growth, not public ranking.

Parents should ask different questions by subject

• English: What language process is becoming more independent?
• Mathematics: Where does the working first break, and is that error recurring?
• Science: Which misconception or evidence problem is currently active?

“How was tuition?” is less useful than a subject-specific learning question.

A parent should not buy three subjects simply because one format is convenient

The child may need English support and not Mathematics, or Science repair and not English. Subject allocation should follow evidence.

A 3-pax model can be excellent and still be unnecessary in one subject.

Subject need should be reviewed term by term

A learner may move one subject into maintenance while another becomes the active repair. Tuition intensity should be allowed to change.

This prevents the household schedule from expanding permanently.

The final cross-subject 3-pax checklist

• Every learner attempts privately before group comparison.
• The tutor can see or hear the subject-specific reasoning.
• Feedback is tied to a named mechanism.
• Prompts differ by learner where needed.
• Reattempt follows correction.
• Delayed retesting occurs.
• School transfer is checked.
• Peer discussion does not replace individual proof.
• Stable repairs move into maintenance.
• Support reduces over time.

The final lesson-design principle

The value of three students is not that English, Mathematics and Science can be taught with one generic small-group formula. The value is that the group is small enough for the tutor to adapt the feedback loop to the subject while still benefiting from peer contrast and independent work.

English needs visible language. Mathematics needs visible reasoning. Science needs visible models and evidence. The group size stays the same; the educational work changes.

For Punggol families, that is the standard worth using when comparing 3-pax classes: do not ask only how many students are in the room. Ask what the tutor can now see, hear, correct and retest in this particular subject because the group is small.

The three subjects should not share one generic lesson timer

English may need longer uninterrupted writing or reading blocks. Mathematics may benefit from several shorter problem cycles with immediate debugging. Science may need time for prediction, evidence inspection and explanation before a final written response.

The class duration can be similar while the internal rhythm differs.

English time should protect deep language production

Constant tutor interruption can make writing fragmented. The tutor may need to observe a longer paragraph before deciding what to correct. Some feedback should be delayed until the child completes the thought.

Small-group attention should not become constant editing.

Mathematics time should protect the first independent attempt

If the tutor explains before the student begins, the class loses information about representation and method selection. Give enough quiet time for the learner to commit to a route.

Only then can the tutor see whether the first error belongs to concept, strategy or execution.

Science time should protect prediction before evidence

If students see the correct result before predicting, the original model disappears from view. A short private prediction gives the tutor diagnostic access to what the learner currently believes.

That diagnostic moment is one of the strongest uses of a small Science group.

Each subject should have its own “first five minutes”

• English: short reading, retrieval or discussion to expose learner state.
• Mathematics: one retrieval problem or representation task.
• Science: one prediction, diagram or concept explanation.

The opening should calibrate the learner before the main lesson begins.

Each subject should have its own “last ten minutes”

• English: revision, retrieval and a clear next writing or reading target.
• Mathematics: fresh independent problem plus personal check.
• Science: changed-condition question or concise explanation from evidence.

The closing should prove that the learner can carry the correction rather than only understand the tutor’s explanation.

The three subjects should use different forms of silence

In English, silence can be drafting or close reading. In Mathematics, it can be private representation and calculation. In Science, it can be prediction or evidence inspection.

Silence is useful when it protects cognition. It is unhelpful when students are merely filling pages without a clear learning job.

The three subjects should use different forms of talk

English talk develops interpretation, vocabulary and expression. Mathematics talk justifies method and exposes structure. Science talk explains models, evidence and causality.

A 3-pax class should make enough room for every student’s subject-specific talk to become visible.

The tutor should avoid the same praise language across subjects

“Good job” is less useful than “your evidence now supports the inference”, “your representation is correct even though the arithmetic slipped”, or “your claim now matches the experiment”.

Specific praise reinforces the process the learner should repeat.

The tutor should avoid the same correction language across subjects

“Careless” does not explain an English evidence omission, a Mathematics sign error and a Science variable mistake. Each needs a mechanism-specific next action.

The small group should increase correction precision, not merely correction frequency.

The tutor should know what not to correct immediately

In English, low-priority style issues may wait until the paragraph is complete. In Mathematics, an isolated arithmetic slip may be left briefly to see whether the student catches it. In Science, imprecise wording may wait until the concept explanation is finished.

Observation before intervention provides valuable diagnostic information.

The tutor should know what to correct immediately

A repeated grammar misconception, invalid algebraic transformation or wrong scientific model may need earlier interruption before the error is rehearsed repeatedly.

