The Tutor Handbook · Volume 0080 · Series ID THB-0080
The Tutor Handbook: Complete series index.
A student arrives with two correct methods.
School has taught one. Tuition has taught another. Both can produce the right answer. The child is now slower than before because every question has acquired an extra hidden task: Which adult system am I supposed to use here?
This is one of the least dramatic and most common ways tutoring can become educationally expensive. Nothing looks obviously wrong. The school method may be sound. The tuition method may be elegant. The tutor may even be able to explain why the tuition method is more general. Yet the learner is carrying two vocabularies, two sequences, two sets of written conventions, or two different interpretations of what a “good answer” looks like. The cost appears later as hesitation, inconsistent notation, unnecessary switching, weaker confidence and difficulty knowing which evidence a teacher will recognise.
The Curriculum Translation Check asks whether a tutor should align, translate, preserve, bridge or deliberately diverge when school and tuition are teaching the same underlying knowledge through different methods, language, representations or answer conventions.
The short answer is not “always follow school” and not “always teach the best method”. The tutor’s first job is to identify what kind of mismatch exists. Some differences are cosmetic and should be translated quickly. Some are legitimate alternative methods and should be connected so the learner can recognise their shared structure. Some reflect an assessment convention that the learner must be able to produce reliably in school. Some reveal a real weakness in the school-facing route and justify a carefully bounded tuition alternative. The professional standard is to prevent the learner from becoming the integration layer for two adults who have not done the integration themselves.
1. This is an interface problem before it is a method problem
When two systems meet, people often ask which one is correct. That question is sometimes necessary, but it is usually too early. First ask whether the systems are actually disagreeing about the underlying knowledge or merely expressing it differently.
One Mathematics teacher writes a ratio as 2:3 and builds a bar model. Another tutor moves directly into algebraic variables. One English teacher uses “point, evidence, explanation”; another tutor asks for “claim, textual proof, reasoning”. One Science teacher says “independent variable”; another uses “changed variable” because that is the language expected at a particular school stage. A learner can understand the same relationship while still being destabilised by the surface change.
The first question is therefore: What must remain invariant across the two systems? The mathematical relationship, causal mechanism, evidence requirement or communicative purpose should survive even when notation or terminology changes. If the invariant is visible, translation becomes possible. If the invariant is not visible, the learner may memorise two procedures without recognising that they solve the same problem.
This volume does not replace the wider owner How Teaching Works, which explains the larger diagnosis-to-independent-performance route. It also does not replace subject-specific hubs, school syllabuses or examination guidance. Its narrower job is the moment when a tutor has to decide what to do with a difference between school-facing instruction and tuition-facing instruction.
2. Name the mismatch before choosing the response
Useful curriculum translation begins with classification. There are at least five common forms of mismatch, and they should not be treated as though they were the same problem.
Vocabulary mismatch occurs when two adults use different names for closely related ideas. The learner may know the mechanism but fail to recognise that the labels refer to the same structure. This is often the easiest form to repair. Build a translation map and move on.
Representation mismatch occurs when school and tuition present the same idea through different external forms: bar model versus equation, prose explanation versus diagram, table versus graph. This can be productive if the learner is taught to translate between forms. It becomes harmful when the learner treats each representation as a separate topic.
Procedure mismatch occurs when both systems use different valid routes to the same result. Here the tutor must decide whether dual-method competence is useful now or whether one method should remain the operational default while the other is kept as a comparison or later extension.
Sequence mismatch occurs when tuition introduces ideas in a different order from school. This may be justified by prerequisites or diagnosis, but it can also leave the learner trying to coordinate two unfinished progressions.
Convention mismatch occurs when the underlying reasoning may be correct but the expected answer form, notation, terminology or level of explanation differs. This matters particularly when the school task is partly assessing whether the learner can communicate in an accepted disciplinary form.
Different mismatches need different responses. A vocabulary mismatch may need one sentence. A procedure mismatch may need a deliberate bridge lesson. A convention mismatch may require the learner to retain one school-facing production form even while understanding alternative routes more deeply.
3. Translation should come before replacement
Tutors sometimes see an awkward school method and immediately replace it. The replacement may be genuinely better for this learner. But replacement should usually come after translation, because the learner still has to survive the school environment where the original method, notation or vocabulary is used.
Translation means making the relationship explicit: “Your teacher’s method and this method are doing the same mathematical move here. In school, keep using the teacher’s notation because that is what your class is practising. I am showing you this second representation because it makes the relationship easier to see. We will connect them, not make you choose sides.”
That sentence performs several jobs. It respects school ownership. It tells the learner the second method has a purpose. It prevents the tuition method from quietly becoming a competing curriculum. Most importantly, it makes the invariant visible. The child is learning a concept that can be represented in more than one way, not memorising two unrelated adult preferences.
Replacement becomes more defensible when the inherited route is demonstrably failing the learner, the alternative addresses the actual weak link, the change can be tested, and the learner can still translate back into the school-facing form when required. This is where The Implementation Fidelity Check matters: do not declare a method ineffective if the learner never actually had a fair opportunity to use it as intended.
4. The school method may carry information the tutor cannot see
A tutor sees the worksheet. The school teacher sees the wider sequence: what came before, what language the class has established, what will be revisited next, what examples have been used, which misconceptions are common, and what answer conventions are being reinforced across several weeks.
This does not make school infallible. It means the tutor should be cautious about judging a method from one artefact. A procedure that looks inefficient in isolation may have been selected because it prepares for a later concept. A vocabulary choice that seems less technically precise may be developmentally appropriate for the current stage. A teacher may ask for a particular representation because the class is learning to connect structure to notation, not because that representation is the only valid one.
Before diverging, gather enough context. Ask the learner what the teacher said the method was for. Look at the worked example before and after the one causing difficulty. Check the current unit or school notes. If a current syllabus or examination rule is relevant, verify it from the official source rather than relying on memory or a tuition-centre convention.
The National Student Support Accelerator’s current guidance on tutoring alignment recommends access to the school’s scope and sequence, unit plans and instructional materials where possible, and frames tutoring as something that should complement and respond to classroom instruction. Its 2025 Tutoring Quality Standards classify curricular alignment and school engagement as areas where the evidence base is still partly emergent rather than universally settled. That distinction matters: alignment is a strong implementation principle, but it is not a licence to assume every school material should be copied without professional judgement.
5. Align exactly when the production convention is part of the task
There are times when the tutor should align very closely with the school-facing convention. If a learner knows the idea but repeatedly loses clarity because they use incompatible notation, the operational problem is not conceptual depth. It is production consistency.
Suppose Alicia understands algebraic equivalence but alternates between two notational systems taught by different adults. Her working becomes hard to read and she makes transcription errors while switching. In this case, teaching a third elegant notation is not helpful. The tutor should select the school-facing notation as the default, explicitly map the alternatives once, and remove unnecessary switching until the learner is stable.
Similarly, Beatrice may understand what counts as textual evidence but use a paragraph structure that conflicts with her school’s taught response routine. If the school routine is reasonable and helps the teacher see the learner’s reasoning, tuition should normally support it. The tutor can deepen the reasoning inside that frame without forcing Beatrice to convert every answer into a competing format.
Exact alignment is strongest when the difference is merely a communication convention, the convention is educationally reasonable, and using it reduces unnecessary coordination cost. The learner should not spend cognitive effort proving independence from a harmless school convention when that effort could be used on the actual reasoning.
6. Translate when the underlying idea is shared but the surface differs
Translation is the correct response when both systems are useful and the learner needs to see the connection.
Ciara’s Science teacher may use a simple causal chain in words while tuition uses a labelled diagram. The tutor can ask Ciara to move both ways: diagram to explanation, explanation to diagram. The purpose is not to make her bilingual in educational jargon. It is to prove that the underlying mechanism survives a change of representation.
This kind of translation can strengthen learning because the learner must identify what is structurally common. However, the tutor should not assume that every additional representation is beneficial. If the learner is still trying to understand the core concept, adding multiple forms can create extra load. Translation should answer a real need: confusion between systems, transfer across formats, or a requirement to recognise the same idea when it appears differently.
A practical translation sequence is: identify the school form; identify the tuition form; state the shared invariant; demonstrate one worked translation; ask the learner to translate a second case; then return to the school form and check whether clarity improved. The final step matters. Translation is not successful because the learner can admire two methods. It is successful when the learner can operate in the required environment with less confusion.
7. Deliberate divergence needs a reason stronger than tutor preference
Sometimes tuition should diverge. A learner may need a different representation to repair a misconception. A school sequence may move on while a prerequisite remains unstable. An alternative method may expose structure that the learner has been copying without understanding. A tutor may need to simplify a task temporarily so the weak component becomes visible.