Feedback timing should depend on the cost of letting the error continue.

A 3-pax English group can share one text but own different errors

One student may misread vocabulary, another infer beyond the evidence, another understand perfectly but write vaguely. The shared passage keeps the discussion efficient while the tutor tracks three different learner states.

A 3-pax Mathematics group can share one problem but own different errors

One student may choose the wrong operation, another choose correctly and miscalculate, another solve accurately but inefficiently. The class can compare why the final answers differ without treating every error as the same.

A 3-pax Science group can share one experiment but own different models

One learner may misidentify the variable, another choose the wrong concept, another understand the concept but overstate the claim. The tutor can compare the layers while preserving individual proof.

The class should record cross-subject independence differently

• English: starts, interprets, plans, writes and revises with less language scaffolding.
• Mathematics: represents, selects, solves and checks without chapter cues.
• Science: predicts, selects concept, reads evidence and explains without keyword prompts.

These are related forms of self-regulation, but the observable behaviours differ.

A parent should not expect one child to need the same class size in every subject

A student may need close small-group English feedback but be highly independent in Mathematics. Another may need intensive algebra repair but only light Science support.

Class format should follow subject need rather than a household preference for one universal package.

The subject budget should be reviewed monthly

1. Which subject has the highest-carry bottleneck?
2. Which subject is stable enough for maintenance?
3. Which subject needs small-group feedback resolution?
4. Which could be handled independently or with lighter support?
5. Does the total weekly load remain sustainable?

This prevents convenience from turning into unnecessary three-subject intensity.

English maintenance should look different from Mathematics maintenance

English maintenance may be sustained reading, occasional vocabulary retrieval and periodic writing. Mathematics maintenance may be spaced mixed problems. Science maintenance may be concept retrieval and changed-context explanation.

Even maintenance should respect subject structure.

Extension should also differ by subject

English extension can deepen reading, argument, vocabulary nuance and writing craft. Mathematics extension can compare methods, generalise relationships and solve less familiar problems. Science extension can evaluate evidence, design investigations and test model limits.

Harder worksheets are not the only form of extension.

Teaching ahead should be subject-sensitive

A strong Mathematics student may benefit from selective preview when algebra is secure. English often benefits more from depth in reading and writing than from simply moving to older-level worksheets. Science preview is useful only when prerequisite concepts are stable.

Acceleration should follow readiness, not a one-size-fits-all philosophy.

The centre should preserve distinct subject expertise

A common small-group philosophy can exist across the centre—diagnose, teach, reattempt, retest, fade—but subject knowledge and pedagogy still matter.

English feedback requires language expertise. Mathematics diagnosis requires mathematical structure. Science correction requires accurate conceptual and evidential reasoning.

The centre should preserve distinct subject materials

Materials can share design principles such as retrieval and transfer, but the tasks themselves should reflect the subject. A reading passage is not a Mathematics problem with words; a Science graph is not merely a data-comprehension exercise.

Subject identity should remain clear.

The centre should preserve one common learner history across subjects

Some cross-subject patterns still matter: task initiation, checking habits, fatigue, prompt dependence and time management. A child who waits for the first step in every subject may have a broader self-management issue.

The tutor team can notice this without flattening the academic differences among subjects.

The family should hear one common philosophy and three subject-specific explanations

The common philosophy might be: diagnose the bottleneck, make thinking visible, correct precisely, require reattempt, retest later and reduce support.

The subject explanation should then show how that philosophy becomes English, Mathematics or Science teaching in practice.

A first-month 3-pax audit

• Has each student produced independent evidence?
• Does the tutor know each learner’s active error family?
• Are prompts different where needed?
• Has feedback produced a fresh attempt?
• Has one correction been retested later?
• Is peer discussion helping rather than hiding learners?

These questions are relevant across subjects, though the evidence looks different.

A term-end 3-pax audit

• Which error has been retired?
• Which prompt has been removed?
• What school transfer is visible?
• Does the group still fit the learner’s pace?
• Is the subject still worth the same tuition intensity?
• What is the next exit condition?

The format should keep earning its place.

The learner should eventually understand why the subjects feel different

An older student can say: “In English I need to support my interpretation. In Maths I need to find the first wrong line. In Science I need to connect the evidence to the concept.”

This metacognitive distinction helps the learner switch strategies instead of applying one generic “study harder” response everywhere.

The final three-subject standard

A small group should not make English, Mathematics and Science look more alike. It should make the important differences easier to teach because the tutor has enough visibility to respond to the subject and the learner at the same time.

Three students is the format. English language production, Mathematics reasoning and Science evidence are the educational work.