But “I teach it differently” is not enough. Divergence should have a learner-specific reason. The tutor should be able to state what school-facing problem is being repaired, why the existing route is not sufficient for that learner now, what the alternative is supposed to change, and how the learner will return to school-compatible performance.
Consider a learner who has memorised a percentage formula but cannot identify the base quantity in changed questions. Tuition may temporarily move away from the school worksheet sequence and use bar models or concrete comparisons. That divergence is defensible because the representation targets the actual weak link. Once the learner can identify the relationship, tuition should reconnect the repaired understanding to the notation and problem forms used in school.
Divergence without a return path creates two curricula. Divergence with a clear repair and reintegration path can be exactly what good tutoring is for.
8. Do not teach two full procedures when one translation sentence will do
Tutors often over-teach differences. The learner asks, “Why does my teacher do it this way?” and receives a twenty-minute comparison of methods. That comparison may be intellectually interesting but educationally unnecessary.
Before expanding, ask whether the learner actually needs dual competence. If the school method is valid and working, the tuition method may only need to exist as a supporting representation. If the difference is terminology, make the equivalence explicit and stop. If the learner is already overloaded, the most professional response may be to remove the tutor’s preferred method from the active route entirely.
This is a form of instructional restraint. Expertise creates many possible explanations. It does not require using all of them. A tutor adds value by selecting the explanation that reduces uncertainty for the learner, not by displaying the number of alternative methods the tutor knows.
The same principle applies in English and Science. A child does not need three competing paragraph acronyms if one school-facing structure already supports sound reasoning. A child does not need every technically precise term immediately if the current school vocabulary is accurate enough for the developmental stage. Add resolution when it helps the learner think, not merely because more resolution is available.
9. Build a bridge lesson when the mismatch is materially affecting performance
When the mismatch has become a genuine source of error or hesitation, use a bounded bridge lesson rather than allowing the conflict to persist implicitly.
Start with a real school example. Ask the learner to show the school method exactly as they understand it. Preserve the attempt. Then ask what each step is doing, not merely what the next step is. This reveals whether the learner understands the method or is reproducing a remembered sequence.
Next, introduce the tuition representation only if it clarifies the hidden structure. Mark the points of equivalence explicitly. “This line in your teacher’s method is the same relationship shown by this bar.” “This phrase in your school paragraph is doing the same job as what I call the evidence link.” “Your teacher’s ‘changed variable’ is the same experimental role described more generally as the independent variable.”
Then translate back. Give a fresh school-style task and ask the learner to use the expected school-facing form. If the bridge worked, the learner should now use that form with greater understanding and less hesitation. If the learner can only succeed in the tuition representation, the bridge is incomplete.
Finally, delay the check. A same-session translation can feel obvious because both forms are visible. Return later with one changed example. Ask the learner to identify whether the difference is only surface or whether the underlying job has changed. This protects against the illusion that recognition of a bridge is the same as owning it.
10. Three-student tutorials require individual translation, not one universal compromise
In a three-student tutorial, the temptation is to standardise everyone onto one tuition method for management convenience. That may be sensible when all three genuinely share the same need. It becomes a problem when the method is only necessary for one learner.
Imagine Alicia’s school teaches one algebraic representation, Beatrice’s school uses another, and Ciara has already become fluent in both. A single “our tuition centre method” may make the class easier to run while increasing the integration burden for two students. The group should share the underlying concept and common discussion where useful, but individual learners may need different school-facing return routes.
This is consistent with the current evidence on small-group tuition: the Education Endowment Foundation’s evidence summary defines small-group tuition as two to five pupils and stresses that effectiveness depends on targeting support to specific pupil needs, with the quality of teaching and group composition mattering alongside size. The average evidence does not prove that any particular three-student arrangement is best for every subject or learner. It does support the principle that small-group advantage comes partly from closer matching of instruction to need.
Translation therefore remains learner-specific. Shared teaching can be efficient; shared confusion is not.
11. Parents should not force a winner between school and tuition
Parents sometimes encounter a method conflict and ask, “Which one should my child follow?” That is a reasonable question, but framing the issue as a contest can increase the child’s burden. The learner may start treating school and tuition as rival authorities and feel responsible for deciding which adult is right.
A better parent question is: “What should my child be able to understand across both approaches, and which form should they use in school right now?” This separates conceptual ownership from operational production.