When the centre understands that distinction, 3-pax becomes more than a class-size claim. It becomes a way to run three different feedback systems at sufficiently high resolution that each learner can eventually carry more of the subject independently.

The final subject-fit question should be asked separately

Before enrolling the same child in several 3-pax subjects, ask whether each subject actually needs that level of feedback. A learner may need close English correction, broad Mathematics maintenance and no extra Science tuition at all.

The class format should be justified subject by subject. Convenience is not enough reason to create a three-subject package.

The learner should not carry one failure label across subjects

A child can struggle with English inference and still be an excellent mathematical reasoner. Another can write fluently and hold a Science misconception. Subject-specific diagnosis protects the learner from turning one weakness into a global identity.

The tutor should name the mechanism and the subject clearly.

The learner should also recognise cross-subject strengths

Strong reading can support Science. Careful checking from Mathematics can support English editing. Causal explanation from Science can strengthen argument writing. These transfers are useful when they are genuine.

The centre should leverage shared learning habits without pretending the subjects are interchangeable.

A family timetable should reflect subject priority

If one subject is in active repair and two are stable, the schedule should not automatically give equal tuition hours to all three. Maintenance can be lighter, independent or temporary.

This protects the child from a timetable that expands simply because multiple classes are available.

The class should become less necessary as self-regulation grows

English students should increasingly revise their own language. Mathematics students should identify and debug their own working. Science students should question their own claims and evidence.

The strongest small-group outcome is therefore not lifelong dependence on close tutor visibility. It is a learner who has internalised enough of the feedback loop to operate when the tutor is absent.

A final parent comparison across English, Mathematics and Science

• What does the tutor need to see in this subject?
• What does the child need to produce?
• What kind of feedback changes the next attempt?
• How is transfer tested?
• Which scaffold is being removed?
• What would justify reducing support?

If the answers are identical for all three subjects, the programme may be too generic.

The final Punggol 3-pax principle

A small class is not a teaching method by itself. It is a high-resolution environment. The educational method appears in what the tutor does with that resolution.

In English, use it to hear and revise language. In Mathematics, use it to inspect and debug reasoning. In Science, use it to expose and revise models from evidence.

When those subject differences are respected, three students can be small enough for close feedback, large enough for useful comparison and spacious enough for each learner to practise independence.

That is when 3-pax becomes educational design rather than a class-size slogan.

The final proof belongs outside the tuition room

After several months, the strongest evidence is not that the child can perform inside a familiar 3-pax lesson. It is that the English revision habit, Mathematics checking process or Science reasoning routine appears in school work without the tutor present.

This is where the different subject loops converge again. Every subject ultimately needs transfer. A small group creates the close feedback that can build the process, but the process must later survive a larger classroom, independent homework and formal assessment.

Parents should therefore look for subject-specific independence rather than one generic feeling that tuition is “helping”. The English learner should need less sentence-by-sentence rescue. The Mathematics learner should identify the first wrong step sooner. The Science learner should connect evidence to explanation without being prompted to use the keyword.

When those changes appear repeatedly, the class size has done its job. The learner is carrying more of the feedback system internally.

That is the final measure of a well-run 3-pax programme across English, Mathematics and Science: the subject differences remain visible, the feedback remains precise, and dependence shrinks as capability grows.

A final family safeguard is to review subject need separately at renewal time. If English still needs close feedback but Mathematics is stable and Science can be maintained independently, keep only the support that continues to earn its place. A small-group philosophy should not become a permanent three-subject package simply because the format is familiar.

The strongest programme therefore becomes more selective as the learner becomes more capable. Shared class-size principles remain—visibility, quick feedback, peer comparison and prompt fading—but the subject, frequency and intensity are allowed to change. This keeps the week sustainable and protects the educational meaning of 3-pax.

Three students is useful when the tutor can see more, diagnose better and return more responsibility to the learner. That standard should be proven separately in English, Mathematics and Science every term.

A final 3-pax principle is worth keeping across the estate: the smaller group should increase information about the learner, not simply reduce the number of chairs. English needs more visible language production, Mathematics needs more visible reasoning, and Science needs more visible models, evidence and claims. If the tutor cannot explain what the small group makes easier to diagnose in each subject, the format is being underused.

The format earns its place when feedback becomes faster, more precise and easier to fade. The learner should leave with stronger subject-specific self-correction and a smaller need for tutor rescue than at the start of the term.

Different subjects, different evidence, different feedback loops—one shared commitment to visible thinking, precise correction and growing learner independence.

Teach the subject, not the format.