If the tutor has deliberately introduced an alternative method, parents should know its job. Is it a temporary scaffold? A diagnostic representation? A more general method that will be connected back later? A genuine replacement because the previous route repeatedly failed? Without that explanation, home practice can accidentally reinforce the wrong version. A parent may insist on the tuition method at home while school is trying to stabilise a different but valid representation.
The parent does not need to become the curriculum arbitrator. The parent needs enough clarity to avoid creating a third system.
12. The learner should eventually own the translation
At first the tutor may perform the bridge: “These two methods are equivalent because…” Later the learner should be able to do it.
Ask: “Your teacher uses this form. I used that form last week. What is the same? What is different? Which one are you expected to produce here? If the numbers or wording change, what idea remains?” These questions move the learner from compliance with one adult to flexible understanding across contexts.
However, translation ownership should not be confused with constant method switching. A learner can understand multiple representations while still using one operational default for fluency. Expertise often includes knowing when not to switch.
The long-term goal is not a child who asks, “Which adult’s method is this?” The goal is a learner who can recognise the underlying structure, choose or produce the form required by the task, and explain why two valid routes agree when they do.
13. A practical Curriculum Translation Check
Before introducing or defending an alternative route, the tutor can run a compact check:
- What is the school form? Vocabulary, notation, method, representation, sequence or answer convention?
- What is the tuition form? Exactly what is different?
- What is invariant? Which concept, relationship, evidence requirement or communicative purpose should remain the same?
- Is the difference actually causing learner difficulty? Or is the tutor solving a conflict the learner does not have?
- Does the school convention carry legitimate assessment or communication value?
- Is tuition translating, supplementing, temporarily scaffolding or replacing?
- What evidence justifies divergence?
- What is the return path to school-compatible performance?
- Can the learner translate between the two forms?
- Can the learner still perform after a delay without both systems visible?
The check is deliberately simple. The point is not to create a new bureaucracy around every difference. The point is to stop casual tutor preference from becoming an extra curriculum the learner has to maintain.
14. What the research can and cannot support
Current high-impact tutoring guidance strongly favours coherence with classroom learning, structured instructional materials, ongoing formative assessment and responsive modification. The National Student Support Accelerator’s November 2025 Tutoring Quality Standards distinguish stronger research-based findings from research-informed and emergent standards. Its curricular-alignment standard is presented as emergent: alignment is recommended because it may reduce confusion and improve coherence, but direct causal evidence about every form of tutor–school alignment is still limited.
NSSA’s curriculum-alignment toolkit recommends access to classroom scope and sequence, relevant plans and materials, and regular review of alignment. It also explicitly recommends using student data to adjust instructional materials. That combination is important: alignment is not blind copying. It is responsive coordination.
The OECD Teaching Compass, published in May 2025, frames teachers as adaptive experts and co-designers rather than passive implementers. That is a policy framework, not a controlled test of private tutoring. Its relevance here is conceptual: professional judgement should operate inside a coherent educational system, not against it for the sake of autonomy.
Nothing in these sources proves that a particular Singapore tuition method, group size or translation routine will improve a particular child. The application in this article is an evidence-informed tutoring protocol. It should be judged by what happens to the learner: less confusion, stronger conceptual recognition, reliable school-facing performance and greater ability to move between representations without adult rescue.
15. Failure modes
Tutor superiority. The tutor treats a different method as evidence that school is wrong. The learner inherits an adult status contest.
School obedience. The tutor copies a school route even when current evidence shows the learner does not understand it. Alignment replaces diagnosis.
Method accumulation. Every alternative is preserved. The learner can list methods but cannot choose or execute one fluently.
Translation without return. The learner succeeds in the tuition representation but cannot use the school-facing form independently.
Convention inflation. The tutor treats a harmless terminology difference as a deep conceptual conflict and spends scarce lesson time resolving a problem that did not matter.
Premature divergence. A new method is introduced before the tutor understands why the school method is there or whether it was actually implemented properly.
Permanent scaffold. A translation device that was meant to clarify one idea becomes an additional routine the learner must carry forever.
16. Final principle
A tutor does not need to become a second school to be useful.
The strongest tutoring often does something more precise. It identifies where the learner’s current school-facing route is working, where it is confusing, where a different representation can reveal structure, and where an alternative method genuinely earns its place. Then it performs the translation work adults are capable of doing so the learner does not have to carry two unintegrated systems alone.
Preserve the concept. Respect the live educational environment. Translate surface differences before replacing routes. Diverge only when learner evidence justifies it. And always give the learner a clear way back to independent performance in the system where the learning must actually be used.
